No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Quantitative MR Imaging using “LiveWire” to Measure Tibiofemoral Articular Cartilage Thickness
Abstract
To assess the reliability and accuracy of manual and semi-automated segmentation methods for quantifying knee cartilage thickness. This study employed both manual and LiveWire-based semi-automated segmentation methods, ex vivo and in vivo, to measure tibiofemoral (TF) cartilage thickness.
The articular cartilage of a cadaver knee and a healthy volunteer's knee were segmented manually and with LiveWire from multiple 3T MR images. The cadaver specimen's cartilage thickness was also evaluated with a 3D laser scanner, which was assumed to be the gold standard. Thickness measurements were made within specific cartilage regions. The reliability of each segmentation method was assessed both ex vivo and in vivo, and accuracy was assessed ex vivo by comparing segmentation results to those obtained with laser scanning.
The cadaver specimen thickness measurements showed mean coefficients of variation (CVs) of 4.16%, 3.02%, and 1.59%, when evaluated with manual segmentation, LiveWire segmentation, and laser scanning, respectively. The cadaver specimen showed mean absolute errors versus laser scanning of 4.07% and 7.46% for manual and LiveWire segmentation, respectively. In vivo thickness measurements showed mean CVs of 2.71% and 3.65% when segmented manually and with LiveWire, respectively.
Manual segmentation, LiveWire segmentation, and laser scanning are repeatable methods for quantifying knee cartilage thickness; however, the measurements are technique-dependent. Ex vivo, the manual segmentation error was distributed around the laser scanning mean, while LiveWire consistently underestimated laser scanning by 8.9%. Although LiveWire offers repeatability and decreased segmentation time, manual segmentation more closely approximates true cartilage thickness, particularly in cartilage contact regions.
... Inter-scan CV ± SD: 2.7 ± 1.0% Bowers et al. (2008) Live wire 7 cadaver and 7 human scans 3D FLASH WE Cadaver CV: 3.02% in vivo CV: 3.65% ...
... User-specific markers was manually inserted through scribbles or boundary points. The priori knowledge was essential to initialize and modify the segmentation (Bowers et al. 2008;Gan et al. 2017Gan et al. , 2019Gougoutas et al. 2004) as well as to compensate the lacking of cost function (Kashyap et al. 2018). Consequently, graph-based methods were often plagued with overdependence on user interaction in order to attain desirable segmentation results. ...
... Overall, classical cartilage segmentation models' DSC ranged from 70 to 88% while classical bone segmentation models' DSC ranged from 90 to 97%. In semiautomatic cartilage segmentation models, both DSC (Gan et al. 2017(Gan et al. , 2019Liukkonen et al. 2017b;Shim et al. 2009a) and CV Bowers et al. 2008;Brem et al. 2009;Duryea et al. 2007;Gougoutas et al. 2004;Lynch et al. 2000;Stammberger et al. 1999;Tang et al. 2006) were two equally important evaluation metrics. DSCs of semiautomatic cartilage segmentation models ranged from 80 to 94% and CV was measured from the perspective of different observers (inter-observer), within observer (intra-observer), different subjects, and different scans. ...
Knee osteoarthritis is a major diarthrodial joint disorder with profound global socioeconomic impact. Diagnostic imaging using magnetic resonance image can produce morphometric biomarkers to investigate the epidemiology of knee osteoarthritis in clinical trials, which is critical to attain early detection and develop effective regenerative treatment/ therapy. With tremendous increase in image data size, manual segmentation as the standard practice becomes largely unsuitable. This review aims to provide an in-depth insight about a broad collection of classical and deep learning segmentation techniques used in knee osteoarthritis research. Specifically, this is the first review that covers both bone and cartilage segmentation models in recognition that knee osteoarthritis is a “whole joint” disease, as well as highlights on diagnostic values of deep learning in emerging knee osteoarthritis research. Besides, we have collected useful deep learning reviews to serve as source of reference to ease future development of deep learning models in this field. Lastly, we highlight on the diagnostic value of deep learning as key future computer-aided diagnosis applications to conclude this review.
... However, the accuracy of the geometry obtained from this method is limited by the resolution of the MRI and the accuracy of the segmentation technique. While sub-pixel accuracy in the segmentation has been reported [21], Van Den Broeck et al. [22] found the errors of the three-dimensional reconstruction from MR images can be as large as two pixels. It is common in MRI to use a slice thickness larger than the in-plane resolution [2,19,20]. ...
... The effects of the geometric uncertainties caused by the limitations of the MRI resolution and accuracy of the segmentation were investigated by varying the meniscal geometry of the intact models (figure 2). As mentioned in the Introduction, the accuracy of segmentation may vary [21,22]. Therefore, a large geometric variation in the meniscus was investigated to cover a wide range of uncertainties. ...
... Most of the current MRI-based FE knee models use a resolution between 0.2 and 0.6 mm [2,9,20]. It has been reported that the errors in the three-dimensional reconstruction from MR images varies from sub-pixel to two pixels [21,22]. Therefore, the errors in the cartilage geometry of the FE knee models resulting from the MRI scanning and segmentation can be easily larger than 0.2 mm, which are at the same magnitude as the RMSE of the P5 imagebased model (0.21-0.36 mm). ...
The geometry of the articular components of the knee is an important factor in predicting joint mechanics in computational models. There are a number of uncertainties in the definition of the geometry of cartilage and meniscus, and evaluating the effects of these uncertainties is fundamental to understanding the level of reliability of the models. In this study, the sensitivity of knee mechanics to geometric uncertainties was investigated by comparing polynomial-based and image-based knee models and varying the size of meniscus. The results suggested that the geometric uncertainties in cartilage and meniscus resulting from the resolution of MRI and the accuracy of segmentation caused considerable effects on the predicted knee mechanics. Moreover, even if the mathematical geometric descriptors can be very close to the imaged-based articular surfaces, the detailed contact pressure distribution produced by the mathematical geometric descriptors was not the same as that of the image-based model. However, the trends predicted by the models based on mathematical geometric descriptors were similar to those of the imaged-based models.
... Established anatomic risk factors for human ACL injury were measured using the MR image stack, including ACL size (length, cross-sectional area and volume) [13,18,32], medial femoral and tibial cartilage thickness [11,12], posterior slope of the tibial plateau in the medial and lateral compartments [24,25], and intercondylar notch size (width, area, and index) [3]. The bicondylar width also was quantified as a measure of overall knee size [58]. ...
... Once segmented, the cartilage was reconstructed, and a 3-D voxel model of each structure was created. The voxel models were wrapped with a triangular mesh to create virtual solid models [11]. Mean cartilage thickness measurements were measured in the weightbearing regions of interest of the medial femoral condyle and medial tibial plateau as previously described [11]. ...
... The voxel models were wrapped with a triangular mesh to create virtual solid models [11]. Mean cartilage thickness measurements were measured in the weightbearing regions of interest of the medial femoral condyle and medial tibial plateau as previously described [11]. The regions of interests were defined using an established technique [11] to assess morphologic features of the cartilage in humans and pigs [12]. ...
Background:
Animal models have long been considered an important modality for studying ACL injuries. However, to our knowledge, the value of these preclinical models to study sex-related phenomena associated with ACL injury and recovery has not been evaluated.
Questions/purposes:
We asked whether (1) prominent anatomic and (2) biomechanical factors differ between female and male porcine knees, particularly those known to increase the risk of ACL injury.
Methods:
Eighteen intact minipig knees (nine males, nine females) underwent MRI to determine the femoral bicondylar width, intercondylar notch size (width, area and index), medial and lateral tibial slope, ACL size (length, cross-sectional area, and volume), and medial compartment tibiofemoral cartilage thickness. AP knee laxity at 30°, 60°, and 90° flexion and ACL tensile structural properties were measured using custom-designed loading fixtures in a universal tensile testing apparatus. Comparisons between males and females were performed for all anatomic and biomechanical measures. The findings then were compared with published data from human knees.
Results:
Female pigs had smaller bicondylar widths (2.9 mm, ratio=0.93, effect size=-1.5) and intercondylar notches (width: 2.0 mm, ratio=0.79, effect size=-2.8; area: 30.8 mm2, ratio=0.76, effect size=-2.1; index: 0.4, ratio=0.84, effect size=-2.0), steeper lateral tibial slope (4.3°, ratio=1.13, effect size=1.1), smaller ACL (length: 2.7 mm, ratio=0.91, effect size=-1.1; area: 6.8 mm2, ratio=0.74, effect size=-1.5; volume: 266.2 mm3, ratio=0.68, effect size=-1.5), thinner medial femoral cartilage (0.4 mm, ratio=0.8, effect size=-1.1), lower ACL yield load (275 N, ratio=0.81, effect size=-1.1), and greater AP knee laxity at 30° (0.7 mm, ratio=1.32, effect size=1.1) and 90° (0.5 mm, ratio=1.24, effect size=1.1) flexion compared with their male counterparts. These differences were significant for all parameters (p≤0.04). Observed sex-related differences were similar to those reported for the human knee.
Conclusions:
Significant differences exist between knees of male and female pigs with respect to prominent anatomic and biomechanical factors. Our findings strongly agreed with published data regarding human knees.
Clinical relevance:
The findings highlight the use of the porcine large animal model to study the role of sex on ACL injuries and surgical outcome. This validated preclinical model may facilitate the development of novel, sex-specific interventions to prevent and treat ACL injuries for male and female patients.
... Four steel spheres were firmly glued to the cortical surfaces at the midshaft of the radius, distant from the distal end to serve as landmarks for integrating each digitized data set and surface registration 12e16 . To prevent cartilage from drying, the articulating end of the radius was covered by gauze moistened with physiological saline, except when cartilage measuring 17 . ...
... First, we used formalin-embalmed cadavers. However, we do not expect this to impact our findings, as formalin fixation has no measurable effect on bone or cartilage 8,12,14,15,17,45,46 . Furthermore, our measurements were in line with those in prior in vivo or fresh-frozen cadaveric studies 27,31,32 . ...
Objective
To quantify the spatial distributions of cartilage and subchondral bone thickness of the distal radius.
Design
Using 17 cadaveric wrists, three types of 3-dimensional models were created: a cartilage-bone model, obtained by laser scanning; a bone model, rescanned after dissolving the cartilage; and a subchondral bone model, obtained using computed tomography. By superimposing the bone model onto the cartilage-bone and the subchondral bone models, the cartilage and subchondral bone thickness were determined. Measurements along with the spatial distribution were made at fixed anatomic points including the scaphoid and lunate fossa, sigmoid notch and interfossal ridge, and compared at each of these four regions.
Results
Cartilage thickness of the interfossal ridge (0.89±0.23mm) had a larger average thickness compared to that of the scaphoid fossa (0.70±0.18mm; p=0.004), lunate fossa (0.75±0.17mm; p=0.044) and sigmoid notch (0.64±0.13mm; p<0.001). Subchondral bone was found to be thickest at the scaphoid (2.18±0.72mm) and lunate fossae (1.94±0.93mm), which were both thicker than that of sigmoid notch (1.63±1.06mm: vs. scaphoid fossa, p=0.020) or interfossal ridge (1.54±0.84mm: vs. scaphoid fossa, p=0.004; vs. lunate fossa, p=0.048). In the volar-ulnar sub-regions of the scaphoid and lunate fossa, the subchondral bone thickened.
Conclusions
Our data can be applied when treating distal radius fractures. Cartilage thickness was less than 1mm across the articular surface, which may give an insight into threshold for an acceptable range of step-offs. The combined findings of subchondral bone appreciate the importance of the volar-ulnar corner of the distal radius in the volar locking plate fixation.
... In fact, cartilage deformation tends to a plateau of about 5% of the original thickness, regardless of the performed activities [56]. Assuming a mean cartilage thickness of 2.5 mm for both the tibia and femur [89,90] will lead to a 0.25 mm of indentation. This value is below the usual in plane resolution of a MR scan, which can be estimated around 0.3 mm [90]. ...
... Assuming a mean cartilage thickness of 2.5 mm for both the tibia and femur [89,90] will lead to a 0.25 mm of indentation. This value is below the usual in plane resolution of a MR scan, which can be estimated around 0.3 mm [90]. Using this value for Δ would make the algorithm too sensitive to the quality of the MRI segmentation. ...
Objective:
The knowledge of individual joint motion may help to understand the articular physiology and to design better treatments and medical devices. Measurements of in-vivo individual motion are nowadays invasive/ionizing (fluoroscopy) or imprecise (skin markers). We propose a new approach to derive the individual knee natural motion from a three-dimensional representation of articular surfaces.
Methods:
We hypothesize that tissue adaptation shapes articular surfaces to optimize load distribution. Thus, the knee natural motion is obtained as the envelope of tibiofemoral positions and orientations that minimize peak contact pressure, i.e. that maximize joint congruence. We investigated four in-vitro and one in-vivo knees. Articular surfaces were reconstructed from a reference MRI. Natural motion was computed by congruence maximization and results were validated versus experimental data, acquired through bone im-planted markers, in-vitro, and single-plane fluoroscopy, in-vivo.
Results:
In two cases, one of which in-vivo, maximum mean absolute error stays below 2.2 and 2.7 mm for rotations and translations, respectively. The remaining knees showed differences in joint internal rotation between the reference MRI and experimental motion at 0 flexion, possibly due to some laxity. The same difference is found in the model predictions, which, however, still replicate the individual knee motion.
Conclusion:
The proposed approach allows the prediction of individual joint motion based on non-ionizing MRI data.
Significance:
This method may help to characterize healthy and, by comparison, pathological knee behavior. Moreover, it may provide an individual reference motion for the personalization of musculoskeletal models, opening the way to their clinical application.
... To obtain baseline data for measurements of cartilage, the cartilage dissolution technique was used to create bone models without cartilage [25][26][27][28][29][30][31][32] . The joint surfaces of the specimens were soaked in 6.0% sodium hypochlorite for 12 to 24 hours to Diagrams of 3-D models, including a cartilage-bone model ( Fig. 1-A), a bone model ( Fig. 1-B), a cartilage model ( Fig. 1-C), and a cartilage model with the cartilage thickness information coded as a thickness map ( Fig. 1-D). ...
... First, we used formalin-embalmed cadavers. Formalin fixation can lead to dehydration of the cartilage; however, several researchers have reported that formalin fixation has no measurable effects on cartilage thickness or on geometric configuration within a joint 21,23,25,31,55,56 . Furthermore, our measurements corresponded with those in previous in vivo cartilage studies [16][17][18][19] . ...
Background: During elbow procedures, reconstruction of the joint (including the articular cartilage) is important in order to restore elbow function; however, the regional distribution of elbow cartilage is not completely understood. The purpose of the present study was to investigate the 3-dimensional (3-D) distribution patterns of cartilage thickness of elbow bones (including the distal part of the humerus, proximal part of the ulna, and radial head) in order to elucidate the morphological relationship among them.
Methods: Two 3-D surface models were created with use of a laser scanner: (1) a cartilage-bone model based on 20 elderly cadaveric elbows exhibiting normal cartilaginous conditions and (2) a bone model that was created after dissolving the cartilage. The 2 models were superimposed, and cartilage thickness was measured as the interval distance on the articular surface. Measurements were made at categorized anatomical points of the individual bones, and 3-D distributions on the entire articular surface were analyzed. The spatial relationship among individual bones was also assessed.
Results: In the distal part of the humerus, the cartilage was thickest in the intermediate region between the capitellum and the trochlea (mean [and standard deviation], 1.27 ± 0.17 mm); in the proximal part of the ulna, it was thickest at the anterolateral edge of the coronoid (2.20 ± 0.39 mm) and the anteroproximal edge of the proximal sigmoid notch (2.49 ± 0.55 mm); and in the radial head, it was thickest at the articular zone on the rim circumference within the dish (1.10 ± 0.17 mm) and on the proximal circumference around the side (1.02 ± 0.17 mm) (p < 0.001 for all). These thicker cartilage regions gathered on the joint center, facing each other.
Conclusions: The present study demonstrated regional variations in elbow cartilage thickness. The combined findings in individual bones showed “cartilage gathering” at the center of the elbow joint, which we believe to be a novel anatomical finding.
Clinical Relevance: An enhanced understanding of elbow cartilage geometry will provide additional insights into elbow procedures in elderly individuals, such as hemiarthroplasties, in which anatomical contours could help to restore normal joint function and improve postoperative outcomes.
... The issue of some soft tissue left on the bones was then addressed. The bones were soaked in a 5.25% bleach solution for approximatley 4 hours in order to remove soft tissue and cartilage (Bowers et al., 2008;DeVries et al., 2008). Bones were stirred periodically to change the portion of the bone touching the surface of the container and therefore not being exposed to the solution. ...
... The volume of the bones were then measured by water displacement. After the measurements, the bones were soaked in a 5.25% bleach solution to chemically remove the remaining soft tissue (Bowers et al., 2008;DeVries et al., 2008). The volumes were again measured by water displacement. ...
Abstract Lisfranc injuries affect at least 1 out of every 55,000 people each year. Although they are rare foot injuries, their effects can be devastating. 20-40% of Lisfranc injuries are missed upon first presentation. This increases the number of poor outcomes, resulting in a disproportionate number of malpractice lawsuits and compensation claims. The Kingston Brace was designed to support an injured foot during a CT scan with the goal of providing the diagnosing physician with the best diagnostic information possible. A prototype was designed and built to support a variety of foot orientations in order to determine which orientation is optimal for Lisfranc joint CT scanning. Three fresh frozen cadaver feet were put through several experiments with the Brace. The 2D diagnostic quality of CT scans taken using the Kingston Brace was compared against that of CT scans using the existing protocol. The Kingston Brace allows for a greater visualization of the injured Lisfranc joint than the existing CT protocol. Lisfranc joint spreading was used as a measure of potential pain in injured patients. The joint spreading was minimal, suggesting that injured patients would not feel as much additional pain during imaging. Also, the adoption of the Kingston Brace resulted in no change in the morphological parameters resulting from more advanced 3D analysis. The experimentally determined optimal Kingston Brace orientation was found to be 9, of eversion. These orientations can be incorporated into the next generation of Kingston Brace design. i Acknowledgments I would like to thank my two co-supervisors, Dr. Chris Mechefske and Dr. Kevin Deluzio.
... The co-first author (Z.Z.) of this paper developed an automatic image analysis program with Matlab R2016a (Mathworks, Natick, MA, USA) to segment each chest wall component area by using interactive automatic segmentation algorithm named Livewire, which was widely used in medical image segmentation [17,18]. We improved the Livewire algorithm to make it become faster and more accurate in our previous work [19]. ...
Objective
By applying computer image processing technology, this study aims to propose better biometric parameters of coronary computed tomography angiography (CCTA) by evaluating correlations between image quality parameters, the body mass index (BMI), and parameters of chest wall components.Methods
One hundred and seventeen subjects (59 males, 58 females, M = 59.3 years, SD = 10.2 years) who underwent CCTA were recruited. A Matlab program was used to measure the chest wall components in chest imaging automatically. In the parameters of the chest wall components, the gray weighted area of the chest wall (ACWgray weighted) was proposed as a new parameter with consideration of the area and CT attenuation of each solid tissue (fat, muscle, and bone). Image quality parameters [image noise, signal to noise ratio, and contrast to noise ratio] were measured on the slices of the aortic root and the maximum heart. The Shapiro–Wilk test was performed to evaluate data distribution. Correlation analyses were conducted to investigate relationships between image quality parameters, the BMI, and parameters of chest wall components. Linear correlation coefficients were used as indicators of the strength of the relationships.ResultsThe gray weighted average area of the chest wall (aACWgray weighted) and the BMI were correlated with the image quality parameters on the slices of the aortic root and the maximum heart (p < 0.05). The correlation coefficients with image noise were 0.635 and 0.516 on the slices of the aortic root, which were 0.672 and 0.543 on the slices of the maximum heart, respectively. Among all the parameters, aACWgray weighted showed the strongest correlation with image noise on both slices.Conclusions
The average quantitative parameters of the chest wall components, particularly aACWgray weighted showed the strongest correlations with all the image quality parameters. Hence, aACWgray weighted can be proposed as a better patient-specific predictor than the BMI of the image quality parameters in CCTA.
... Research has reported that deformation models or active counters can be modelled based on either parametric or geometric information available in the image. Segmentation from MR images is an active field of research where the segmentation is being performed using several deformation models, such as active con- tours [42,[82][83][84], Snake Contours [71,85], Statistical Shape Models (SSM) [37,64], Active Shape Models (ASM) [37,86], Global Statistical Model (GSM), Active Appearance Models (AAM) [15,31,37], and Statistical Deformation Models (SDM). Of these methods, ASM and SSM are most commonly used for segmentation of AC from 3D MR images. ...
Articular cartilage (AC) is a flexible and soft yet stiff tissue that can be visualized and interpreted using magnetic resonance (MR) imaging for the assessment of knee osteoarthritis. Segmentation of AC from MR images is a challenging task that has been investigated widely. The development of computational methods to segment AC is highly dependent on various image parameters, quality, tissue structure, and acquisition protocol involved. This review focuses on the challenges faced during AC segmentation from MR images followed by the discussion on computational methods for semi/fully automated approaches, whilst performances parameters and their significances have also been explored. Furthermore, hybrid approaches used to segment AC are reviewed. This review indicates that despite the challenges in AC segmentation, the semi-automated method utilizing advanced computational methods such as active contour and clustering have shown significant accuracy. Fully automated AC segmentation methods have obtained moderate accuracy and show suitability for extensive clinical studies whilst advanced methods are being investigated that have led to achieving significantly better sensitivity. In conclusion, this review indicates that research in AC segmentation from MR images is moving towards the development of fully automated methods using advanced multi-level, multi-data, and multi-approach techniques to provide assistance in clinical studies.
... However, cartilage surface models segmented from MR tended to over-estimate cartilage thickness for both femur and tibia by 1.15 mm. Such deviation is significant, considering overall cartilage thickness between femur and tibia at the joint is about 6 mm (25). ...
Background:
The aim of this study is to evaluate the accuracy of bone cuts and the resultant alignment, using the MyKnee patient specific cutting blocks.
Methods:
We retrospectively reviewed 132 patients undergoing primary TKR for osteoarthritis by one single surgeon. The operative time, the preoperative Hip-Knee-Ankle (HKA) axis based on the CT-scan, the postoperative HKA axis based on long axis standing x-rays, the planned and the actual size of the femoral and the tibial components, and the number of the recuts which has been made intraoperative were measured.
Results:
The average preoperative HKA axis was 177.50 (range 163.50 to 1940), whereas the average postoperative HKA axis was 179.40 (range 177.10 to 182.70). No outliers were reported in the study (0%). Intraoperatively, 4 femoral components (3.03%), and 7 tibial components (5.30%) applied to the patients were different than the planned size. There was no need of recuts in any of our cases intraoperatively.
Conclusion:
The MyKnee system evaluated in this study was shown to be remarkable reliable in the coronal plane alignment, and the prediction of the component size. However, further studies are needed to determine whether there are any clinically important improvements in outcomes or patient satisfaction when using patient-specific cutting blocks for TKA.
... On the other side, fissures could not sufficiently be detected by the MRI in our study, so its contribution to the MRI score was insignificant. The fact that the articular cartilage of the Minipig knee joints used in our study is thinner [87,123] than the cartilage of human knee joints [123,124] may be another explanation for the difficulties in finding the discovered histological degeneration in the MRI images. Wei et al. [6] showed differences between the control group and the treated group in quantitative MRI analysis by delayed gadolinium enhanced MRI of cartilage with the dGEMRIC technique over a time period of 2 to 6 weeks with a positive correlation of the GAG content to T1, Gd values and negative correlation to T2 values [6]. ...
Introduction
The Göttingen Minipig (GM) is used as large animal model in articular cartilage research. The aim of the study was to introduce osteoarthritis (OA) in the GM by resecting the anterior cruciate ligament (ACLR) according to Pond and Nuki, verified by histological and magnetic resonance imaging (MRI) scoring as well as analysis of gene and protein expression.
Materials and Methods
The eight included skeletally mature female GM were assessed after ACLR in the left and a sham operation in the right knee, which served as control. 26 weeks after surgery the knee joints were scanned using a 3-Tesla high-field MR tomography unit with a 3 T CP Large Flex Coil. Standard proton-density weighted fat saturated sequences in coronal and sagittal direction with a slice thickness of 3 mm were used. The MRI scans were assessed by two radiologists according to a modified WORMS-score, the X-rays of the knee joints by two evaluators. Osteochondral plugs with a diameter of 4mm were taken for histological examination from either the main loading zone or the macroscopic most degenerated parts of the tibia plateau or condyle respectively. The histological sections were blinded and scored by three experts according to Little et al. Gene expression analysis was performed from surrounding cartilage. Expression of adamts4, adamts5, acan, col1A1, col2, il-1ß, mmp1, mmp3, mmp13, vegf was determined by qRT-PCR. Immunohistochemical staining (IH) of Col I and II was performed. IH was scored using a 4 point grading (0—no staining; 3-intense staining).
Results and Discussion
Similar signs of OA were evident both in ACLR and sham operated knee joints with the histological scoring result of the ACLR joints with 6.48 ± 5.67 points and the sham joints with 6.86 ± 5.84 points (p = 0.7953) The MRI scoring yielded 0.34 ± 0.89 points for the ACLR and 0.03 ± 0.17 for the sham knee joints. There was no correlation between the histological and MRI scores (r = 0.10021). The gene expression profiles as well as the immunohistochemical findings showed no significant differences between ACLR and sham knee joints. In conclusion, both knee joints showed histological signs of OA after 26 weeks irrespective of whether the ACL was resected or not. As MRI results did not match the histological findings, MRI was obviously unsuitable to diagnose the OA in GM. The analysis of the expression patterns of the 10 genes could not shed light on the question, whether sham operation also induced cartilage erosion or if the degeneration was spontaneous. The modified Pond-Nuki model may be used with reservation in the adult minipig to induce an isolated osteoarthritis.
... MRI sequences were segmented in Mimics software (using the LiveWire function) to produce surface models of femur and tibia articular cartilage and bones [38]. Boolean operations were performed between tibia and femur cartilage segmentation masks to ensure no overlap occurred between models. ...
Combining accurate bone kinematics data from biplane radiography with cartilage models from magnetic resonance imaging, it is possible to estimate tibiofemoral cartilage contact area and centroid location. Proper validation of such estimates, however, has not been performed under loading conditions approximating functional tasks, such as gait, squatting, and stair descent. The goal of this study was to perform an in vitro validation to resolve the accuracy of cartilage contact estimations in comparison to a laser scanning gold standard. Results demonstrated acceptable reliability and accuracy for both contact area and centroid location estimates. Root mean square errors in contact area averaged 8.4% and 4.4% of the medial and lateral compartmental areas, respectively. Modified Sorensen-Dice agreement scores of contact regions averaged 0.81 ± 0.07 for medial and 0.83 ± 0.07 for lateral compartments. These validated methods have applications for in vivo assessment of a variety of patient populations and physical activities, and may lead to greater understanding of the relationships between knee cartilage function, effects of joint injury and treatment, and the development of osteoarthritis.
Copyright © 2015 IPEM. Published by Elsevier Ltd. All rights reserved.
... This value is below the one usual for the in plane resolution of a magnetic resonance imaging (MRI) scan, which can be estimated around 0.3 mm. 39 Furthermore, in order to observe the effect of a deformation, two pictures for the same joint are needed, one in an unloaded condition and other in a loaded condition. This would increase both the cost and the complexity of the measurements. ...
In the medical world, the term "congruence" is used to describe by visual inspection how the articular surfaces mate each other, evaluating the joint capability to distribute an applied load from a purely geometrical perspective. Congruence is commonly employed for assessing articular physiology and for the comparison between normal and pathological states. A measure of it would thus represent a valuable clinical tool. Several approaches for the quantification of joint congruence have been proposed in the biomechanical literature, differing on how the articular contact is modeled. This makes it difficult to compare different measures. In particular, in previous articles a congruence measure has been presented which proved to be efficient and suitable for the clinical practice, but it was still empirically defined. This article aims at providing a sound theoretical support to this congruence measure by means of the Winkler elastic foundation contact model which, with respect to others, has the advantage to hold also for highly conforming surfaces as most of the human articulations are. First, the geometrical relation between the applied load and the resulting peak of pressure is analytically derived from the elastic foundation contact model, providing a theoretically sound approach to the definition of a congruence measure. Then, the capability of congruence measure to capture the same geometrical relation is shown. Finally, the reliability of congruence measure is discussed.
... The 3D models were exported in STL format to a computer running the computer-aided design (CAD) software (PowerSHAPE, Delcam, UK) to determine the circumference and angle of the ROI. The validation of this method and the use of this computer program for measuring cartilage volume and bone areas have previously been described [6,15]. From our previously unpublished study, reliability tests for intra-class correlation coefficients between assessors and intra-assessor were 0.76 and 0.89, respectively. ...
High tibial osteotomy (HTO) is a surgical procedure used to correct abnormal mechanical loading of the knee joint; additionally, intra-articular hyaluronic acid injections have been shown to restore the viscoelastic properties of synovial fluid and balance abnormal biochemical processes. It was hypothesized that combining HTO with intra-articular hyaluronic acid injections would have benefit to improve the cartilage volume of knee joints.
Forty patients with medial compartment knee osteoarthritis (OA) were randomly placed into 1 of 2 groups. The study group (n = 20) received 2 cycles (at 6-month intervals) of 5 weekly intra-articular hyaluronic acid injections after HTO operation. The control group (n = 20) did not receive any intra-articular injections after HTO surgery. Cartilage volume (primary outcome) was assessed by magnetic resonance imaging (MRI) pre-operatively and 1 year post-operatively. Treatment efficacy (secondary outcomes) was evaluated with the Western Ontario and McMaster Universities OA Index (WOMAC) and by the comparison of the total rescue medication (paracetamol/diclofenac) used (weeks 6, 12, 24, 48).
MRI studies showed a significant increase in total cartilage volume (p = 0.033), lateral femoral cartilage volume (p = 0.044) and lateral tibial cartilage volume (p = 0.027) in the study group. Cartilage volume loss was detected at the lateral tibial plateau in the control group. There were significant improvements after surgery in both groups for all subscales of WOMAC scores (p < 0.001) compared to the baseline. However, no difference was found between the two groups. The study group had significantly lower amounts of diclofenac consumption (p = 0.017).
Based on the findings of the present study, intra-articular hyaluronic acid injections may be beneficial for increasing total cartilage volume and preventing the loss of lateral tibiofemoral joint cartilage after HTO.
Therapeutic study, Level I.
... A number of previous studies have used MR imaging to measure articular cartilage volume and thickness in human subjects (Bowers et al., 2008a;Cohen et al., 1999;Eckstein et al., 2005a;Van Ginckel et al., 2011b;Waterton et al., 2000). The methodology used in the present study has been previously validated to measure cartilage thickness distributions in the tibiofemoral joint ( Van de Velde et al., 2009). ...
Due to the biphasic viscoelastic nature of cartilage, joint loading may result in deformations that require times on the order of hours to fully recover. Thus, cartilaginous tissues may exhibit cumulative strain over the course of each day. The goal of this study was to assess the magnitude and spatial distribution of strain in the articular cartilage of the knee with daily activity. Magnetic resonance (MR) images of 10 asymptomatic subjects (six males and four females) with mean age of 29 years were obtained at 8:00 AM and 4:00 PM on the same day using a 3T magnet. These images were used to create 3D models of the femur, tibia, and patella from which cartilage thickness distributions were quantified. Cartilage thickness generally decreased from AM to PM in all areas except the patellofemoral groove and was associated with significant compressive strains in the medial condyle and tibial plateau. From AM to PM, cartilage of the medial tibial plateau exhibited a compressive strain of -5.1±1.0% (mean±SEM) averaged over all locations, while strains in the lateral plateau were slightly lower (-3.1±0.6%). Femoral cartilage showed an average strain of -1.9±0.6%. The findings of this study show that human knee cartilage undergoes diurnal changes in strain that vary with site in the joint. Since abnormal joint loading can be detrimental to cartilage homeostasis, these data provide a baseline for future studies investigating the effects of altered biomechanics on diurnal cartilage strains and cartilage physiology.
... MRI-based quantitative cartilage thickness measurement techniques have become increasingly important in clinical applications as well as in fundamental research. Although such studies have been performed for approximately a decade, most of them have focus on the knee joint (7)(8)(9)(10)(11). There have been, to date, very few studies on the human hip due to a lack of an appropriate MRI-acquisition protocol that specifically targets the ball and socket hip joint. ...
To test the accuracy of the most widely used technique based on edge detection for thickness measurement of the hip joint cartilage in MR images, and to improve the measurement accuracy by developing a new measurement method based on a model of the MRI process.
MRI was performed in 3 normal cadaver hips, 25 hips of normal volunteers, and 25 hips of patients with osteoarthritis. In general, thickness was defined as the distance between the two sides of the hip cartilage along the normal directions of the cartilage surfaces. In this article this is referred to as the "edge detection method." A theoretical simulation analysis revealed that the edge detection method considerably underestimated the cartilage thickness of the hip joint. A new measurement method based on a model of the MR imaging process was accordingly proposed for correcting the measurement errors.
In the experiment using the cadaver hips, anatomical measurement of cartilage thickness was used as reference standard. For measurements at 35 sites, the proposed model-based method gave results similar to those presented from anatomic section, while the edge detection method gave underestimation compared with the anatomic thickness. The underestimation biases for the edge detection method were consistent with the biases predicted by theoretical simulation. In the experiment using the hips of volunteers and patients, the edge detection result was an underestimation compared with the result generated by using the model-based method.
The edge detection method underestimated the hip cartilage thickness in MR images. The proposed model-based method was more accurate than the edge detection method for thickness measurement of the hip cartilage.
Segmentation of medical images into different tissue types is essential for many advancements in orthopaedic research; however, manual segmentation techniques can be time- and cost-prohibitive. The purpose of this work was to develop a semi-automatic segmentation algorithm that leverages gradients in spatial intensity to isolate the patella bone from magnetic resonance (MR) images of the knee that does not require a training set. The developed algorithm was validated in a sample of four human participants (in vivo) and three porcine stifle joints (ex vivo) using both magnetic resonance imaging (MRI) and computed tomography (CT). We assessed the repeatability (expressed as mean ± standard deviation) of the semi-automatic segmentation technique on: (1) the same MRI scan twice (Dice similarity coefficient = 0.988 ± 0.002; surface distance = − 0.01 ± 0.001 mm), (2) the scan/re-scan repeatability of the segmentation technique (surface distance = − 0.02 ± 0.03 mm), (3) how the semi-automatic segmentation technique compared to manual MRI segmentation (surface distance = − 0.02 ± 0.08 mm), and (4) how the semi-automatic segmentation technique compared when applied to both MRI and CT images of the same specimens (surface distance = − 0.02 ± 0.06 mm). Mean surface distances perpendicular to the cartilage surface were computed between pairs of patellar bone models. Critically, the semi-automatic segmentation algorithm developed in this work reduced segmentation time by approximately 75%. This method is promising for improving research throughput and potentially for use in generating training data for deep learning algorithms.
OBJECTIVE:
The present quantitative study aimed to assess the three-dimensional (3-D) cartilage wear patterns of the first metacarpal and trapezium in the advanced stage of osteoarthritis (OA) and compare cartilage measurements with radiographic severity.
DESIGN:
Using 19 cadaveric trapeziometacarpal (TMC) joints, 3-D cartilage surface models of the first metacarpal and trapezium were created with a laser scanner, and 3-D bone surface model counterparts were similarly created after dissolving the cartilage. These two models were superimposed, and the interval distance on the articular surface as the cartilage thickness was measured. All measurements were obtained in categorized anatomic regions on the articular surface of the respective bone, and we analyzed the 3-D wear patterns on the entire cartilage surface. Furthermore, we compared measurements of cartilage thickness with radiographic OA severity according to the Eaton grading system using Pearson correlation coefficients (r).
RESULTS:
In the first metacarpal, the cartilage thickness declined volarly (the mean cartilage thickness of the volar region was 0.32 ± 0.16 mm, whereas that of the dorsal region was 0.53 ± 0.18 mm). Conversely, the cartilage evenly degenerated throughout the articular surface of the trapezium. Measurements of the categorized regions where cartilage thinning was remarkable exhibited statistical correlations with radiographic staging (r = -0.48 to -0.72).
CONCLUSIONS:
Our findings indicate that cartilage wear patterns differ between the first metacarpal and trapezium in the late stage of OA. There is a need for further studies on cartilage degeneration leading to symptomatic OA in the TMC joint.
Background:
Osteochondral grafts are indicated for reconstructing the finger middle phalanx base when there is greater than 50% involvement of the articular surface and significant comminution. This study aims to compare the cartilage thickness of the distal surface of the hamate to the finger middle phalanx base to assess its suitability as an osteochondral graft.
Methods:
A 3-dimensional laser scanner and computer modelling techniques were utilized to determine the cartilage thickness of the distal surface of the hamate, and finger middle phalanx base using cadaver specimens. The mean, maximum, and coefficient of variation (CV%; a measure of uniformity of cartilage distribution), as well as cartilage distribution maps were determined.
Results:
The mean cartilage thickness of the hamate was 0.73 ± 0.08 mm compared to the average mean thickness of the finger middle phalanx base of 0.40 ± 0.12 mm. The maximum cartilage thickness of the hamate was 1.27 ± 0.14 mm compared to the average maximum of the finger middle phalanx base of 0.67 ± 0.14 mm. The CV% of the hamate was 27.8 ± 4.2 compared to the average CV% for the finger middle phalanx base of 26.6 ± 8.1. The hamate and finger middle phalanx base have maximum areas that were most frequently at or spanning the median ridge; however, this was more consistently observed with the hamate.
Conclusion:
The distal surface of the hamate is a suitable osteochondral graft with respect to cartilage thickness and distribution providing sufficient cartilage for reconstruction of the finger middle phalanx base.
L'arthrose est une maladie commune observée dans les populations vieillissantes caractérisée par une affection dégénérative du cartilage articulaire. Le diagnostic clinique de la maladie repose sur la radiographie conventionnelle. Cette technique permet de mettre en évidence les modifications de l'os liées à l'arthrose (géodes, condensation de l'os sous-chondral, et ostéophytes) mais n'offre pas une vision directe du cartilage. Grâce à sa résolution spatiale et son contraste tissulaire élevé, l'IRM individualise le cartilage et le différencie des structures adjacentes (os, tissu synovial, ménisques et liquide synovial). Nous avons mis à profit les potentialités de l'IRM à haut champ (7 Teslas) pour suivre, in vivo, les modifications du cartilage de l'articulation fémoro-tibiale chez le rat, au cours du processus de maturation/vieillissement ainsi que dans un modèle d'arthrose expérimentale (section du ligament croisé antérieur). Ces travaux ont montré une diminution du volume et des épaisseurs cartilagineuses liée à l'âge, tout comme des pertes chondrales fémorales et un oedème du cartilage tibial dans le genou arthrosique. Dans une seconde partie, nous avons appliqué les méthodes de cartographie T2 et de mesures volumiques (à 1,5 T) afin de déterminer les variations survenant au sein du cartilage rotulien humain vieillissant. Ces travaux attestent de la capacité de la cartographie T2 à détecter des modifications matricielles avant l'apparition de réelles pertes chondrales. Enfin, notre dernière étude, toujours à 1,5 T, concerne la quantification du volume et de l'activité de la membrane synoviale inflammatoire dans une cohorte de patients souffrant de gonarthrose.
Lifestyle and genetics are known to be the major factors causing cerebral aneurysms, but some studies suggest that the shape of cerebral arteries might be correlated with the risk of aneurysm occurrence. This study focuses on the shape of cerebral arteries where cerebral aneurysms tend to occur. First, it extracts the shape feature of the cerebral artery ring, which is a predilection site of cerebral aneurysm, from 3-D magnetic resonance angiography images, and calculates four types of shape feature vectors – 3-D shape, bifurcation angle, degree of meandering, and direction of the branch points. Then, it estimates the risk of cerebral aneurysms occurring, based on the extracted features using support vector machine. To validate the proposed method, we conducted a leave-one-out cross validation test using 80 subjects (40 subjects with and 40 subjects without cerebral aneurysms). The method using a 3-D artery shape achieved 75% sensitivity and 75% specificity; the one using the bifurcation angle showed 47% sensitivity and 41% specificity. The method using the degree of meandering showed 55% sensitivity and 53% specificity, and the one that used the direction of the six branch points showed 30% sensitivity and 27% specificity. These results show that the 3-D artery shape could be a possible indicator for predicting the risk of developing cerebral aneurysms.
Magnetic resonance imaging (MRI) provides an effective and noninvasive means by which to evaluate articular cartilage within the knee. Existing techniques can be utilized to detect and monitor longitudinal changes in cartilage status due to injury or progression of degenerative disease. Quantitative MRI (qMRI) techniques can provide a metric by which to evaluate the efficacy of cartilage repair techniques and offer insight into the composition of cartilage and cartilage repair tissue. In this review, we provide background on MR signal generation and decay, the utility of morphologic MRI assessment, and qMRI techniques for the biochemical assessment of cartilage (dGEMRIC, T2, T2*, T1ρ, sodium, gagCEST). Finally, the description and utility of these qMRI techniques for the evaluation of cartilage repair are discussed.
Background:
Image registration (IR) is an important process of developing a spatial relationship between pre-operative data and the physical patient in the operation theatre. Current IR techniques for Computer Assisted Orthopaedic Surgery (CAOS) are time consuming and costly. There is a need to automate and accelerate this process.
Methods:
Bespoke quick, cost effective, contactless and automated 3D laser scanning techniques based on the DAVID Laserscanner method were designed. 10 cadaveric knee joints were intra-operatively laser scanned and were registered with the pre-operative MRI scans. The results are supported with a concurrent validity study.
Results:
The average absolute errors between scan models were systematically less than 1 mm. Errors on femoral surfaces were higher than tibial surfaces (P<0.05). Additionally, scans acquired through the large exposure produced higher errors than the smaller exposure (P<0.05).
Conclusion:
This study has provided proof of concept for a novel automated shape acquisition and registration technique for CAOS.
Despite the rapid growth of three-dimensional (3D) printing applications in medicine, the accuracy and reproducibility of 3D printed medical models have not been thoroughly investigated. Although current technologies enable 3D models to be created with accuracy within the limits of clinical imaging spatial resolutions, this is not always achieved in practice. Inaccuracies are due to errors that occur during the imaging, segmentation, postprocessing, and 3D printing steps. Radiologists' understanding of the factors that influence 3D printed model accuracy and the metrics used to measure this accuracy is key in directing appropriate practices and establishing reference standards and validation procedures. The authors review the various factors in each step of the 3D model printing process that contribute to model inaccuracy, including the intrinsic limitations of each printing technology. In addition, common sources of model inaccuracy are illustrated. Metrics involving comparisons of model dimensions and morphology that have been developed to quantify differences between 3D models also are described and illustrated. These metrics can be used to define the accuracy of a model, as compared with the reference standard, and to measure the variability of models created by different observers or using different workflows. The accuracies reported for specific indications of 3D printing are summarized, and potential guidelines for quality assurance and workflow assessment are discussed. Online supplemental material is available for this article. (©)RSNA, 2017.
Osteoarthritis (OA) is a common disease observed in elderly population and characterized by a progressive destruction of cartilaginous tissue. The clinical diagnosis of this disease is realized by conventional radiography. This method allows visualizing bone modifications related to OA disease (cysts, subchondral bone thickening, and osteophytes) but is unable to assess directly cartilage structure.
Due to its high spatial resolution and high contrast between tissues, Magnetic Resonance Imaging (MRI) is able
to visualize the cartilage structure and to differentiate it from adjacent structures (bone, synovial tissue, menisci, and synovial fluid).
We have employed MRI potentialities at high magnetic field (7 Teslas) to follow, in vivo, cartilage modifications
in the rat femoro-tibial articulation. This methodology was used to evaluate normal cartilage ageing-process and to assess an experimental OA model (anterior cruciate ligament transaction). These works showed an age-related cartilage volumetric and thickness decrease, as well as femoral cartilage damages and tibial cartilage oedema in OA knees.
In a second part of our work, we applied T2 mapping and volumetric techniques (at 1.5 T) to determine variations which occur in the elderly human patellar cartilage. Results demonstrated the capacity of T2 mapping to early detect matricial modifications before any cartilage volumetric impact can be found.
At least, our last study, always at 1.5 T, focused on the synovial membrane volume and inflammatory activity by
taking into account a human population suffering from knee OA.
(Ref.: univ-lorraine-ori-5619).
This unique book describes the latest information in the fundamental understanding of the biophysics and biochemistry of articular cartilage using the state-of-the-art practices in NMR and MRI. This is the first book of its kind, written by physicists and chemists on this important tissue, whose degradation contributes to osteoarthritis and related joint diseases. Connecting the fundamental science with the clinical imaging applications, the experts Editors provide an authoritative addition to the literature.
Ideal for practising physical scientists and radiologists with an interest in the fundamental science as well as instrument manufacturers and clinical researchers working with articular cartilage.
Lisfranc injuries affect at least 1 out of every 55,000 people each year. Although they are relatively uncommon foot injuries, their effects can be severe: 20-40% of Lisfranc injuries are missed upon first presentation. This increases the likelihood of poor outcomes, especially if associated with a workplace injury. These injuries result in a disproportionate number of malpractice lawsuits and compensation claims. The Kingston Brace was designed to support an injured foot during a computed tomography (CT) scan with the goal of providing the diagnosing physician with the best diagnostic information possible. A prototype was designed and built to support a variety of foot orientations. It was used to determine which orientation is optimal for Lisfranc joint CT scanning. Three fresh frozen cadaver feet were put through several experiments with the Brace. The 2D diagnostic quality of CT scans taken using the Kingston Brace was compared against that of CT scans taken using the existing protocol. The Kingston Brace allows for a better visualization of the injured Lisfranc joint than the existing CT protocol. Lisfranc joint spreading was used as a measure of potential pain in injured patients. The joint spreading in the Kingston Brace was minimal, suggesting that additional pain felt by injured patients during imaging would also be minimal. The adoption of the Kingston Brace resulted in no change in the morphological parameters resulting from more advanced 3D analysis. The best experimentally determined Kingston Brace orientation was found to be 10° of plantarflexion and 11° of eversion.
Biological tissues are plastic with respect to the mechanical environment to which they are exposed. This makes them able to modify their architecture and inner structure in order to respond to different loading conditions with the smallest biological effort (functional adaptation). As a result, tissues can optimally adapt their structures to the task they have to perform.
Based on these concepts, a kinetic model of the ankle joint has been recently developed. The tibio-talar relative motion was obtained by imposing the congruence maximization as a biological optimum throughout the entire flexion range.
The aim of this work is to investigate the applicability of the proposed approach to the knee and to evaluate the weight of the meniscal contribution to the global femoro-tibial congruence.
Objectives
We sought to determine if anterior cruciate ligament (ACL)-injured subjects demonstrated side-to-side differences in tibial cartilage thickness soon after injury, and if uninjured-control subjects displayed side-to-side symmetry in cartilage thickness. Second, we aimed to investigate associations between body mass index (BMI), cross-sectional area (CSA) of the proximal tibia, and articular cartilage thickness differences.
Methods
Bilateral Magnetic Resonance Images (MRIs) were obtained on 88 ACL-injured subjects (27 male; 61 female) a mean 27 days post-injury, and 88 matched uninjured control subjects. Within ACL-injured and uninjured control subjects, side-to-side differences in medial and lateral tibial articular cartilage thickness were analyzed with adjustment for tibial position relative to the femur during MRI acquisition. Associations between tibial CSA and cartilage thickness differences were tested within high and low BMI groups.
Results
Within the medial tibial compartment, ACL-injured females displayed significant increases: mean (confidence interval (CI)) = +0.18 mm (0.17, 0.19) and decreases: mean (CI) = −0.14 mm (−0.13, −0.15) in tibial cartilage thickness within the central and posterior cartilage regions respectively.
Adjustment for tibial position revealed a decreased area of significant cartilage thickness differences, though 46% of points maintained significance.
In the lateral compartment anterior region, there was a significantly different relationship between cartilage thickness differences and CSA, within high and low BMI groups (BMI group*CSA interaction, P = 0.007). Within the low BMI group, a significant negative correlation between cartilage thickness and CSA was identified (P = 0.03).
Conclusions
ACL-injured females displayed cartilage thickness differences in the central, and posterior medial tibial cartilage regions. Tibial position effected thickness differences, but did not account for all significant differences.
Many diseases impair the function of the heart by a ffecting its anatomy and physiology. Patient-specif ic surgeries are often required to repair complex diseases of th e heart, but it is difficult for the physicians to plan these surgeries using only 2D images. The integration of 3D models into the surgical planning process has been hindere d by many obstacles. The advent of rapid prototyping (RP) all ows for the creation of quick and accurate patient- specific 3D models from medical imaging data. The process to convert the files has not been properly documented ma king it difficult for future researchers to build upon. Thi s leaves very little information pertaining to the steps required to convert the file, best practices found, nor a tool to use in future research. The goal of this researc h was to document and streamline the file conversion process from med ical image file to RP model in industry recognized software packages. Detailed medical image files were acquire d from patients and archived data sets. The medical images were converted to stereolithography (STL) files and sent to RP machines for model production. Mimics software package was used to separate the heart from the sur rounding tissues. Manual segmentation functions were explored, and a protocol was written to facilitate the conver sion from medical image to STL file. Freeform Modeling system was used to finalize the STL file. Since finalizing tasks on Freeform are file-dependant, common final izing tasks were documented. Stereolithography and 3D printing RP technologies were used to create various models of the heart. The protocol created will be an invaluable t ool for researchers to further the use of 3D modeli ng in the medical community. In addition, the robustness of t his protocol makes it useful for a large range of p atient-specific 3D modeling applications in the medical field.
Background:
The effects of exercise on osteoarthritic cartilage remain elusive.
Objective:
The objective of this study was to investigate the effect of dynamic in vivo squatting exercise on the magnitude and spatial pattern of acute cartilage responses in people with tibiofemoral osteoarthritis (ie, Kellgren-Lawrence grades 1 and 2).
Design:
This investigation was a case-control study.
Methods:
Eighteen people with radiographic signs of doubtful to mild medial tibiofemoral osteoarthritis were compared with 18 people who were middle-aged and healthy (controls). Three-dimensional magnetic resonance imaging was used to monitor deformation and recovery on the basis of 3-dimensional cartilage volume calculations (ie, total volume and volumes in anterior, central, and posterior subregions) before and after a 30-repetition squatting exercise. Three-dimensional volumes were estimated after semiautomatic segmentation and were calculated at 4 time points (1 before and 3 after scans). Scans obtained after the exercise were separated by 15-minute intervals.
Results:
In both groups, significant deformation was noted in the medial compartment (-3.4% for the femur and -3.2% for the tibia in people with osteoarthritis versus -2.8% for the femur and -3.8% for the tibia in people in the control group). People with osteoarthritis had significant deformation in the lateral femur (-3.9%) and a tendency toward significant deformation in the lateral tibia (-3.1%). From 15 minutes after exercise cessation onward, volume changes were no longer significantly different from the baseline. At all time points, no significant between-group differences were revealed for volume changes. People with osteoarthritis showed a tendency toward slower recovery preceded by larger deformations in entire cartilage plates and subregions. Spatial subregional deformation patterns were similar between groups.
Limitations:
Generalizability is limited to people with doubtful to mild osteoarthritis and low levels of pain.
Conclusions:
Tibiofemoral cartilage deformation appeared similar in magnitude and spatial pattern in people who were middle-aged and either had or did not have tibiofemoral osteoarthritis (ie, Kellgren-Lawrence grades 1 and 2). Restoration of volumes required a 15-minute recovery, especially in the presence of osteoarthritic cartilage degeneration.
Knee osteoarthritis deals with the loss of cartilaginous substances. Biomaterial implant using mesenchymalstromal cells was usually used to perform cartilage regeneration.The biointegration and biofunctionality assessments of the used biomaterial implant on the injury sites require the longitudinal monitoring of the cartilage over times. The aim of this present study was to develop a characterization framework of the cartilage in knee osteoarthritis especially its morphological properties and contact areas derived from high-resolution magnetic resonance imaging (MRI) technique. Manual segmentation procedure was applied for extracting cartilage from surrounding biological tissues. Geometrical computing techniques were applied to calculate characterization properties of the rat cartilage. As clinical case study, curvature, thickness, volume and relative contact area properties of the rat cartilage in knee osteoarthritis were quantified on each knee full model and/or its region of interest (ROI) models.We found that the cartilage thickness and volume of three cartilage compartments decrease at the beginning (from 7 to 28 days). Tibial cartilage thickness increases at the end of the osteoarthritis process (from 28 to 56 days).This is due to the fact that the cartilage swelling occurred during osteoarthritis process. These findings were confirmed when observing cartilage curvature; we noted that there is an increasing curvature due to the incurvation effect of the cartilage during osteoarthritis process. The relative contact areas of three cartilage compartments increase slightly. However, the difference between control knees and osteoarthritis knees is significant. ROIs analyses of the thickness and curvature properties were also reported on the femoral and tibial cartilages.
Brain-skull relative motion plays a pivotal role in the etiology of traumatic brain injury (TBI). The present study aims to assess brain-skull relative motion in quasistatic circumstances, and to correlate cortical regions with high motion amplitudes with sites prone to cerebral contusions. The study includes 30 healthy volunteers scanned using a clinical 3-T MR scanner in four different head positions. Through image processing and 3D model registration, pairwise comparisons were performed to calculate the brain shift between sagittal and coronal head positional change. Next, local brain deformation was evaluated by comparison between cortical and ventricular amplitudes. Finally, the influence of age, sex, and skull geometry on the cortical and ventricular motion was investigated. The results describe complex brain shift patterns, with high regional and inter-individual variations, outweighing age and sex patterns. Regions with maximum motion amplitudes were identified at the inferolateral aspects of the frontal and temporal lobes, congruent with predilection sites for contusions. No significant influences of age and sex on the cortical shift amplitudes were detected. The 3D cortical deviations varied from -7.86 mm to +5.71 mm for the sagittal head movement, and from -11.46 mm to +7.30 mm for head movement in the coronal plane, for a 95% confidence interval. The present study contributes to a better understanding of the mechanopathogenesis of frontotemporal contusions, and is useful for the optimization of finite-element head models and neurosurgical navigation procedures. Moreover, our results prove that in vivo MRI allows for accurate assessment of brain-skull relative motion in quasistatic conditions.
The present study aimed to develop a method to measure three-dimensional (3-D) thickness of cartilage (Tc) at the femoral condyle using B-mode ultrasonography (US) and to clarify the feasibility of US in clinical evaluations of articular cartilage by comparing the results with 3-D measurement values using magnetic resonance imaging (MRI) and assessing repeatability. The medial surface of the right knees of two healthy male volunteers (age, 37 and 59 years) and the knees on affected side of three male patients with osteoarthritis (OA) (age, 73, 81 and 83 years) were scanned using B-mode US with the knee flexed at 120°. The range of the angle of probe rotation for the arm was 0-80° and B-mode images (total, 101 images) were acquired every 0.8°. MRI of the knees was also performed using the double echo steady-state sequence. Both US and MRI images were used to create 3-D models of medial femoral condyle articular cartilage. Tc was determined at points 1 mm apart from one another in the US model (Tc-US) and MRI model (Tc-MRI). Tc-US was compared with Tc-MRI and the repeatability of Tc-US was assessed by mean Tc in the specific region of interest of the femoral condyle. Tc-US correlated significantly with Tc-MRI both in volunteers and in OA patients (p < 0.0001 each) and coefficients of correlation were 0.976 and 0.964 for volunteers and OA patients, respectively. The coefficient of variance for mean Tc-US was 4.90%. Our results show that 3-D US measurements of femoral cartilage are reproducible and correlate strongly with MRI measurements.
To compare the cartilage thickness of the third toe joints to the finger proximal interphalangeal (PIP) joints to assess the appropriateness of using third toe osteochondral grafts for finger PIP joint reconstruction.
A laser scanner was used to construct 3-dimensional computer models of 6 matched cadaver right third toe PIP joints, condyles of the third toe middle phalanx, and finger PIP joints with and without cartilage. Cartilage distribution patterns were computed and analyzed for each surface. The cartilage thickness of both sides of the third toe PIP joint and the third toe middle phalanx condyles were compared to the PIP joint of the fingers. A total of 18 third toe and 48 finger joint surfaces were analyzed.
For the third toe middle phalanx condyles, the mean thickness was 0.20 ± 0.09 mm with a maximum of 0.52 ± 0.18 mm, and a coefficient of variation (CV%; a measure of uniformity of cartilage distribution) of 62. For the third toe proximal phalanx condyles, the mean cartilage thickness was 0.26 ± 0.10 mm with a maximum thickness of 0.56 ± 0.14 mm and a CV% of 44. The mean thickness, maximum thickness, and CV% of the finger proximal phalanx condyles was 0.43 ± 0.11 mm, 0.79 ± 0.16 mm, and 31, respectively. For the third toe middle phalanx base, the mean thickness was 0.28 ± 0.06 mm with a maximum of 0.47 ± 0.09 mm and a CV% of 34, compared to the finger middle phalanx base mean of 0.40 ± 0.12 mm with a maximum of 0.67 ± 0.14 mm and a CV% of 27.
There were significant differences in cartilage thickness between the third toe and the fingers in this study. However, fewer differences were observed with the third toe middle phalanx base cartilage thickness than with the third toe condyles in comparison to the fingers.
The corpus callosum (CC) is frequently compromised in patients with multiple sclerosis (MS). Structural and functional measurements of the CC may be useful to monitor the progression of the disease. The aim of this pilot study was to determine if bimanual tactile temporal thresholds correlates with CC volume. A tactile temporal threshold is the longest temporal interval that separates the onsets of two tactile stimuli when they are judged by the observer as simultaneous. Judgments to bimanual stimulations require interhemispheric transfer via the CC.
Thresholds were examined in MS patients and matched controls. Magnetic resonance (MR) images were acquired on a 3T MR system within 48 hours of clinical assessment and measurement of thresholds.
Corpus callosum volume was assessed by using a semiautomatic livewire algorithm. The CC volume was smaller (by 21% on average, p < 0.01) and thresholds were higher (by 49% on average, p < 0.03) in MS patients when compared to controls. A significant correlation (r = -0.66, p = 0.01) between CC volume and thresholds emerged for the MS patients.
Measuring treatment benefits of neuroprotective and repair therapies is a well recognized challenge in MS research. The overall findings of this study suggest that these measurements, which involve the transfer of information interhemispherically via the CC, may be promising outcome measures that warrant further scientific exploration to develop a model to measure recovery.
Quantifying changes in meniscal volume in vivo before and after partial meniscectomy (PM) could help elucidate the mechanisms involved in osteoarthritis development after meniscal injury and its surgical treatment. Purpose/
To determine whether quantitative magnetic resonance imaging (qMRI) can detect the immediate reduction in meniscal volume created by PM, while ruling out changes in unresected structures. We hypothesized that qMRI would be reliable for determining meniscal volume within the repeated images of unresected menisci. Additionally, we expected no significant difference in volume between the uninjured menisci of the injured knees and the same menisci of the uninjured knees.
Cohort study (Diagnosis); Level of evidence, 2.
Ten subjects with meniscal tears were evaluated with 3-T MRI before and after arthroscopic PM. Manual segmentation was used to create models of the menisci and to determine the preoperative and postoperative meniscal volumes for each subject. The responsiveness and reliability of qMRI for determining meniscal volume in vivo were evaluated using these measurements. We expected a decrease in volume of the resected menisci, but not in the uninjured menisci, after surgery.
The mean preoperative volume of the injured menisci was significantly greater than the mean postoperative volume (2896 +/- 277 vs 2480 +/- 277 mm(3); P = .000). There was no significant difference between the mean preoperative and postoperative volumes of the uninjured menisci (2687 +/- 256 vs 2694 +/- 256 mm(3); P = 1.000).
Manual segmentation demonstrated a significant reduction in the volume of the surgically resected menisci after PM, but no significant change in the volume of unresected meniscal tissue, indicating that the manual segmentation method is responsive.
This approach offers a novel, reliable method to study the relationship between the volume of meniscal tissue removed during PM and subsequent patient outcomes during long-term clinical studies.
Different methods have been used to cross-validate cartilage thickness measurements from magnetic resonance images (MRIs); however, a majority of these methods rely on interpolated data points, regional mean and/or maximal thickness, or surface mean thickness for data analysis. Furthermore, the accuracy of MRI cartilage thickness measurements from commercially available software packages has not necessarily been validated and may lead to an under- or overestimation of cartilage thickness. The goal of this study was to perform a matching point-to-point validation of indirect cartilage thickness calculations using a magnetic resonance (MR) image data set with direct cartilage thickness measurements using biomechanical indentation testing at the same anatomical locations. Seven bovine distal femoral condyles were prepared and a novel phantom filled with dilute gadolinium solution was rigidly attached to each specimen. High resolution MR images were acquired, and thickness indentation analysis of the cartilage was performed immediately after scanning. Segmentation of the MR data and cartilage thickness calculation was performed using semi-automated software. Registration of MR and indentation data was performed using the fluid filled phantom. The inter- and intra-examiner differences of the measurements were also determined. A total of 105 paired MRI-indentation thickness data points were analyzed, and a significant correlation between them was found (r=0.88, p<0.0001). The mean difference (+/-std. dev.) between measurement techniques was 0.00+/-0.23 mm, with Bland-Altman limits of agreement of 0.45 mm and -0.46 mm. The intra- and inter-examiner measurement differences were 0.03+/-0.22 mm and 0.05+/-0.24 mm, respectively. This study validated cartilage thickness measurements from MR images with thickness measurements from indentation by using a novel phantom to register the image-based and laboratory-based data sets. The accuracy of the measurements was comparable to previous cartilage thickness validation studies in literature. The results of this study will aid in validating a tool for clinical evaluation of in-vivo cartilage thickness.
Despite the increasing interest in using knee cartilage volume as an outcome measure in studies of osteoarthritis (OA), it is unclear what components of knee cartilage will be most useful as markers of structure in the tibiofemoral (TF) joint.
To compare the changes that occur in femoral and tibial cartilage volume in normal and osteoarthritic knees and how they relate to radiological grade.
82 subjects (44 female, 38 male, age range 35-69 years) with a spectrum of radiological knee OA were examined. Each subject had femoral and tibial cartilage volume in the medial and lateral TF joint determined from T(1) weighted fat saturated magnetic resonance images of the knee. Radiological grade of OA was determined from standing knee radiographs.
There was strong correlation between femoral and tibial cartilage volume measured in both the medial (R=0.75, p<0.001) and lateral TF joint (R=0.77, p<0.001). Similar correlations persisted when those with normal and those with OA joints were examined separately at both the medial and lateral TF joint. For each increase in radiological grade of joint space narrowing (0-3), there was a mean (SD) reduction in tibial cartilage volume of 1.00 (0.32) ml in the medial compartment and 0.53 (0.25) ml in the lateral compartment, after adjusting for differences in bone size. Similar changes were seen in the femoral cartilage.
The amounts of tibial and femoral cartilage are strongly related. It may be that for TF joint disease, measuring tibial cartilage alone may be adequate, given that measurements of the total femoral cartilage are less reproducible and the difficulties inherent in identifying the most appropriate component of femoral cartilage to measure.
The primary objective of this study was to develop a computer-aided method for the quantification of three-dimensional (3-D) cartilage changes over time in knees with osteoarthritis (OA). We introduced a local coordinate system (LCS) for the femoral and tibial cartilage boundaries that provides a standardized representation of cartilage geometry, thickness, and volume. The LCS can be registered in different data sets from the same patient so that results can be directly compared. Cartilage boundaries are segmented from 3-D magnetic resonance (MR) slices with a semi-automated method and transformed into offset-maps, defined by the LCS. Volumes and thickness are computed from these offset-maps. Further anatomical labeling allows focal volumes to be evaluated in predefined subregions. The accuracy of the automated behavior of the method was assessed, without any human intervention, using realistic, synthetic 3-D MR images of a human knee. The error in thickness evaluation is lower than 0.12 mm for the tibia and femur. Cartilage volumes in anatomical subregions show a coefficient of variation ranging from 0.11% to 0.32%. This method improves noninvasive 3-D analysis of cartilage thickness and volume and is well suited for in vivo follow-up clinical studies of OA knees.
There is emerging evidence that knee alignment is associated with progression of osteoarthritis (OA). The aim of this study was to examine the relationship between baseline knee angle and the rate of cartilage loss in subjects with knee OA.
One hundred and seventeen subjects with knee OA had standing radiographs and MRI on their symptomatic knee at baseline and at the 1.9+/-0.2 yr follow-up. Knee cartilage volume was measured at baseline and follow-up. Knee angle was defined as the angle subtended by a line drawn through the mid-shaft of the femur with respect to one drawn through the mid-shaft of the tibia.
At baseline, in the medial compartment, as the angle decreased (i.e. was less varus) the tibial and femoral cartilage volume increased. In the lateral compartment, as the angle became more valgus, there was a reduction in tibial and femoral cartilage volume. In the longitudinal study, for every 1 degrees increase in baseline varus angulation there was an average annual loss of medial femoral cartilage of 17.7 micro l [95% confidence interval (CI) 6.5-28.8]. Although not statistically significant, there was a trend for a similar relationship between loss of medial tibial cartilage volume and baseline knee angle. In the lateral compartment, there was an average loss of tibial cartilage volume of 8.0 micro l (95% CI 0.0-16.0) for every 1 degrees increase in valgus angle.
Baseline knee angle is associated with the rate of cartilage loss in the knee. Further work will be needed to determine whether therapies aimed at modifying the knee angle will reduce the progression of knee OA.
No consistent relationship between the severity of symptoms of knee osteoarthritis (OA) and radiographic change has been demonstrated.
To determine the relationship between symptoms of knee OA and tibial cartilage volume, whether pain predicts loss of cartilage in knee OA, and whether change in cartilage volume over time relates to change in symptoms over the same period.
132 subjects with symptomatic, early (mild to moderate) knee OA were studied. At baseline and 2 years later, participants had MRI scans of their knee and completed questionnaires quantifying symptoms of knee OA (knee-specific WOMAC: pain, stiffness, function) and general physical and mental health (SF-36). Tibial cartilage volume was determined from the MRI images.
Complete data were available for 117 (89%) subjects. A weak association was found between tibial cartilage volume and symptoms at baseline. The severity of the symptoms of knee OA at baseline did not predict subsequent tibial cartilage loss. However, weak associations were seen between worsening of symptoms of OA and increased cartilage loss: pain (r(s) = 0.28, p = 0.002), stiffness (r(s) = 0.17, p = 0.07), and deterioration in function (r(s) = 0.21, p = 0.02).
Tibial cartilage volume is weakly associated with symptoms in knee OA. There is a weak association between loss of tibial cartilage and worsening of symptoms. This suggests that although cartilage is not a major determinant of symptoms in knee OA, it does relate to symptoms.
We describe a technique to axially compress a sheep knee joint in an MRI scanner and measure articular cartilage deformation. As an initial application, tibial articular cartilage deformation patterns after 2 h of static loading before and after medial meniscectomy are compared.
Precision was established for repeated scans and repeated segmentations. Accuracy was established by comparing to micro-CT measurements. Four sheep knees were then imaged unloaded, and while statically loaded for 2 h at 1.5 times body weight before and after medial meniscectomy. Images were obtained using a 3D gradient echo sequence in a 4.7 T MRI. Corresponding 3D cartilage thickness models were created. Nominal strain patterns for the intact and meniscectomized conditions were compared.
Coefficients of variation were all 2% or less. Root mean squared errors of MR cartilage thickness measurements averaged less than 0.09 mm. Meniscectomy resulted in a 60% decrease in the contact area (P=0.001) and a 13% increase in maximum cartilage deformation (P=0.01). Following meniscectomy, there were greater areas of articular cartilage experiencing abnormally high and low nominal strains. Areas of moderate nominal strain were reduced.
Medial meniscectomy resulted in increased medial tibial cartilage nominal strains centrally and decreased strains peripherally. Areas of abnormally high nominal strain following meniscectomy correlated with areas that are known to develop fibrillation and softening 16 weeks after medial meniscectomy. Areas of abnormally low nominal strain correlated with areas of osteophyte formation. Studies of articular cartilage deformation may prove useful in elucidating the mechanical etiology of osteoarthritis.
We propose an algorithm for surface reconstruction from unorganized points based on a view of the sampling process as a deformation from the original surface. In the course of this deformation the Medial Scaffold (MS) - a graph representation of the 3D Medial Axis (MA) - of the original surface undergoes abrupt changes (transitions) such that the MS of the unorganized point set is significantly different from that of the original surface. The algorithm seeks a sequence of transformations of the MS to invert this process. Specifically, some MS curves (junctions of 3 MA sheets) correspond to triplets of points on the surface and represent candidates for generating a (Delaunay) triangle to mesh that portion of the surface. We devise a greedy algorithm that iteratively transforms the MS by "removing" suitable candidate MS curves (gap transform) from a rank-ordered list sorted by a combination of properties of the MS curve and its neighborhood context. This approach is general and applicable to surfaces which are: non-closed, non-orientable, non-uniformly sampled. In addition, the method is comparable in speed and complexity to current popular Voronoi/Delaunay-based algorithms, and is applicable to very large datasets.
We present a fully automatic method for articular cartilage segmentation from magnetic resonance imaging (MRI) which we use as the foundation of a quantitative cartilage assessment. We evaluate our method by comparisons to manual segmentations by a radiologist and by examining the interscan reproducibility of the volume and area estimates. Training and evaluation of the method is performed on a data set consisting of 139 scans of knees with a status ranging from healthy to severely osteoarthritic. This is, to our knowledge, the only fully automatic cartilage segmentation method that has good agreement with manual segmentations, an interscan reproducibility as good as that of a human expert, and enables the separation between healthy and osteoarthritic populations. While high-field scanners offer high-quality imaging from which the articular cartilage have been evaluated extensively using manual and automated image analysis techniques, low-field scanners on the other hand produce lower quality images but to a fraction of the cost of their high-field counterpart. For low-field MRI, there is no well-established accuracy validation for quantitative cartilage estimates, but we show that differences between healthy and osteoarthritic populations are statistically significant using our cartilage volume and surface area estimates, which suggests that low-field MRI analysis can become a useful, affordable tool in clinical studies
Dynamic programming is discussed as an approach to solving variational problems in vision. Dynamic programming ensures global optimality of the solution, is numerically stable, and allows for hard constraints to be enforced on the behavior of the solution within a natural and straightforward structure. As a specific example of the approach's efficacy, applying dynamic programming to the energy-minimizing active contours is described. The optimization problem is set up as a discrete multistage decision process and is solved by a time-delayed discrete dynamic programming algorithm. A parallel procedure for decreasing computational costs is discussed
Background: Quantifying the in vivo cartilage contact mechanics of the knee may improve our understanding of the mechanisms of joint degeneration and may therefore improve the surgical repair of the joint after injury. Objective: To measure tibiofemoral articular cartilage contact kinematics during in vivo knee flexion. Study Design: Descriptive laboratory study. Methods: Orthogonal fluoroscopic images and magnetic resonance image-based computer models were used to measure the motion of the cartilage contact points during a quasi-static lunge in 5 human subjects. Results: On the tibial plateau, the contact point moved in both the anteroposterior and the mediolateral directions during knee flexion. On the medial tibial plateau, flexion angle did not have a statistically significant effect on the location of the contact points. The total translation of the contact point from full extension to 90 degrees of flexion was less than 1.5 mm in the anteroposterior direction, whereas the translation in the mediolateral direction was more than 5.0 mm. In the anteroposterior direction, the contact points were centered on the medial tibial plateau. On the lateral tibial plateau, there was a statistically significant difference between the location of the contact point at full extension and the locations of the contact points at other flexion angles in the anteroposterior direction. No significant difference was detected between the location of the contact points at other flexion angles. The overall range of contact point motion was about 9.0 mm in the anteroposterior direction and about 4.0 mm in the mediolateral direction. The contact points were primarily on the inner half of the medial and lateral tibial plateaus (the half closest to the tibial spine). The contact points on both femoral condyles were also on the inner half of the condyles (near the condylar notch). Conclusions: The tibiofemoral contact points move in 3 dimensions during weightbearing knee flexion. The medial tibiofemoral contact points remained within the central portion of the tibial plateau in the anteroposterior direction. Both the medial and lateral tibiofemoral contact points were located on the inner portions of the tibial plateau and femoral condyles (close to the tibial spine), indicating that the tibial spine may play an important role in knee stability. Clinical Relevance: The results of this study may provide important insight as to the mechanisms contributing to the development of osteoarthritis after ligament injuries.
A snake is an energy-minimizing spline guided by external constraint forces and influenced by image forces that pull it toward features such as lines and edges. Snakes are active contour models: they lock onto nearby edges, localizing them accurately. Scale-space continuation can be used to enlarge the capture region surrounding a feature. Snakes provide a unified account of a number of visual problems, including detection of edges, lines, and subjective contours, motion tracking, and stereo matching. The authors have used snakes successfully for interactive interpretation, in which user-imposed constraint forces guide the snake near features of interest.
Live-wire segmentation is a new interactive tool for efficient, accurate, and reproducible
boundary extraction which requires minimal user input with a mouse. Optimal boundaries are
computed and selected at interactive rates as the user moves the mouse starting from a manually
specified seed point. When the mouse position comes in proximity to an object edge, a “live-wire”
boundary snaps to, and wraps around the object of interest. Input of a new seed point “freezes” the
selected boundary segment, and the process is repeated until the boundary is complete. Two novel
enhancements to the basic live wire methodology include boundary cooling and on-the-fly training.
Data-driven boundary cooling generates seed points automatically and further reduces user input.
On-the-fly training adapts the dynamic boundary to edges of current interest.
Using the live wire technique, boundaries are extracted in one-fifth of the time required for
manual tracing, but with 4.4 times greater accuracy and 4.8 times greater reproducibility. In
particular, interobserver reproducibility using the live wire tool is 3.8 times greater than
intraobserver reproducibility using manual tracing.
The segmentation of structure from images is an inherently difficult problem in computer vision and a bottleneck to its widespread application, e.g., in medical imaging. This paper presents an approach for integrating local evidence such as regional homogeneity and edge response to form global structure for figure–ground segmentation. This approach is motivated by a shock-based morphogenetic language, where the growth of four types of shocks results in a complete description of shape. Specifically, objects are randomly hypothesized in the form of fourth-order shocks (seeds) which then grow, merge, split, shrink, and, in general, deform under physically motivated “forces,” but slow down and come to a halt near differential structures. Two major issues arise in the segmentation of 3D images using this approach. First, it is shown that the segmentation of 3D images by 3D bubbles is superior to a slice-by-slice segmentation by 2D bubbles or by “2D bubbles” which are inherently 2D but use 3D information for their deformation. Specifically, the advantages lie in an intrinsic treatment of the underlying geometry and accuracy of reconstruction. Second, gaps and weak edges, which frequently present a significant problem for 2D and 3D segmentation, are regularized by curvature-dependent curve and surface deformations which constitute diffusion processes. The 3D bubbles evolving in the 3D reaction–diffusion space are a powerful tool in the segmentation of medical and other images, as illustrated for several realistic examples.
Live-wire segmentation is a new interactive tool for efficient, accurate and reproducible boundary extraction which requires minimal user input with a mouse. Optimal boundaries are computed and selected at interactive rates as the user moves the mouse starting from a manually specified seed point. When the mouse position comes into the proximity of an object edge, a ‘live-wire’ boundary snaps to, and wraps around the object of interest. The input of a new seed point ‘freezes’ the selected boundary segment and the process is repeated until the boundary is complete. Two novel enhancements to the basic live-wire methodology include boundary cooling and on-the-fly training. Data-driven boundary cooling generates seed points automatically and further reduces user input. On-the-fly training adapts the dynamic boundary to edges of current interest. Using the live-wire technique, boundaries are extracted in one-fifth of the time required for manual tracing, but with 4.4 times greater accuracy and 4.8 times greater reproducibility. In particular, interobserver reproducibility using the live-wire tool is 3.8 times greater than intraobserver reproducibility using manual tracing.
A snake is an energy-minimizing spline guided by external constraint forces and influenced by image forces that pull it toward features such as lines and edges. Snakes are active contour models: they lock onto nearby edges, localizing them accurately. Scale-space continuation can be used to enlarge the capture region surrounding a feature. Snakes provide a unified account of a number of visual problems, including detection of edges, lines, and subjective contours; motion tracking; and stereo matching. We have used snakes successfully for interactive interpretation, in which user-imposed constraint forces guide the snake near features of interest.
In the past, we have presented three user-steered image segmentation
paradigms: live wire, live lane, and the 3D extension of the live-wire
method. In this paper, we introduce an ultra-fast live-wire method,
referred to as live-wire-on-the-fly, for further reducing user's time
compared to live wire. For both approaches, given a slice and a 2D
boundary of interest in this slice, we translate the problem of finding
the best boundary segment between any two points specified by the user
on this boundary to the problem of finding the minimum-cost path between
two vertices in a weighted and directed graph. The entire 2D boundary is
identified as a set of consecutive boundary segments, each specified and
detected in this fashion. A drawback in live wire is that the speed for
optimal path computation depends on image size, compromising the overall
segmentation efficiency. In this work, we solve this problem by
exploiting some properties of graph theory to avoid unnecessary
minimum-cost path computation during segmentation. Based on 164
segmentation experiments from an actual medical application, we
demonstrate that live-wire- on-the-fly is about 1.5 to 33 times faster
than live wire for actual segmentation, although the pure computational
part alone is found to be over a hundred times faster.
Previously reported studied to quantify articular cartilage have used labor-intensive manual or semi-automatic data-driven techniques, demonstrating high accuracy and precision. However, none has been able to automate the segmentation process. This paper describes a fast, automatic, model-based approach to segmentation and thickness measurement of the femoral cartilage in 3D T1-weighted images using active shape models (ASMs). Systematic experiments were performed to assess the accuracy and precision of the technique with in vivo images of both normal and abnormal knees. Segmentation accuracy was determined by comparing the results of the segmentation with the boundaries delineated by a radiologist. The mean error in locating the boundary was 0.57 pixels. To assess the precision of the measurement technique, the mean thickness of the femoral cartilage was calculated for repeated scans of five healthy volunteers. A mean coefficient of variation (CV) of 2.8% was obtained for the thickness measurements.
This work demonstrates a new method for computerized measurement of the dimensions (thickness and volume) of articular cartilage for any specified region of the human knee joint. Three-dimensional magnetic resonance (MR) images optimized for cartilage contrast have been analyzed using computerized edge-detection techniques, and the reproducibility of articular cartilage thickness and volume measurements is assessed.
A fat-suppressed, three-dimensional SPoiled GRass MR sequence (45/7.5/30 degrees) with total scan time of approximately 12 minutes was used to acquire volume images of human knee joints at spatial resolution of 0.6 x 1.2 x 1.2 mm. Measurements were made using six repeated scans for three healthy volunteers over a period of 2 months. The subsequent semi-automated image processing to establish total cartilage volume and cartilage thickness maps for the femur required approximately 60 minutes of operator time.
The mean coefficient of variation for total cartilage volume for the six repeated scans for the three volunteers was 3.8%, and the average coefficient of variation for the user-selected cartilage plugs was 2.0%. The cartilage thickness maps from the repeated scans of the same knee were similar.
Standard resolution MR images with fat-suppressed contrast lead to an objective and reproducible measurement of spatial dimensions of articular cartilage when analyzed semi-automatically using computerized edge-detection methods.
A noninvasive imaging technique for quantifying articular cartilage is needed for diagnosis, monitoring, and therapy control in osteoarthritis. In this study the accuracy of three-dimensional cartilage volume and thickness measurements in the knee with magnetic resonance imaging was analyzed. Eight cadaveric specimens had sagittal imaging with a fat suppressed gradient echo sequence. After a contrast agent was injected, two sagittal computed tomography data sets were obtained, with the knees being repositioned between the examinations. The cartilage thickness was determined, after three-dimensional reconstruction, using a minimal distance algorithm. The mean absolute volume deviation between magnetic resonance imaging and computed tomography arthrography was 3.3% and that between the two computed tomography data sets was 3.6%. The absolute error in determining the maximal cartilage thickness with magnetic resonance imaging was on average 0.6 intervals (of 0.5-mm thickness) and that between the computed tomography examinations was 0.5 intervals. In a patient with anterior knee pain, a focal cartilage defect was seen with magnetic resonance imaging, and this was verified by arthroscopic examination. Using three-dimensional image processing, magnetic resonance imaging can provide accurate data on cartilage volume and thickness in the human knee joint surfaces. This imaging technique potentially may be valuable in the treatment of patients with joint disease.
(C) Lippincott-Raven Publishers.
The objective of this work was to develop a computational approach for quantifying the three-dimensional (3D) thickness distribution of articular cartilage with magnetic resonance (MR) imaging, independent of the imaging plane, and to test the reproducibility of the method in the living. An algorithm was implemented, based on a 3D Euclidean distance transformation, and its accuracy was assessed in geometric test objects, for which an analytic solution was available. The precision of the method was evaluated in six replicated MR data sets of the knee joint cartilage of eight volunteers. The algorithm produced 3D thickness values identical to those of the analytic solutions in the test objects. The reproducibility of the mean cartilage thickness in the patellar and tibial cartilages was 1.5-3.4% (root-mean-square average of the individual coefficient of variation percent), that of the maximal thickness 2.1-7.9%, and that of the thickness distribution 2.3-6.1%. The method presented allows for noninvasive analysis of 3D cartilage thickness from MR images in biomechanical and clinical investigations.
The objective of this work was to develop a segmentation technique for thickness measurements of the articular cartilage in MR images and to assess the interobserver reproducibility of the method in comparison with manual segmentation. The algorithm is based on a B-spline snakes approach and is able to delineate the cartilage boundaries in real time and with minimal user interaction. The interobserver reproducibility of the method, ranging from 3.3 to 13.6% for various section orientations and joint surfaces, proved to be significantly superior to manual segmentation.
The rate of change in osteoarthritic (OA) tibial articular cartilage and the factors that influence it are not known. We examined a cohort of subjects with OA to determine the change in articular knee cartilage volume over the course of 2 years and to identify factors which might influence such change and its rate.
One hundred twenty-three subjects with OA underwent baseline knee radiography and magnetic resonance imaging (MRI) on their symptomatic knee. They were followed up 2 years later with a repeat MRI of the same knee. Knee cartilage volume was measured at baseline and at followup. Risk factors assessed at baseline were tested for their association with change in knee cartilage volume over time.
Mean +/- SD total tibial articular cartilage decreased by 5.3 +/- 5.2% (95% confidence interval [95% CI] 4.4%, 6.2%) per year. The annual percentages of loss of medial and lateral tibial cartilage were 4.7 +/- 6.5% (95% CI 3.6%, 5.9%) and 5.3 +/- 7.2% (95% CI 4.1%, 6.6%), respectively. Initial cartilage volume was the most significant determinant of loss of tibial cartilage in all compartments, while age was a significant determinant of lateral tibial cartilage loss, when possible confounders were accounted for.
In OA, tibial cartilage volume is lost at a rate of approximately 5% per year. The main factor affecting cartilage loss is initial cartilage volume. Our results suggest that cartilage loss may be more rapid early in disease. Further study is required to determine whether the rate of cartilage loss in OA is steady or phasic, and to identify factors amenable to intervention to reduce cartilage loss.
To evaluate the rate of progression of cartilage loss in the knee joint using magnetic resonance imaging (MRI) and to evaluate potential risk factors for more rapid cartilage loss.
We evaluated baseline and followup MRIs of the knees in 43 patients (minimum time interval of 1 year, mean 1.8 years, range 52–285 weeks). Cartilage loss was graded in the anterior, central, and posterior regions of the medial and lateral knee compartments. Knee joints were also evaluated for other pathology. Data were analyzed using analysis of variance models.
Patients who had sustained meniscal tears showed a higher average rate of progression of cartilage loss (22%) than that seen in those who had intact menisci (14.9%) (P ≤ 0.018). Anterior cruciate ligament (ACL) tears had a borderline significant influence (P ≤ 0.06) on the progression of cartilage pathology. Lesions located in the central region of the medial compartment were more likely to progress to more advanced cartilage pathology (progression rate 28%; P ≤ 0.003) than lesions in the anterior (19%; P ≤ 0.564) and posterior (17%; P ≤ 0.957) regions or lesions located in the lateral compartment (average progression rate 15%; P ≤ 0.707). Lesions located in the anterior region of the lateral compartment showed less progression of cartilage degradation (6%; P ≤ 0.001). No specific grade of lesion identified at baseline had a predilection for more rapid cartilage loss (P ≤ 0.93).
MRI can detect interval cartilage loss in patients over a short period (<2 years). The presence of meniscal and ACL tears was associated with more rapid cartilage loss. Cartilage lesions located in the central region of the medial compartment showed more rapid progression of cartilage loss than cartilage lesions in the anterior and posterior portions of the medial compartment. The findings in this study suggest that patients entering clinical trials investigating antiarthritis regimens may need to be randomized based on location of the lesion.
The in vivo investigation of joint kinematics in normal and injured wrist requires the segmentation of carpal bones from 3D (CT) images, and their registration over time. The non-uniformity of bone tissue, ranging from dense cortical bone to textured spongy bone, the irregular shape of closely packed carpal bones, small inter-bone spaces compared to the resolution of CT images, along with the presence of blood vessels, and the inherent blurring of CT imaging render the segmentation of carpal bones a challenging task. We review the performance of statistical classification, deformable models (active contours), region growing, region competition, and morphological operations for this application. We then propose a model which combines several of these approaches in a unified framework. Specifically, our approach is to use a curve evolution implementation of region growing from initialized seeds, where growth is modulated by a skeletally-mediated competition between neighboring regions. The inter-seed skeleton, which we interpret as the predicted boundary of collision between two regions, is used to couple the growth of seeds and to mediate long-range competition between them. The implementation requires subpixel representations of each growing region as well as the inter-region skeleton. This method combines the advantages of active contour models, region growing, and both local and global region competition methods. We demonstrate the effectiveness of this approach for our application where many of the difficulties presented above are overcome as illustrated by synthetic and real examples. Since this segmentation method does not rely on domain-specific knowledge, it should be applicable to a range of other medical imaging segmentation tasks.
The high prevalence of osteoarthritis continues to demand improved accuracy in detecting cartilage injury and monitoring its response to different treatments. MRI is the most accurate noninvasive method of diagnosing cartilage lesions. However, MR imaging of cartilage is limited by scan time, signal-to-noise ratio (SNR), and image contrast. Recently, there has been renewed interest in SNR-efficient imaging sequences for imaging cartilage, including various forms of steady-state free-precession as well as driven-equilibrium imaging. This work compares several of these sequences with existing methods, both theoretically and in normal volunteers. Results show that the new steady-state methods increase SNR-efficiency by as much as 30% and improve cartilage-synovial fluid contrast by a factor of three. Additionally, these methods markedly decrease minimum scan times, while providing 3D coverage without the characteristic blurring seen in fast spin-echo images.
The aim of this study was to evaluate the reliability of a software tool that assesses knee cartilage volumes using magnetic resonance (MR) images. The objectives were to assess measurement reliability by: (1) determining the differences between readings of the same image made by the same reader 2 weeks apart (test-retest reliability), (2) determining the differences between the readings of the same image made by different readers (between-reader agreement), and (3) determining the differences between the cartilage volume readings obtained from two MR images of the same knee image acquired a few hours apart (patient positioning reliability).
Forty-eight MR examinations of the knee from normal subjects, patients with different stages of symptomatic knee osteoarthritis (OA), and a subset of duplicate images were independently and blindly quantified by three readers using the imaging system. The following cartilage areas were analyzed to compute volumes: global cartilage, medial and lateral compartments, and medial and lateral femoral condyles.
Between-reader agreement of measurements was excellent, as shown by intra-class correlation (ICC) coefficients ranging from 0.958 to 0.997 for global cartilage (P<0.0001), 0.974 to 0.998 for the compartments (P<0.0001), and 0.943 to 0.999 for the condyles(P<0.0001). Test-retest reliability of within-reader data was also excellent, with Pearson correlation coefficients ranging from 0.978 to 0.999 (P<0.0001). Patient positioning reliability was also excellent, with Pearson correlation coefficients ranging from 0.978 to 0.999 (P<0.0001).
The results of this study establish the reliability of this MR imaging system. Test-retest reliability, between-reader agreement, and patient positioning reliability were all extremely high. This study represents a first step in the overall validation of an imaging system designed to follow progression of human knee OA.
To describe the association between early radiographic osteoarthritis of the knee (ROA), knee cartilage volume and tibial bone surface area.
Cross-sectional convenience sample of 372 male and female subjects (mean age 45 years, range 26-61). Articular cartilage volume, bone area and volume were determined at the patella, medial tibial and lateral tibial compartments by processing images acquired in the sagittal plane using T1-weighted fat saturation MRI. ROA was assessed with a standing semiflexed radiograph and the OARSI atlas for joint space narrowing and osteophytosis. Both radiographs and MRIs were performed in the right knee and read by different observers.
ROA (predominantly grade 1) was present in 17% of subjects of which medial joint space narrowing was most common (14%) followed by medial osteophytes (6%). Grade one medial joint space narrowing was associated with substantial reductions in cartilage volume at both the medial and lateral tibial and patellar sites within the knee (adjusted mean difference 11-13%, all P<0.001) while grade one osteophytosis was associated with substantial increases in both lateral and medial tibial joint surface area (adjusted mean difference 10-16%, all P<0.001). In contrast, osteophytosis was not associated with a significant change in cartilage volume and joint space narrowing was not associated with a significant change in tibial bone area (all P>0.05).
Early medial compartment ROA is associated with substantial reductions in cartilage volume and increases in bone area. These large changes, when combined with similar measurement error for MRI and radiographs, suggest that MRI may be superior at detecting and hence understanding early osteoarthritis of the knee in humans.
Despite the increasing interest in knee cartilage volume as an outcome measure in studies of osteoarthritis (OA), it is unclear what components of knee cartilage will be most useful as markers of structural change in the tibiofemoral joint. This study was undertaken to longitudinally compare changes in femoral and tibial cartilage volume in patients with OA.
One hundred seventeen patients with knee OA (58.1% women; mean +/- SD age 63.7 +/- 10.2 years) were examined. Femoral and tibial cartilage volumes (medial and lateral tibiofemoral joints) were determined from T1-weighted fat-saturated magnetic resonance images of the knee from coronal views.
The study population was followed up for a mean +/- SD of 1.9 +/- 0.2 years. In the medial tibiofemoral joint, the mean +/- SD loss of cartilage was 0.15 +/- 0.30 ml/year for femoral cartilage and 0.10 +/- 0.25 ml/year for tibial cartilage. In the lateral tibiofemoral joint, the average loss was 0.15 +/- 0.22 and 0.12 +/- 0.16 ml/year for femoral and tibial cartilage, respectively. There was a significant correlation between the degree of loss of tibial cartilage and the degree of loss of femoral cartilage, in both tibiofemoral joints (r = 0.81, P < 0.001 at the medial tibiofemoral joint; r = 0.71, P < 0.001 at the lateral tibiofemoral joint).
Longitudinal changes in tibial cartilage and those in femoral cartilage are strongly related to one another. This suggests that in tibiofemoral disease, measuring tibial cartilage alone may be adequate, given the facts that measurements of the total femoral cartilage are less reproducible and there are difficulties inherent in identifying the most appropriate component of femoral cartilage to measure.
To evaluate the change in osteoarthritic (OA) knee cartilage volume over a two-year period with the use of magnetic resonance imaging (MRI) and to correlate the MRI changes with radiologic changes.
Thirty-two patients with symptomatic knee OA underwent MRI of the knee at baseline and at 6, 12, 18, and 24 months. Loss of cartilage volumes were computed and contrasted with changes in clinical variables for OA and with standardized semiflexed knee radiographs at baseline at 1 and 2 years.
Progression of cartilage loss at all followup points was statistically significant (P < 0.0001), with a mean +/- SD of 3.8 +/- 5.1% for global cartilage loss and 4.3 +/- 6.5% for medial compartment cartilage loss at 6 months, 3.6 +/- 5.1% and 4.2 +/- 7.5% at 12 months, and 6.1 +/- 7.2% and 7.6 +/- 8.6% at 24 months. Discriminant function analysis identified 2 groups of patients, those who progressed slowly (<2% of global cartilage loss; n = 21) and those who progressed rapidly (>15% of global cartilage loss; n = 11) over the 2 years of study. At baseline, there was a greater proportion of women (P = 0.001), a lower range of motion (P = 0.01), a greater circumference and higher level of pain (P = 0.05) and stiffness in the study knee, and a higher body mass index in the fast progressor group compared with the slow progressor group. No statistical correlation between loss of cartilage volume and radiographic changes was seen.
Quantitative MRI can measure the progression of knee OA precisely and can help to identify patients with rapidly progressing disease. These findings indicate that MRI could be helpful in assessing the effects of treatment with structure-modifying agents in OA.
A reduction in cartilage volume is characteristic of osteoarthritis and hence there exists a need for an accurate and reproducible method to measure in vivo cartilage volume. Quantification of cartilage volume from magnetic resonance (MR) images requires a segmentation technique such as the user-driven "Live Wire" strategy that can reliably delineate object volumes in a time-efficient manner. In the present work, the accuracy and reproducibility of the Live Wire method for the quantification of cartilage volume in MR images is evaluated.
The accuracy of the Live Wire method was assessed by comparing the MR-based volume measurement of a patellar cartilage-shaped phantom versus data calculated via water displacement. The inter- and intra-operator reproducibility of the technique was evaluated from Live Wire segmentation of the patellar cartilage volume from fat-suppressed 3-dimensional spoiled-gradient-echo images of five healthy human volunteers performed by three operators. To provide data for analysis of inter-scan reproducibility, the human scans were repeated five times with the aid of a leg-restraining jig to minimize repositioning error.
The volume of the patellar cartilage-shaped phantom measured via Live Wire segmentation of MR images was within 97.8% of its true volume. The average inter- and intra-operator coefficients of variation of three operators were 3.0% and 0.4%, respectively. The average inter-scan coefficient of variation of five repeated scans of each volunteer was 2.7%.
The data suggest that the Live Wire strategy is an accurate, reproducible, and efficient technique to measure cartilage volume in vivo in a feasible amount of operator time.
Quantifying the in vivo cartilage contact mechanics of the knee may improve our understanding of the mechanisms of joint degeneration and may therefore improve the surgical repair of the joint after injury.
To measure tibiofemoral articular cartilage contact kinematics during in vivo knee flexion.
Descriptive laboratory study.
Orthogonal fluoroscopic images and magnetic resonance image-based computer models were used to measure the motion of the cartilage contact points during a quasi-static lunge in 5 human subjects.
On the tibial plateau, the contact point moved in both the anteroposterior and the mediolateral directions during knee flexion. On the medial tibial plateau, flexion angle did not have a statistically significant effect on the location of the contact points. The total translation of the contact point from full extension to 90 degrees of flexion was less than 1.5 mm in the anteroposterior direction, whereas the translation in the mediolateral direction was more than 5.0 mm. In the anteroposterior direction, the contact points were centered on the medial tibial plateau. On the lateral tibial plateau, there was a statistically significant difference between the location of the contact point at full extension and the locations of the contact points at other flexion angles in the anteroposterior direction. No significant difference was detected between the location of the contact points at other flexion angles. The overall range of contact point motion was about 9.0 mm in the anteroposterior direction and about 4.0 mm in the mediolateral direction. The contact points were primarily on the inner half of the medial and lateral tibial plateaus (the half closest to the tibial spine). The contact points on both femoral condyles were also on the inner half of the condyles (near the condylar notch).
The tibiofemoral contact points move in 3 dimensions during weightbearing knee flexion. The medial tibiofemoral contact points remained within the central portion of the tibial plateau in the anteroposterior direction. Both the medial and lateral tibiofemoral contact points were located on the inner portions of the tibial plateau and femoral condyles (close to the tibial spine), indicating that the tibial spine may play an important role in knee stability.
The results of this study may provide important insight as to the mechanisms contributing to the development of osteoarthritis after ligament injuries.
The primary goal of this study was to describe and evaluate conditions that could influence the precision and accuracy of measuring in vivo cartilage thickness in the weight bearing regions of the knee from magnetic resonance imaging (MRI).
Three-dimensional (3D) models of the femoral cartilage were created from segmented MR images. The weight bearing regions on femoral cartilage were selected for the portion of the tibiofemoral joint that sustains contact during walking. Six regions of interest (three on each condyle) were located on the femur. Average cartilage thickness was calculated over each region. The sensitivity of the precision of the measurements to observer variability was evaluated using intra- and inter-observer reproducibility tests of cartilage thickness measurements from the MRI-derived 3D models. In addition, the quantitative influence of a rule-based protocol for segmentation was evaluated using the inter-observer reproducibility protocol. Accuracy tests were conducted on porcine knees by comparing 3D models from MR images and laser scans across weight bearing and non-weight bearing regions.
The precision was substantially better for the intra-observer tests (Coefficient of variation (CV) = 1-3%) than the inter-observer tests. Adding a rule-based protocol reduced variability in inter-observer tests substantially (CV = 6.6% vs 8.3%). Accuracy tests showed that the central and weight bearing regions on each condyle were more accurate than boundary and non-weight bearing regions. In addition, these results indicate that care should be taken when determining cartilage thickness of weight bearing regions with cartilage degenerations, since the thickness of thinner cartilage can be systematically overestimated in MR images.
A rule-based approach can substantially increase inter-observer reproducibility when measuring cartilage thickness from multiple observers. This improvement in inter-observer reproducibility could be an important consideration for longitudinal studies of disease progression. In quantifying cartilage thickness, central and weight bearing regions on each condyle can provide more accurate measurement than boundary and non-weight bearing regions with average accuracy of +/-0.2-0.3 mm. An important finding of this study was that the weight bearing regions, which are usually of the greatest clinical interest, were measured most accurately by sagittal plane imaging.
To determine the relationship between radiographic progression of joint space narrowing and cartilage loss on magnetic resonance imaging (MRI) in patients with symptomatic knee osteoarthritis (OA), and to investigate the location of MRI-based cartilage loss in the knee and its relation to radiographic progression.
Two hundred twenty-four men and women (mean age 66 years) were studied. Radiographs and MRI of the more symptomatic knee were obtained at baseline and at 15- and 30-month followup. Radiographs of the knee (with weight-bearing) were read for joint space narrowing (scale 0-3), with progression defined as any worsening in score. We used a semiquantitative method to score cartilage morphology in all 5 regions of the tibiofemoral joint, and defined cartilage loss as an increase in score (scale 0-4) at any region. We examined the relationship between progression of joint space narrowing on radiographic images and cartilage loss on MRI, using a generalized estimating equation proportional odds logistic regression, adjusted for baseline cartilage score, age, body mass index, and sex. The medial and lateral compartments were analyzed separately.
In the medial compartment, 104 knees (46%) had cartilage loss detected by MRI. The adjusted odds ratio was 3.7 (95% confidence interval 2.2-6.3) for radiographic progression being predictive of cartilage loss on MRI. However, there was still a substantial proportion of knees (80 of 189 [42%]) with cartilage loss visible on MRI when no radiographic progression was apparent. Cartilage loss occurred frequently in the central regions of the femur and tibia as well as the posterior femur region, but radiographic progression was less likely to be observed when posterior femur regions showed cartilage loss. Radiographic progression appeared specific (91%) but not sensitive (23%) for cartilage loss. Overall findings were similar for the lateral compartment.
While our results provide longitudinal evidence that radiographic progression of joint space narrowing is predictive of cartilage loss assessed on MRI, radiography is not a sensitive measure, and if used alone, will miss a substantial proportion of knees with cartilage loss.
The assessment of articular cartilage using magnetic resonance imaging has seen considerable advances in recent years. Cartilage morphologic characteristics can now be evaluated with a high degree of accuracy and reproducibility using dedicated pulse sequences, which are becoming standard at many institutions. These techniques detect clinically unsuspected traumatic cartilage lesions, allowing the physician to study their natural history with longitudinal evaluation and also to assess disease status in degenerative osteoarthritis. Magnetic resonance imaging also provides a more objective assessment of cartilage repair to augment the information obtained from more subjective clinical outcome instruments. Newly developed methods that provide detail at an ultrastructural level offer an important addition to cartilage evaluation, particularly in the detection of early alterations in the extracellular matrix. These methods have created an undeniably important role for magnetic resonance imaging in the reproducible, noninvasive, and objective evaluation and monitoring of cartilage. An overview of the advances, current techniques, and impact of magnetic resonance imaging in the setting of trauma, degenerative arthritides, and surgical treatment for cartilage injury is presented.
Quantitative magnetic resonance imaging (qMRI) of knee cartilage morphology is a powerful research tool but relies on expensive and often inaccessible 1.5 T whole-body equipment. Here we examine the reproducibility and accuracy of qMRI at 1.0 T by direct comparison with previously validated technology.
Coronal images of the knee were obtained in six healthy and six osteoarthritic participants. Two data sets were acquired with a 1.5T whole-body magnetic resonance imaging (MRI) system and two with a 1.0 T peripheral MRI system, with repositioning between scans. Proprietary software was used to analyze surface area, volume, and thickness of femoral and tibial cartilage.
At 1.0 T, precision errors for surface areas (root-mean-square (RMS) coefficient of variation (CV%)=1.7-2.6%) were higher than those at 1.5 T (1.0-2.1%). For volume and thickness, precision errors were 2.9-5.5% at 1.0 T compared to 1.6-3.4% at 1.5 T. High levels of agreement were found between the two scanners over all plates. With the exception of lateral femoral cartilage (volume and thickness), no statistically significant systematic bias was found between 1.0 T and 1.5 T.
This is the first reported study to show that knee cartilage morphology can be determined with a reasonable degree of accuracy and precision using a 1.0 T peripheral scanner. Peripheral MRI is less costly, can be performed in clinical offices, and is associated with higher patient comfort and tolerance than 1.5 T whole-body MRI. Implementation of qMRI with peripheral systems may thus permit its more widespread use in clinical research and patient care.
To compare the precision and accuracy of 3.0-T and 1.5-T magnetic resonance (MR) imaging in the quantification of cartilage volume by using direct volumetric measurements as a reference standard.
The local animal experimentation committee did not require its approval for this study. Porcine knees were obtained from an abattoir. These specimens were used to optimize imaging parameters regarding effective signal-to-noise ratio (SNRE) and contrast-to-noise ratio (CNRE) for a fat-saturated spoiled gradient-recalled acquisition in the steady state (SPGR) sequence, a water excitation SPGR sequence, and a fast spin-echo sequence at 3.0 T and a fat-saturated SPGR sequence at 1.5 T. By using the optimized sequences, 18 specimens were imaged in less than 6 minutes per sequence. A fivefold repetition of measurements of four specimens was performed for precision analysis. Cartilage was segmented by using semiautomatic software to calculate the volume. After imaging, the cartilage was scraped off and the volume was measured directly by using a saline-displacement method to calculate accuracy. Precision and accuracy errors were calculated as the root-mean-squares of the single errors per specimen.
SNRE and CNRE values, respectively, were highest for the water excitation sequence at 3.0 T (1.81 sec(-1/2) and 1.27 sec(-1/2)), followed by the fat-saturated SPGR sequence (1.52 sec(-1/2) and 1.07 sec(-1/2)). The fast spin-echo sequence and the fat-saturated SPGR sequence at 1.5 T had lower SNR(E) (1.27 sec(-1/2) and 0.59 sec(-1/2), respectively). Accuracy error for MR-based volume calculation at the femur was 5.0%, 3.0%, 21%, and 16% for the water excitation, fat-saturated SPGR, and fast spin-echo sequences at 3.0 T and the fat-saturated SPGR sequence at 1.5 T, respectively.
MR imaging at 3.0 T was shown in our study to better quantify cartilage volume. SNRE and CNRE were substantially improved, resulting in significantly higher accuracy in determining cartilage volume.
Magnetic resonance imaging (MRI) and quantitative image analysis technology has recently started to generate a great wealth of quantitative information on articular cartilage and bone physiology, pathophysiology and degenerative changes in osteoarthritis. This paper reviews semiquantitative scoring of changes of articular tissues (e.g. WORMS = whole-organ MRI scoring or KOSS = knee osteoarthritis scoring system), quantification of cartilage morphology (e.g. volume and thickness), quantitative measurements of cartilage composition (e.g. T2, T1rho, T1Gd = dGEMRIC index) and quantitative measurement of bone structure (e.g. app. BV/TV, app. TbTh, app. Tb.N, app. Tb.Sp) in osteoarthritis. For each of these fields we describe the hardware and MRI sequences available, the image analysis systems and techniques used to derive semiquantitative and quantitative parameters, the technical accuracy and precision of the measurements reported to date and current results from cross-sectional and longitudinal studies in osteoarthritis. Moreover, the paper summarizes studies that have compared MRI-based measurements with radiography and discusses future perspectives of quantitative MRI in osteoarthritis. In summary, the above methodologies show great promise for elucidating the pathophysiology of various tissues and identifying risk factors of osteoarthritis, for developing structure modifying drugs (DMOADs) and for combating osteoarthritis with new and better therapy.
Phased-array (PA) coils generally provide higher signal-to-noise ratios (SNRs) than quadrature knee coils. In this pilot study for the Osteoarthritis Initiative (OAI) we compared these two types of coils in terms of contrast-to-noise ratio (CNR), precision, and consistency of quantitative femorotibial cartilage measurements. Test-retest measurements were acquired using coronal fast low-angle shot with water excitation (FLASHwe) and coronal multiplanar reconstruction (MPR) of sagittal double-echo steady state with water excitation (DESSwe) at 3T. The precision errors for cartilage volume and thickness were <or=2.6% for the quadrature coil and <or=2.3% for the PA coil with FLASHwe, and <or=2.3%/<or=2.5% with DESSwe. The precision for aggregate medial and lateral cartilage measures was significantly higher than that for single plates, independently of coil and sequence. The PA coil measurements did not always fully agree with the quadrature coil measurements, and some differences were significant. The higher CNR of the PA coil did not translate directly into improved precision of cartilage measurement; however, summing up cartilage plates within the medial and lateral compartment reduced precision errors.
The aim of this study was to follow, over a 4(1/2)-month period, the medial tibia cartilage thickness on a meniscectomy (MNX) guinea pig osteoarthritis (OA) model and to compare with control animals, using three-dimensional high-resolution magnetic resonance imaging (3D HR-MRI).
MRI experimentations were performed in vivo at 7 T on guinea pig knee joints. 3D HR-MR images were acquired in 60 controls (SHAM) and 45 osteoarthritic animals (MNX) at four time-points (15, 45, 90 and 135 days) after surgery. Medial tibial cartilage thickness was measured from MRI images using in-house dedicated 3D software followed by a statistical analysis. At each time-point 15 SHAM and 15 MNX animals were sacrificed for histomorphometric assessments.
No significant difference of mean cartilage thickness between the groups was found at early stage (D45) using MRI; however, significant differences were found between the groups at D90 (P<0.001) and D135 (P<0.001). Histomorphometry data confirmed the pathological status of the animals and was well correlated with MRI at D15 (r=0.79, P<0.01), D45 (r=0.67, P<0.01), and D135 (r=0.39, P<0.05) for SHAM, and at D45 (r=0.63, P<0.01), and D135 (r=0.81, P<0.01) for MNX.
Medial tibial cartilage measurement based on HR-MR images enables the monitoring of longitudinal cartilage thickness changes. This technique showed significant differences between SHAM and MNX as from D90 after surgery. It could be used as a noninvasive and reproducible tool to monitor therapeutic response in this OA model.
Objective:
Cartilage loss as determined either by magnetic resonance imaging (MRI) or by joint space narrowing in X-rays is the result of cartilage erosion. However, metabolic processes within the cartilage that later result in cartilage loss may be a more accurate assessment method for early changes. Early biological processes of cartilage destruction are among other things, a combination of proteoglycan turnover, as a result of altered charge distributions, and local alterations in water content (edema). As water distribution is detectable by MRI, the aim of this study was to investigate cartilage homogeneity visualized by MRI related to water distribution, as a potential very early marker for early detection of knee osteoarthritis (OA).
Design:
One hundred and fourteen right and left knees from 71 subjects aged 22-79 years were scanned using a Turbo 3D T(1) sequence on a 0.18T MRI Esaote scanner. The medial compartment of the tibial cartilage sheet was segmented using a fully automatic voxel classification scheme based on supervised learning. From the segmented cartilage sheet, homogeneity was quantified by measuring entropy from the distribution of signal intensities inside the compartment. For each knee an X-ray was acquired and the knees were categorized by the Kellgren and Lawrence (KL) index and the joint space width (JSW) was measured. The P-values for separating the groups by each of JSW, cartilage volume, cartilage mean intensity, and cartilage homogeneity were calculated using the unpaired t-test.
Results:
The P-value for separating the group diagnosed as KL 0 from the group being KL 1 based on JSW, volume and mean signal intensity the values were P=0.9, P=0.4 and P=0.0009, respectively. In contrast, the P-value for homogeneity was P=0.0004. The precision of the measures assessed, as a test-retest root mean square coefficient of variation (RMS-CV%) was 3.9% for JSW, 7.4% for volume, 3.9% for mean signal intensity and 3.0% for homogeneity quantification.
Conclusion:
These data demonstrate that the distribution of components of the articular matrix precedes erosion, as measured by cartilage homogeneity related to water concentration. We show that homogeneity was able to separate early OA from healthy individuals in contrast to traditional volume and JSW quantifications. These data suggest that cartilage homogeneity quantification may be able to quantify early biochemical changes in articular cartilage prior to cartilage loss and thereby provide better identification of patients for OA trials who may respond better to medicinal intervention of some treatments. In addition, this study supports the feasibility of using low-field MRI in clinical studies.
Fast low angle shot (FLASH) and double echo steady state (DESS) magnetic resonance imaging (MRI) acquisitions were recently cross-calibrated for quantification of cartilage morphology at 3T. In this pilot study for the osteoarthritis (OA) initiative we compare their test-retest-precision and sensitivity to longitudinal change.
Nine participants with mild to moderate clinical OA were imaged twice each at baseline, year 1 (Y1) and year 2 (Y2). Coronal 1.5mm FLASH and sagittal 0.7mm DESS sequences were acquired; 1.5mm coronal multiplanar reformats (MPR) were obtained from the DESS. Patellar, femoral and tibial cartilage plates were quantified in a paired fashion, with blinding to time point.
In the weight-bearing femorotibial joint, average precision errors across plates were 1.8% for FLASH, 2.6% for DESS, and 3.0% for MPR-DESS. Volume loss at Y1 was not significant; at Y2 the average change across the femorotibial cartilage plates was -1.7% for FLASH, -2.8% for DESS, and -0.3% for MPR-DESS. Volume change in the lateral tibia (-5.5%; P<0.03), and in the medial (-2.9%; P<0.04) and lateral femorotibial compartments (-3.8%; P<0.03) were significant for DESS.
FLASH, DESS and MPR-DESS all displayed adequate test-retest precision. Although the comparison between protocols is limited by the small number of participants and by the relatively small longitudinal change in cartilage morphology in this pilot study, the data suggest that significant change can be detected with MRI in a small sample of OA subjects over 2 years.
MRI-based cartilage morphometry was previously validated in the absence of gadopentate dimeglumine (Gd-DTPA). However, Gd-DTPA is required for compositional (proteoglycan) imaging using delayed gadolinium-enhanced MRI of cartilage (dGEMRIC). Therefore, the effect of Gd-DTPA on cartilage morphometry was studied. A total of 165 female participants (67 with and 98 without osteoarthritis [OA]) were imaged at 3.0 Tesla before and 2 hr after intravenous Gd-DTPA injection. Flip angles in post-Gd-DTPA scans varied between 12 degrees and 35 degrees . Cartilage volume and thickness of post- vs. pre-Gd-DTPA scans showed intraclass correlation coefficients (ICCs) of 0.85 > or = r > or = 0.95, mean differences between -2.1% and +1.1%, and standard deviations (SDs) of differences between 4.7% and 9.2%. Mixed-effect models found no consistent impact of flip angle and OA status on post- vs. pre-Gd-DTPA differences. Accurate morphological measurements of cartilage can be obtained after Gd-DTPA injection, allowing compositional and morphological imaging to be combined into one session.
The in vivo precision (reproducibility) of quantitative MRI is of particular importance in osteoarthritis (OA) progression of small magnitude and response to therapy. In this study, three-dimensional high-resolution MRI performed at 7 T was used to assess the short-term reproducibility of measurements of mean tibial cartilage thickness in a meniscectomized guinea pig model of OA. MR image acquisition was repeated five times in nine controls (SHAM) and 10 osteoarthritic animals 3 months after meniscectomy (MNX), in vivo. The animals were then killed for histomorphometric assessment and correlation with the MRI-based measurements. Medial tibial cartilage thickness was measured on MR images using semi-automatic dedicated 3D software developed in-house. The reproducibility of measurements of cartilage thickness was assessed by five repeated MRI examinations with a short recovery delay between examinations (48 h). The computed coefficients of variation were 8.9% for the SHAM group and 8.2% for the MNX group. The coefficients of variation were compatible with expected thickness variations between normal and pathological animals. A positive agreement and significant partial correlation (Spearman r' = 0.74; P < 0.01) between the MRI and histomorphometric data was established. Three-dimensional high-resolution MRI is a promising non-invasive research tool for in vivo follow-up. This modality could be used for staging and monitoring therapy response in small-animal models of OA.
To study the articular cartilage surface curvature determined automatically from magnetic resonance (MR) knee scans, evaluate accuracy of the curvature estimates on digital phantoms, and an evaluation of their potential as disease markers for different stages of osteoarthritis (OA).
Knee MR data were acquired using a low-field 0.18T scanner, along with posteroanterior x-rays for evaluation of radiographic signs of OA according to the Kellgren-Lawrence index (KL). Scans from a total of 114 knees from test subjects with KL 0-3, 59% females, ages 21-79 years were evaluated. The surface curvature for the medial tibial compartment was estimated automatically on a range of scales by two different methods: Euclidean shortening flow and boundary normal comparison on a cartilage shape model. The curvature estimates were normalized for joint size for intersubject comparisons. Digital phantoms were created to establish the accuracy of the curvature estimation methods.
A comparison of the two curvature estimation methods to ground truth yielded absolute pairwise differences of 1.1%, and 4.8%, respectively. The interscan reproducibility for the two methods were 2.3% and 6.4% (mean coefficient of variation), respectively. The surface curvature was significantly higher in the OA population (KL > 0) compared with the healthy population (KLi = 0) for both curvature estimates, with P values of .000004 and .000006, respectively. The shape model based curvature estimate could also separate healthy from borderline OA (KL = 1) populations (P = .005).
The phantom study showed that the shape model method was more accurate for a coarse-scale analysis, whereas the shortening flow estimated fine scales better. Both the fine- and the coarse-scale curvature estimates distinguished between healthy and OA populations, and the coarse-scale curvature could even distinguish between healthy and borderline OA populations. The highly significant differences between populations demonstrate the potential of cartilage curvature as a disease marker for OA.
MRI-based measures of cartilage morphology are being increasingly used as surrogate markers in osteoarthritis. In contrast to other knee joint surfaces, quantitative analysis of the femoral condyles from sagittal MRI suffers from limited precision. The objective, therefore, was to develop a technique for reproducible assessment of femoral cartilage morphology from coronal image data. Coronal MR images (3D T(1)-w FLASHwe) of the knee were obtained in 16 healthy volunteers and in 7 patients with severe osteoarthritis (OA, prior to knee arthroplasty), with repositioning between repeated scans. After segmentation the cartilage volume, thickness, and joint surface areas were quantified in the tibia and in an anatomically defined region of the femoral condyle. Immediate test-retest interscan precision errors (CV%) for femoral cartilage volume were 3.0% (SD = 26 microl) and 3.2% (29 microl) medially and laterally in volunteers, and 3.0% (34 microl) and 7.0% (37 microl) in OA patients. The estimated loss, from cross sectional data, in the patients in the medial femoral condyle (-61%/-4.4 SD) was higher than that in the medial tibia (-45%/-3.1 SD) and compared favorably with precision errors (ratio > 16:1). The technique proposed overcomes some of the problems associated with sagittal scans and thus shows high promise for reliable assessment of femoro-tibial cartilage loss in OA.
To assess the effects of interference screws, which are commonly used to surgically fix an anterior cruciate ligament (ACL) graft in the ACL-deficient knee, and magnetic field strength on cartilage volume and thickness measurements with quantitative magnetic resonance imaging (qMRI).
Five cadaver knees were imaged using a cartilage-sensitive sequence (T1-weighted water-excitation, three-dimensional (3D) fast low-angle shot) on 1.5T and 3T scanners with and without interference screws implanted. The tibiofemoral articular cartilage was segmented and reconstructed from the magnetic resonance images, and volume and thickness measurements were made on the resulting 3D models.
Although several load-bearing regions showed significant differences in volume and thickness between magnet strengths, most showed no significant difference between screw conditions. The medial tibial cartilage showed a mean decrease in volume of 5.9% and 8.0% in the presence of interference screws at 3T and 1.5T, respectively. At 3T and 1.5T, the medial tibial cartilage showed a mean decrease in thickness of 7.0% and 12.0%, respectively, in the presence of interference screws.
Caution should be used when interpreting thickness and volume of cartilage at 3T in the presence of interference screws, particularly in the medial tibial compartment. Additionally, 3T and 1.5T qMRI should not be used interchangeably to assess structural changes in tibiofemoral articular cartilage during longitudinal studies.
Quantitative diagnostic tools for osteoarthritis (OA) are important for evaluating the treatment response to structure-modifying drugs. This study was undertaken to test the technical validity (accuracy) of quantitative magnetic resonance imaging (qMRI) for reliable determination of the total bone interface area, percentage of cartilaginous (denuded) joint surface area, and cartilage thickness in OA.
High-resolution MRIs of femorotibial and patellar cartilage were acquired in 21 patients prior to total knee arthroplasty, using a T1-weighted gradient-echo sequence with water excitation. After segmentation of original bone interface areas (before disease onset) and the actual cartilage layer, the percentages of cartilaginous joint surface area, cartilage thickness, and cartilage volume were determined using proprietary software. During surgery, the patella and the medial and lateral tibia were resected. Results obtained with qMRI were compared with those obtained by direct image analysis of surface area, cartilage thickness, and cartilage volume of the surgically removed tissue.
Pairwise differences between results obtained with qMRI and morphologic analysis were +/-4.6% for percentage of cartilaginous surface area, +/-8.9% for cartilage thickness, and +/-9.1% for cartilage volume. Correlation coefficients ranged from 0.92 (thickness) to 0.98 (volume).
Quantitative MRI permits technically accurate and differential assessment of increases in eroded joint surface area and reductions in cartilage thickness in OA. The surrogate validity of these parameters requires testing in longitudinal studies. These parameters may be advantageous over determination of cartilage volume alone when diagnosing OA, exploring its progression, or testing responsiveness to new therapies.
The segmentation of structure from 3D images is an inherently
difficult problem and a bottleneck to the widespread use of computer
vision in such applications as medical imaging. Local low-level
voxel-based features must somehow be integrated to obtain global
object-based descriptions. Deformable models in the form of snakes,
balloons, level sets, and bubbles have been proposed for this task. In
this paper, we extend the reaction-diffusion segmentation bubble
technique to three dimensions. In generalizing this approach to 3D, two
separate issues arise. First, should segmentations be achieved by
treating images as a series of disjoint 2D slides, as 2D slices with
interslice interactions, or intrinsically as a 3D image? We will show
that the existence of saps of information in low-level features guides
us to make maximal use of continuity in all directions, thus advocating
an intrinsic 3D approach. The treatment of bubbles in 3D, however,
requires the generalization of the reaction-diffusion space. While the
reaction process is trivially extended, the generalization of diffusion
is not straightforward. We utilize a particular mean-Gauss curvature
deformation to serve as the regularizing diffusion process. The
resulting 3D reaction-diffusion bubbles are intrinsic, can deal with a
variety of gaps, and place captured surfaces in a hierarchy of scale.
The process is illustrated on MRI images of the ventricle cavity and the
vascular structure in MRA images
Quantitative imaging of cartilage morphology at
- F Eckstein
- Rj Buck
- Bt Wyman
- Kotyk Jj
- Le Hellio
- Mp
- Ae
Eckstein F, Buck RJ, Wyman BT, Kotyk JJ, Hellio Le Graverand MP, Remmers AE, et al. Quantitative imaging of cartilage morphology at
Volumetric cartilage measurements of Porcine knee at 1.5-T and
- J Bauer
- Ross S C Krause
- R J Krug
- E Ozhinsky
Bauer J, Krause S, Ross C, Krug R, Carballido-Gamio J, Ozhinsky E, et al. Volumetric cartilage measurements of Porcine knee at 1.5-T and