Figure 2 - uploaded by Elizabeth Louise Thomas
Content may be subject to copyright.
Bland-Altman plot (a) of iliopsoas muscle volumes determined with CNN-based and manual segmentations ( n = 90 ), using a six-fold cross-validation experiment. Dotted lines represent the average bias ( −0.2 %) and the 95% limits of agreement. Overlays of the CNN-based and manual segmentations for two subjects (b,c), where the manual annotation is red, the CNN segmentation is green and the overlap is yellow.
Source publication
Abstract Psoas muscle measurements are frequently used as markers of sarcopenia and predictors of health. Manually measured cross-sectional areas are most commonly used, but there is a lack of consistency regarding the position of the measurement and manual annotations are not practical for large population studies. We have developed a fully automa...
Contexts in source publication
Context 1
... A summary of the cross-validation experiment may be found in Supplementary Table S2. The average bias was −0.2 % with upper and lower limits of agreement being 13.3% and −13.7 %, respectively (Fig. 2). The overlap between the CNN-based and manual segmentations for two subjects is also provided in Fig. 2, where the DSCs are 0.85 (left) and 0.90 (right) for (b) and 0.96 in both for (c). With consistent DSCs from the cross- www.nature.com/scientificreports/ validation experiment showing a robust model performance on both muscles, we ...
Context 2
... A summary of the cross-validation experiment may be found in Supplementary Table S2. The average bias was −0.2 % with upper and lower limits of agreement being 13.3% and −13.7 %, respectively (Fig. 2). The overlap between the CNN-based and manual segmentations for two subjects is also provided in Fig. 2, where the DSCs are 0.85 (left) and 0.90 (right) for (b) and 0.96 in both for (c). With consistent DSCs from the cross- www.nature.com/scientificreports/ validation experiment showing a robust model performance on both muscles, we trained a final model using the entire 90 available manual annotations. Example segmentations from our ...
Context 3
... final set of three subjects represent subjects whose left and right iliopsoas muscles differ in volume ( difference in volume ≈ 93 ml for j and k, difference in volume = 182 ml for l). We can see that the model performs well for all of them, with additional details regarding model validation provided in Supplementary Fig. S2. ...
Similar publications
Purpose
Nutrition is an important outcome predictor in oncology patients including treatment response, physical disability, quality of life, and overall survival. Sarcopenia (loss of skeletal muscle mass and function) is a demonstrated marker of nutritional status in adults, but data are more limited in children. The purpose of this study was to ev...
Citations
... We speculate that the poor outcome of THAs with lower PMI may be due to the large T1PA in these groups, resulting in significant kyphosis and persistent low back pain (Fig. 5). The iliopsoas muscle, which flexes the hip and stabilizes the lumbar spine, plays a crucial role in maintaining upright posture and initiating hip flexion during gait [19]. The iliopsoas muscle also interacts with surrounding structures, such as the hip joint capsule and the lumbar spine, impacting overall movement coordination and spinal alignment. ...
... There is growing interest in the association between the crosssectional area of the iliopsoas and therapeutic outcomes. The psoas cross-sectional area has been used as an index of total skeletal muscle mass and identified as a prognostic factor for cardiovascular, thoracic, gastrointestinal, colorectal, oncologic, and transplantation surgery outcomes [19]. Moreover, the cross-sectional area of the iliopsoas muscle may provide an indirect measure of nutritional status, providing an avenue to innovative areas of nutritional research Fig. 4 The ROC curves showing the accuracy of the cut-offs of the psoas muscle index for females (solid) and males (dotted). ...
... We also found age as a predictor of outcomes post-THA. This supported previous studies which have reported a decrease in the cross-sectional area of the iliopsoas muscle related to the development of age-related sarcopenia [1,19]. Understanding the effects of aging on muscle size can inform interventions to prevent muscle loss. ...
The aim of this study is to assess the association between the psoas muscle index (PMI) and total hip arthroplasty (THA) outcomes. This is a critical issue as sarcopenia has been associated with poor patient satisfaction post-THA.
This was a retrospective case–control study of 205 THAs, with a mean follow-up of 3.6 (range, 2.0–5.5) years. Age, sex, serum immune markers, spinopelvic parameters, PMI (quantified as the cross-sectional area of the psoas, bilaterally, at L3 divided by the individual’s height squared), and patient-reported outcomes were compared between patients ‘with’ (n = 118) and ‘without’ (n = 87) achievement of a minimum clinically important difference (MCID) improvement in the EuroQol 5-Dimension (EQ-5D), post-THA. Logistic regression and receiver operating characteristic curve analyses were used to identify predictive factors.
A ≥ MCID improvement in the EQ-5D was associated with the PMI (odds ratio, 0.75; 95% confidence interval, 0.63–0.91; P = 0.028), prognostic nutritional index (odds ratio, 0.85; 95% confidence interval, 0.45–0.94; P = 0.043), and age (odds ratio, 1.09; 95% confidence interval, 1.01–1.18; P = 0.044). After adjusting the PMI threshold to 4.0 cm2/m2 for females and 6.4 cm2/m2 for males, there were significant differences in serum factors (P = 0.041 for albumin and P = 0.016 for a prognostic nutritional index < 40), MCID (P < 0.001 for EQ-5D, P < 0.001 for low back pain, and P = 0.008 for the Hip Disability and Osteoarthritis Outcome Score Joint Replacement score), patient satisfaction (P = 0.003), and T1 pelvic angle (P = 0.030).
The PMI, which is associated with nutritional status and global sagittal spinal deformity, does predict THA outcomes. Therefore, it can be useful when discussing THA expectations with patients.
... Other indicators of physical fitness, such as walking distance, functional leg strength, grip strength, inspiratory muscle strength, and gait speed, have also been correlated with improved surgical outcomes [4]. Psoas muscle volume (which is negatively correlated with age [5] and a marker for frailty [6]) predicts mortality and difficulty with ventilator weaning in critically ill surgical patients [7]. The impact of physical fitness is not limited to surgical, oncologic, or critically ill patients, with higher VO 2 max improving all-cause mortality in adults with no detectable limit [8]. ...
Background:
There is a clear relationship between quantitative measures of fitness (e.g., VO2 max) and outcomes after surgical procedures. Whether or not fitness is a modifiable risk factor and what underlying biological processes drive these changes are not known. The purpose of this study was to evaluate the moderate exercise training effect on sepsis outcomes (survival) as well as the hepatic biological response. We chose to study the liver because it plays a central role in the regulation of immune defense during systemic infection and receives blood flow directly from the origin of infection (gut) in the cecal ligation and puncture (CLP) model.
Methods:
We randomized 50 male (♂) and female (♀) Sprague-Dawley rats (10 weeks, 340 g) to 3 weeks of treadmill exercise training, performed CLP to induce polymicrobial "sepsis," and monitored survival for five days (Part I). In parallel (Part II), we randomized 60 rats to control/sedentary (G1), exercise (G2), exercise + sham surgery (G3), CLP/sepsis (G4), exercise + CLP [12 h (G5) and 24 h (G6)], euthanized at 12 or 24 h, and explored molecular pathways related to exercise and sepsis survival in hepatic tissue and serum.
Results:
Three weeks of exercise training significantly increased rat survival following CLP (polymicrobial sepsis). CLP increased inflammatory markers (e.g., TNF-a, IL-6), which were attenuated by exercise. Sepsis suppressed the SOD and Nrf2 expression, and exercise before sepsis restored SOD and Nrf2 levels near the baseline. CLP led to increased HIF1a expression and oxidative and nitrosative stress, the latter of which were attenuated by exercise. Haptoglobin expression levels were increased in CLP animals, which was significantly amplified in exercise + CLP (24 h) rats.
Conclusions:
Moderate exercise training (3 weeks) increased the survival in rats exposed to CLP, which was associated with less inflammation, less oxidative and nitrosative stress, and activation of antioxidant defense pathways.
... We processed all available image data for cases and controls using our previously published image processing and segmentation pipelines [41]. We subsequently generated imagederived phenotypes (IDPs) of abdominal organs, adipose tissue, and muscles using convolutional neural networks [41][42][43]. As in [41] we performed validation procedures that included a visual inspection in three separate groups of scans: the smallest fifty by volume, the largest fifty by volume and fifty randomly-selected scans not included in the previous two groups. ...
The main drivers of COVID-19 disease severity and the impact of COVID-19 on long-term health after recovery are yet to be fully understood. Medical imaging studies investigating COVID-19 to date have mostly been limited to small datasets and post-hoc analyses of severe cases. The UK Biobank recruited recovered SARS-CoV-2 positive individuals (n = 967) and matched controls (n = 913) who were extensively imaged prior to the pandemic and underwent follow-up scanning. In this study, we investigated longitudinal changes in body composition, as well as the associations of pre-pandemic image-derived phenotypes with COVID-19 severity. Our longitudinal analysis, in a population of mostly mild cases, associated a decrease in lung volume with SARS-CoV-2 positivity. We also observed that increased visceral adipose tissue and liver fat, and reduced muscle volume, prior to COVID-19, were associated with COVID-19 disease severity. Finally, we trained a machine classifier with demographic, anthropometric and imaging traits, and showed that visceral fat, liver fat and muscle volume have prognostic value for COVID-19 disease severity beyond the standard demographic and anthropometric measurements. This combination of image-derived phenotypes from abdominal MRI scans and ensemble learning to predict risk may have future clinical utility in identifying populations at-risk for a severe COVID-19 outcome.
... Measurement of total body muscle volume and distribution by MRI has been typically limited to small cohorts, due to the cost and time-consuming requirements of image acquisition and analysis. The lack of automated techniques led researchers to rely on the measurement of single or multiple crosssectional areas as an index of overall muscle mass, with a single slice at the third lumbar vertebra (L3) [5], individual muscles groups such as the iliopsoas [6] or anatomical groupings such as the 'thigh muscles' [4] being among the most popular approaches. Whether proxies of muscle volume provide sufficient information to discern the overall impact of muscle volume and quality on health and ageing is yet to be fully ascertained. ...
Background
Magnetic resonance imaging (MRI) enables direct measurements of muscle volume and quality, allowing for an in-depth understanding of their associations with anthropometric traits, and health conditions. However, it is unclear which muscle measurements: total muscle volume, regional measurements or measurements of muscle quality, such as intermuscular adipose tissue (IMAT) or proton density fat fraction (PDFF), are most informative to detect changes and associations with relevant health outcomes such as sarcopenia and frailty.
Methods
We developed a pipeline to automatically segment and extract image-derived phenotypes (IDPs) including total and regional muscle volumes and measures of muscle quality, and applied it to the neck-to-knee Dixon images in 44,520 UK Biobank participants. We also segmented paraspinal muscle from 2D quantitative MRI to quantify muscle PDFF and iron concentration. We performed linear regression to assess associations with anthropometric and lifestyle factors. We further applied logistic regression to investigate the association between these IDPs and sarcopenia and frailty.
Results
All muscle measurements were negatively associated with age, waist-to-hip ratio, and Townsend deprivation index while they were positively associated with body mass index, alcohol intake, hand grip strength, physical activity and were significantly higher in men. Additionally, IMAT, (corrected for muscle volume) and paraspinal muscle PDFF were significantly higher in female compared with male participants. Sarcopenia was associated with reduced muscle volume, while frailty was associated with increased IMAT and paraspinal PDFF.
Conclusions
Our fully automated method enables the quantification of muscle volumes and quality suitable for large population-based studies. While the choice of muscle measurements is important particularly when investigating associations with health conditions, most provide consistent differences relating to the detection of age, sex, body composition and lifestyle changes on muscularity.
... We predicted the F and Ŵ volumes for more than 70 participants that were affected by swaps identified when selecting the training datasets, and performed visual inspection on the volumes to identify whether or not the models were able to successfully perform fat-water separation where the scanner software failed. As a final qualitative evaluation of our model on swapped data, we used the predicted fat and water channels (F ,Ŵ ) to perform segmentations using publicly available code [8,55] in order to highlight the importance of correct fat-water separation. We also implemented the dixonfix method by Glocker et al. [20] to compare our method with a closelyrelated alternative where the code is publicly available. ...
Unlabelled:
Chemical-shift encoded MRI (CSE-MRI) is a widely used technique for the study of body composition and metabolic disorders, where derived fat and water signals enable the quantification of adipose tissue and muscle. The UK Biobank is acquiring whole-body Dixon MRI (a specific implementation of CSE-MRI) for over 100,000 participants. Current processing methods associated with large whole-body volumes are time intensive and prone to artifacts during fat-water separation performed by the scanner, making quantitative analysis challenging. The most common artifacts are fat-water swaps, where the labels are inverted at the voxel level. It is common for researchers to discard swapped data (generally around 10%), which is wasteful and may lead to unintended biases. Given the large number of whole-body Dixon MRI acquisitions in the UK Biobank, thousands of swaps are expected to be present in the fat and water volumes from image reconstruction performed on the scanner. If they go undetected, errors will propagate into processes such as organ segmentation, and dilute the results in population-based analyses. There is a clear need for a robust method to accurately separate fat and water volumes in big data collections like the UK Biobank. We formulate fat-water separation as a style transfer problem, where swap-free fat and water volumes are predicted from the acquired Dixon MRI data using a conditional generative adversarial network, and introduce a new loss function for the generator model. Our method is able to predict highly accurate fat and water volumes free from artifacts in the UK Biobank. We show that our model separates fat and water volumes using either single input (in-phase only) or dual input (in-phase and opposed-phase) data, with the latter producing superior results. Our proposed method enables faster and more accurate downstream analysis of body composition from Dixon MRI in population studies by eliminating the need for visual inspection or discarding data due to fat-water swaps.
Supplementary information:
The online version contains supplementary material available at 10.1186/s40537-022-00677-1.
... Support with elbow flexion and keep both lower limbs pressed down with a wide sling. 4 4) Omega. Under the premise of maintaining the plank posture, retract the abdomen and levator anus. ...
Introduction
Lumbar muscle strain is a chronic injury to soft tissues such as the lumbar muscles, ligaments, and fascia. Functional exercise has specific applications in treating lumbar muscle injuries caused by sports. However, analyses on the treatment results in the psoas muscle are inconclusive.
Objective
Analyze the clinical efficacy of functional exercise in treating psoas muscle dysfunction.
Methods
10 athletes diagnosed with lumbar muscle strain received continuous training with a functional exercise protocol for two weeks, five times a week. Clinical efficacy was assessed by visual analog scale for pain score and Prokin254 for proprioception ability indices before and after treatment. The article adopts a mathematical statistics analysis method to analyze the therapeutic effect of motor function exercise with SPSS 13.0.
Results
Patients reported a reduction of pain in the muscles under exertion after functional exercise. The results were significantly different (P<0.05). Patients’ lumbar strength was significantly improved. This index has a considerable statistical difference (P<0.05).
Conclusion
Functional exercise showed a positive effect on the treatment of psoas muscle injury. The research results of this article can provide an effective training protocol for the rehabilitation of people with a psoas muscle strain. Evidence Level II; Therapeutic Studies - Investigating the result.
Psoas Muscles; Athlete; Sports; Human Physical Training
... Also, the psoas muscle intensity to cerebrospinal fluid ratio was evaluated, which is a valid representative measure of myosteatosis 26 . Fitzpatrick et al. (2020) found that there is an accelerated muscle volume decline in aging males 27 . They created a strong as well as dependable model with the use of Dixon sequence and a convolutional neural network to automatically segment iliopsoas muscles. ...
... Also, the psoas muscle intensity to cerebrospinal fluid ratio was evaluated, which is a valid representative measure of myosteatosis 26 . Fitzpatrick et al. (2020) found that there is an accelerated muscle volume decline in aging males 27 . They created a strong as well as dependable model with the use of Dixon sequence and a convolutional neural network to automatically segment iliopsoas muscles. ...
... They created a strong as well as dependable model with the use of Dixon sequence and a convolutional neural network to automatically segment iliopsoas muscles. They proved that the specific method can be applied in a large cohort, a fact that gives the ability to conduct future population-wide studies on the usefulness of iliopsoas muscle as a prognostic factor regarding the course of health problems 27 . ...
... This operation is cumbersome and timeconsuming, which is why there is intense research on the automation of this operation [7][8][9][10][11]. However, all automatic segmentation algorithms are validated assuming the manual segmentation as the true value [12,13]; thus, it becomes very important to quantify the repeatability of the manual segmentation of skeletal muscle volume on MRI images. ...
The manual segmentation of muscles on magnetic resonance images is the gold standard procedure to reconstruct muscle volumes from medical imaging data and extract critical information for clinical and research purposes. (Semi)automatic methods have been proposed to expedite the otherwise lengthy process. These, however, rely on manual segmentations. Nonetheless, the repeatability of manual muscle volume segmentations performed on clinical MRI data has not been thoroughly assessed. When conducted, volumetric assessments often disregard the hip muscles. Therefore, one trained operator performed repeated manual segmentations (n = 3) of the iliopsoas (n = 34) and gluteus medius (n = 40) muscles on coronal T1-weighted MRI scans, acquired on 1.5 T scanners on a clinical population of patients elected for hip replacement surgery. Reconstructed muscle volumes were divided in sub-volumes and compared in terms of volume variance (normalized variance of volumes - nVV), shape (Jaccard Index-JI) and surface similarity (maximal Hausdorff distance-HD), to quantify intra-operator repeatability. One-way repeated measures ANOVA (or equivalent) tests with Bonferroni corrections for multiple comparisons were conducted to assess statistical significance. For both muscles, repeated manual segmentations were highly similar to one another (nVV: 2-6%, JI > 0.78, HD < 15 mm). However, shape and surface similarity were significantly lower when muscle extremities were included in the segmentations (e.g., iliopsoas: HD -12.06 to 14.42 mm, P < 0.05). Our findings show that the manual segmentation of hip muscle volumes on clinical MRI scans provides repeatable results over time. Nonetheless, extreme care should be taken in the segmentation of muscle extremities.
... Future research can therefore focus on muscle volume instead of muscle area. Some studies with deep learning have already been performed correctly determining volume of iliopsoas (26). However, it is currently unknown if complete volume is also predictive of clinical outcomes. ...
Background
Manual muscle mass assessment based on Computed Tomography (CT) scans is recognized as a good marker for malnutrition, sarcopenia, and adverse outcomes. However, manual muscle mass analysis is cumbersome and time consuming. An accurate fully automated method is needed. In this study, we evaluate if manual psoas annotation can be substituted by a fully automatic deep learning-based method.
Methods
This study included a cohort of 583 patients with severe aortic valve stenosis planned to undergo Transcatheter Aortic Valve Replacement (TAVR). Psoas muscle area was annotated manually on the CT scan at the height of lumbar vertebra 3 (L3). The deep learning-based method mimics this approach by first determining the L3 level and subsequently segmenting the psoas at that level. The fully automatic approach was evaluated as well as segmentation and slice selection, using average bias 95% limits of agreement, Intraclass Correlation Coefficient (ICC) and within-subject Coefficient of Variation (CV). To evaluate performance of the slice selection visual inspection was performed. To evaluate segmentation Dice index was computed between the manual and automatic segmentations (0 = no overlap, 1 = perfect overlap).
Results
Included patients had a mean age of 81 ± 6 and 45% was female. The fully automatic method showed a bias and limits of agreement of −0.69 [−6.60 to 5.23] cm², an ICC of 0.78 [95% CI: 0.74–0.82] and a within-subject CV of 11.2% [95% CI: 10.2–12.2]. For slice selection, 84% of the selections were on the same vertebra between methods, bias and limits of agreement was 3.4 [−24.5 to 31.4] mm. The Dice index for segmentation was 0.93 ± 0.04, bias and limits of agreement was −0.55 [1.71–2.80] cm².
Conclusion
Fully automatic assessment of psoas muscle area demonstrates accurate performance at the L3 level in CT images. It is a reliable tool that offers great opportunities for analysis in large scale studies and in clinical applications.
... We processed all available image data for cases and controls using our previously published image processing pipeline [37]. We subsequently generated image-derived phenotypes (IDPs) of abdominal organs, adipose tissue, and muscles using convolutional neural networks [37][38][39]. We included a total of twelve IDPs in this study: volumes of abdominal subcutaneous adipose tissue (ASAT), visceral adipose tissue (VAT), liver, lungs, iliopsoas muscles, kidneys, pancreas, spleen, as well as proton density fat fraction (PDFF) measures of liver and pancreas fat content, and organ iron concentration of the liver and pancreas. ...
The main drivers of COVID-19 disease severity and the impact of COVID-19 on long-term health after recovery are yet to be fully understood. Medical imaging studies investigating COVID-19 to date have mostly been limited to small datasets and post-hoc analyses of severe cases. The UK Biobank recruited recovered SARS-CoV-2 positive individuals (n=967) and matched controls (n=913) who were extensively imaged prior to the pandemic and underwent follow-up scanning. In this study, we investigated longitudinal changes in body composition, as well as the associations of pre-pandemic image-derived phenotypes with COVID-19 severity. Our longitudinal analysis, in a population of mostly mild cases, associated a decrease in lung volume with SARS-CoV-2 positivity. We also observed that increased visceral adipose tissue and liver fat, and reduced muscle volume, prior to COVID-19, were associated with COVID-19 disease severity. Finally, we trained a machine classifier with demographic, anthropometric and imaging traits, and showed that visceral fat, liver fat and muscle volume have prognostic value for COVID-19 disease severity beyond the standard demographic and anthropometric measurements. This combination of image-derived phenotypes from abdominal MRI scans and ensemble learning to predict risk may have future clinical utility in identifying populations at-risk for a severe COVID-19 outcome.
