Fig 2 - available via license: Creative Commons Attribution 3.0 Unported
Content may be subject to copyright.
CT images of the multipurpose chest phantom showing the 8 mm low-attenuation (anterior) and high-attenuation (posterior) simulated ground-glass opacities (GGOs). These images were acquired using (a) a clinical lung cancer screening protocol (51 effective mAs) and reconstructed with filtered back-projection (FBP), or (b) and (c) a reduced tube current protocol (20 effective mAs) and reconstructed with adaptive statistical iterative reconstruction (ASIR) or model-based iterative reconstruction (MBIR), respectively.
Source publication
The purpose of this study was to reduce the radiation dosage associated with computed tomography (CT) lung cancer screening while maintaining overall diagnostic image quality and definition of ground‐glass opacities (GGOs). A lung screening phantom and a multipurpose chest phantom were used to quantitatively assess the performance of two iterative...
Context in source publication
Citations
... where | − | is the distance between the location of the mth pixel and the signal center, R is the signal radius, and As is the parameter that controls the signal amplitude. The parameters were set as follows: R was fixed at 4 mm, z was fixed at 4, and signal amplitude was set to −870 HU, mimicking the attenuation of low-attenuation ground-glass opacities (GGO) [32]. In this task, the signal was invariant, referring to the signal-known-exactly (SKE) paradigm. ...
... where |r m − r c | is the distance between the location of the mth pixel and the signal center, R is the signal radius, and A s is the parameter that controls the signal amplitude. The parameters were set as follows: R was fixed at 4 mm, z was fixed at 4, and signal amplitude was set to −870 HU, mimicking the attenuation of low-attenuation ground-glass opacities (GGO) [32]. In this task, the signal was invariant, referring to the signal-known-exactly (SKE) paradigm. ...
Many new reconstruction techniques have been deployed to allow low-dose CT examinations. Such reconstruction techniques exhibit nonlinear properties, which strengthen the need for a task-based measure of image quality. The Hotelling observer (HO) is the optimal linear observer and provides a lower bound of the Bayesian ideal observer detection performance. However, its computational complexity impedes its widespread practical usage. To address this issue, we proposed a self-supervised learning (SSL)-based model observer to provide accurate estimates of HO performance in very low-dose chest CT images. Our approach involved a two-stage model combining a convolutional denoising auto-encoder (CDAE) for feature extraction and dimensionality reduction and a support vector machine for classification. To evaluate this approach, we conducted signal detection tasks employing chest CT images with different noise structures generated by computer-based simulations. We compared this approach with two supervised learning-based methods: a single-layer neural network (SLNN) and a convolutional neural network (CNN). The results showed that the CDAE-based model was able to achieve similar detection performance to the HO. In addition, it outperformed both SLNN and CNN when a reduced number of training images was considered. The proposed approach holds promise for optimizing low-dose CT protocols across scanner platforms.
... Within the United Kingdom the TLHC Standard Protocol 7 states: "The calculated radiation dose delivered to each individual is below 2 mSv (based on a median standard 70-kg adult)." Image quality levels for lung screening CT are mentioned infrequently in the published literature [15][16][17][18][19] ; however, there are many articles, which present indications of the patient doses delivered. 4, [15][16][17][18][19][20][21][22][23][24][25][26][27] There are notable differences in how these doses are presented, with some studies presenting solely values of effective dose, 18,20,22,25,26 some presenting values of CTDIvol or dose length product (DLP) from patient scans 4,20,21,26 whilst others present these same metrics for scans of standardized phantoms. ...
... Image quality levels for lung screening CT are mentioned infrequently in the published literature [15][16][17][18][19] ; however, there are many articles, which present indications of the patient doses delivered. 4, [15][16][17][18][19][20][21][22][23][24][25][26][27] There are notable differences in how these doses are presented, with some studies presenting solely values of effective dose, 18,20,22,25,26 some presenting values of CTDIvol or dose length product (DLP) from patient scans 4,20,21,26 whilst others present these same metrics for scans of standardized phantoms. 16,18,19,25 Lung cancer mortality in the Yorkshire and Humber region is higher that the UK national rate (age standardized mortality per 100 000 persons is 64.7 in Yorkshire and Humber versus 54.7 in England). ...
... 4, [15][16][17][18][19][20][21][22][23][24][25][26][27] There are notable differences in how these doses are presented, with some studies presenting solely values of effective dose, 18,20,22,25,26 some presenting values of CTDIvol or dose length product (DLP) from patient scans 4,20,21,26 whilst others present these same metrics for scans of standardized phantoms. 16,18,19,25 Lung cancer mortality in the Yorkshire and Humber region is higher that the UK national rate (age standardized mortality per 100 000 persons is 64.7 in Yorkshire and Humber versus 54.7 in England). 28,29 The Yorkshire Lung Screening Trial (YLST) 5 was established in response to these higher rates, and recruitment to the trial started in Leeds in 2018 using community-based screening in mobile units, using a model similar to those successfully used elsewhere in the United Kingdom. ...
Objectives
To evaluate radiation doses for all low-dose CT scans performed during the first year of a lung screening trial.
Methods
For all lung screening scans that were performed using a CT protocol that delivered image quality meeting the RSNA QIBA criteria, radiation dose metrics, participant height, weight, gender, and age were recorded. Values of volume CT dose index (CTDIvol) and dose length product (DLP) were evaluated as a function of weight in order to assess the performance of the scan protocol across the participant cohort. Calculated effective doses were used to establish the additional lifetime attributable cancer risks arising from trial scans.
Results
Median values of CTDIvol, DLP, and effective dose (IQR) from the 3521 scans were 1.1 mGy (0.70), 42.4 mGycm (24.9), and 1.15 mSv (0.67), whilst for 60-80kg participants the values were 1.0 mGy (0.30), 35.8 mGycm (11.4), and 0.97 mSv (0.31). A statistically significant correlation between CTDIvol and weight was identified for males (r = 0.9123, P < .001) and females (r = 0.9052, P < .001), however, the effect of gender on CTDIvol was not statistically significant (P = .2328) despite notable differences existing at the extremes of the weight range. The additional lifetime attributable cancer risks from a single scan were in the range 0.001%-0.006%.
Conclusions
Low radiation doses can be achieved across a typical lung screening cohort using scan protocols that have been shown to deliver high levels of image quality. The observed dose levels may be considered as typical values for lung screening scans on similar types of scanners for an equivalent participant cohort.
Advances in knowledge
Presentation of typical radiation dose levels for CT lung screening examinations in a large UK trial. Effective radiation doses can be of the order of 1 mSv for standard sized participants. Lifetime attributable cancer risks resulting from a single low-dose CT scan did not exceed 0.006%.
... Nowadays, an iterative reconstruction (IR) method is generally used for CT image reconstruction in order to produce a high-quality image with low dose. The IR results in images with non-uniform noise [21][22][23][24][25]. In addition, many non-linear filters, such as the bilateral filter (BF) [26,27] and the nonlocal mean (NLM) filter [28,29] are implemented in CT image, where filtering is performed aggressively in homogeneous regions with zero gradients, but less aggressively in non-homogeneous regions or edge areas with high gradients [30,31]. ...
In computed tomography (CT), the noise is sometimes non-uniform, i.e. the noise magnitude may vary with the gradient CT number within the image. However, the noise fluctuations due to the magnitude of the image gradient is not considered. The purpose of this study was to quantify the noise non-uniformity in CT images using appropriate 1D and 2D computational gradient phantoms, and to validate the effectiveness of the proposed concept in images filtered by the bilateral filter (BF), as an example of a non-linear filter. We first developed 1D and 2D computational gradient phantoms, and Gaussian noises with several noise levels were then added to the phantoms. In addition, to simulate the real form of noise from images obtained in a real CT scanner, a homogeneous water phantom image was used. These noise levels were referred to as ground truth noise (σG). The phantoms were filtered by the bilateral filter with various pixel value spreads (σ) to produce non-uniform noise. The original gradient phantoms (G) were subtracted from both the noisy phantoms (IN) and the filtered noisy phantoms (IBF), and the magnitudes of the resulting noise for each gradient were computed. The noise-gradient dependency (NGD) curve was used to display the dependency of noise magnitude on image gradient in the non-uniform noise. It was found that for uniform noise, the magnitude of noise was constant for all gradients. However, for non-uniform noise, the measured noise was dependent on the gradient levels and on the strength of the BF for every ground truth noise (σG). It was found that the noise magnitude was large for the large gradients and decreased with the magnitude of the image gradient.
... Recent techniques have set a focus on iterative image reconstruction models such as adaptive statistical iterative reconstruction or model-based iterative reconstruction 16,17,22,[36][37][38] . These algorithms are useful options for further dose reductions, not only in diagnostic CT but also in CT-guided interventions [39][40][41] . ...
This study aimed to systematically evaluate the impact of dose reduction on image quality and confidence for intervention planning and guidance regarding computed tomography (CT)-based intervertebral disc and vertebral body biopsies. We retrospectively analyzed 96 patients who underwent multi-detector CT (MDCT) acquired for the purpose of biopsies, which were either derived from scanning with standard dose (SD) or low dose (LD; using tube current reduction). The SD cases were matched to LD cases considering sex, age, level of biopsy, presence of spinal instrumentation, and body diameter. All images for planning (reconstruction: “IMR1”) and periprocedural guidance (reconstruction: “iDose4”) were evaluated by two readers (R1 and R2) using Likert scales. Image noise was measured using attenuation values of paraspinal muscle tissue. The dose length product (DLP) was statistically significantly lower for LD scans regarding the planning scans (SD: 13.8 ± 8.2 mGy*cm, LD: 8.1 ± 4.4 mGy*cm, p < 0.01) and the interventional guidance scans (SD: 43.0 ± 48.8 mGy*cm, LD: 18.4 ± 7.3 mGy*cm, p < 0.01). Image quality, contrast, determination of the target structure, and confidence for planning or intervention guidance were rated good to perfect for SD and LD scans, showing no statistically significant differences between SD and LD scans (p > 0.05). Image noise was similar between SD and LD scans performed for planning of the interventional procedures (SD: 14.62 ± 2.83 HU vs. LD: 15.45 ± 3.22 HU, p = 0.24). Use of a LD protocol for MDCT-guided biopsies along the spine is a practical alternative, maintaining overall image quality and confidence. Increasing availability of model-based iterative reconstruction in clinical routine may facilitate further radiation dose reductions.
... To generate the signal-present images, each signal object (f s ) was inserted into step (a) of the image generation process. The signal's parameters A s , z, and R, as described in Equation (3), were set as follows: signal amplitude was set to −870 HU (close to the attenuation coefficient of low-attenuation ground-glass opacities (GGO) 50 ), R was fixed at 4 mm, and z ranged from 1 to 4. Note that according to the Astra Toolbox tool, lower Poisson noise values lead to higher noise levels added to a sinogram. 46 ...
Background
The current paradigm for evaluating computed tomography (CT) system performance relies on a task‐based approach. As the Hotelling observer (HO) provides an upper bound of observer performances in specific signal detection tasks, the literature advocates HO use for optimization purposes. However, computing the HO requires calculating the inverse of the image covariance matrix, which is often intractable in medical applications. As an alternative, dimensionality reduction has been extensively investigated to extract the task‐relevant features from the raw images. This can be achieved by using channels, which yields the channelized‐HO (CHO). The channels are only considered efficient when the channelized observer (CO) can approximate its unconstrained counterpart. Previous work has demonstrated that supervised learning‐based methods can usually benefit CO design, either for generating efficient channels using partial least squares (PLS) or for replacing the Hotelling detector with machine‐learning (ML) methods.
Purpose
Here we investigated the efficiency of a supervised ML‐algorithm used to design a CO for predicting the performance of unconstrained HO. The ML‐algorithm was applied either (1) in the estimator for dimensionality reduction, or (2) in the detector function.
Methods
A channelized support vector machine (CSVM) was employed and compared against the CHO in terms of ability to predict HO performances. Both the CSVM and the CHO were estimated with channels derived from the singular value decomposition (SVD) of the system operator, principal component analysis (PCA), and PLS. The huge variety of regularization strategies proposed by CT system vendors for statistical image reconstruction (SIR) make the generalization capability of an observer a key point to consider upfront of implementation in clinical practice. To evaluate the generalization properties of the observers, we adopted a 2‐step testing process: (1) achieved with the same regularization strategy (as in the training phase) and (2) performed using different reconstruction properties. We generated simulated‐ signal‐known‐exactly/background‐known‐exactly (SKE/BKE) tasks in which different noise structures were generated using Markov random field (MRF) regularizations using either a Green or a quadratic, function.
Results
The CSVM outperformed the CHO for all types of channels and regularization strategies. Furthermore, even though both COs generalized well to images reconstructed with the same regularization strategy as the images considered in the training phase, the CHO failed to generalize to images reconstructed differently whereas the CSVM managed to successfully generalize. Lastly, the proposed CSVM observer used with PCA channels outperformed the CHO with PLS channels while using a smaller training data set.
Conclusion
These results argue for introducing the supervised‐learning paradigm in the detector function rather than in the operator of the channels when designing a CO to provide an accurate estimate of HO performance. The CSVM with PCA channels proposed here could be used as a surrogate for HO in image quality assessment.
... The alternative hypothesis follows a noninferiority approach for different pathologic lesions at different radiation dose levels. 8,16,22,23 ...
Objectives:
The purpose of this study was to evaluate the minimum diagnostic radiation dose level for the detection of high-resolution (HR) lung structures, pulmonary nodules (PNs), and infectious diseases (IDs).
Materials and methods:
A preclinical chest computed tomography (CT) trial was performed with a human cadaver without known lung disease with incremental radiation dose using tin filter-based spectral shaping protocols. A subset of protocols for full diagnostic evaluation of HR, PN, and ID structures was translated to clinical routine. Also, a minimum diagnostic radiation dose protocol was defined (MIN). These protocols were prospectively applied over 5 months in the clinical routine under consideration of the individual clinical indication. We compared radiation dose parameters, objective and subjective image quality (IQ).
Results:
The HR protocol was performed in 38 patients (43%), PN in 21 patients (24%), ID in 20 patients (23%), and MIN in 9 patients (10%). Radiation dose differed significantly among HR, PN, and ID (5.4, 1.2, and 0.6 mGy, respectively; P < 0.001). Differences between ID and MIN (0.2 mGy) were not significant (P = 0.262). Dose-normalized contrast-to-noise ratio was comparable among all groups (P = 0.087). Overall IQ was perfect for the HR protocol (median, 5.0) and decreased for PN (4.5), ID-CT (4.3), and MIN-CT (2.5). The delineation of disease-specific findings was high in all dedicated protocols (HR, 5.0; PN, 5.0; ID, 4.5). The MIN protocol had borderline IQ for PN and ID lesions but was insufficient for HR structures. The dose reductions were 78% (PN), 89% (ID), and 97% (MIN) compared with the HR protocols.
Conclusions:
Personalized chest CT tailored to the clinical indications leads to substantial dose reduction without reducing interpretability. More than 50% of patients can benefit from such individual adaptation in a clinical routine setting. Personalized radiation dose adjustments with validated diagnostic IQ are especially preferable for evaluating ID and PN lesions.
... The image reconstruction techniques (IRTs) such as adaptive statistical iterative reconstruction (ASIR) in conjunction with lowering tube current will minimize radiation doses, while maintaining ground glass opacity (GGO) definition and overall image quality [21,22] . Quantitative image analysis using ASIR allows a decrease of about 40% in the tube current and the radiation dose with the same result in image noise magnitude and contrast-tonoise ratio compared with the conventional filtered back projection (FBP) [23][24][25][26] . To date, many types of iterative reconstruction (IR) algorithms have been introduced according to the CT machines or vendors, and widely used in clinical practice. ...
... Nevertheless, it would be possible to further reduce exposure by improving the diagnostic flowchart and excluding low-risk patients to minimize unnecessary radiologic examinations (17). Furthermore, new CT scanners with optimized acquisition protocols that can reduce the dose by up to 40% and new reconstruction algorithms that can reduce the radiation dose by up to 80% and obtain an equivalent image quality are now available (18)(19)(20). ...
Lung cancer is the leading cause of cancer-related death worldwide, and its early detection is critical to achieving a curative treatment and to reducing mortality. Low-dose computed tomography (LDCT) is a highly sensitive technique for detecting noninvasive small lung tumors in high-risk populations. We here analyze the current status of lung cancer screening (LCS) from a European point of view. With economic burden of health care in most European countries resting on the state, it is important to reduce costs of screening and improve its effectiveness. Current cost-effectiveness analyses on LCS have indicated a favorable economic profile. The most recently published analysis reported an incremental cost-effectiveness ratio (ICER) of €3,297 per 1 life-year gained adjusted for the quality of life (QALY) and €2,944 per life-year gained, demonstrating a 90% probability of ICER being below €15,000 and a 98.1% probability of being below €25,000. Different risk models have been used to identify the target population; among these, the PLCOM2012 in particular allows for the selection of the population to be screened with high sensitivity. Risk models should also be employed to define screening intervals, which can reduce the general number of LDCT scans after the baseline round. Future perspectives of screening in a European scenario are related to the will of the policy makers to implement policy on a large scale and to improve the effectiveness of a broad screening of smoking-related disease, including cardiovascular prevention, by measuring coronary calcium score on LDCT. The employment of artificial intelligence (AI) in imaging interpretation, the use of liquid biopsies for the characterization of CT-detected undetermined nodules, and less invasive, personalized surgical treatments, will improve the effectiveness of LCS.
... Our study showed that subsolid nodules can be detected without a decrement in performance at the National Lung Screening Trial dose level (2). However, very-low-dose levels resulted in loss of detection of some part-solid nodules, as demonstrated previously (31,32). Our finding that IR was not helpful at lower-dose levels differs from some previous results (33)(34)(35), but is consistent with that of Vardhanabhuti et al (36) and others (20), and a prior metaanalysis concluding that dose can be reduced to less than 1 mSv when IR is used for a variety of indications including detection of pulmonary nodules (37). ...
Background There is a wide variation in radiation dose levels that can be used with chest CT in order to detect indeterminate pulmonary nodules. Purpose To compare the performance of lower-radiation-dose chest CT with that of routine dose in the detection of indeterminate pulmonary nodules 5 mm or greater. Materials and Methods In this retrospective study, CT projection data from 83 routine-dose chest CT examinations performed in 83 patients (120 kV, 70 quality reference mAs [QRM]) were collected between November 2013 and April 2014. Reference indeterminate pulmonary nodules were identified by two nonreader thoracic radiologists. By using validated noise insertion, five lower-dose data sets were reconstructed with filtered back projection (FBP) or iterative reconstruction (IR; 30 QRM with FBP, 10 QRM with IR, 5 QRM with FBP, 5 QRM with IR, and 2.5 QRM with IR). Three thoracic radiologists circled pulmonary nodules, rating confidence that the nodule was a 5-mm-or-greater indeterminate pulmonary nodule, and graded image quality. Analysis was performed on a per-nodule basis by using jackknife alternative free-response receiver operating characteristic figure of merit (FOM) and noninferiority limit of -0.10. Results There were 66 indeterminate pulmonary nodules (mean size, 8.6 mm ± 3.4 [standard deviation]; 21 part-solid nodules) in 42 patients (mean age, 51 years ± 17; 21 men and 21 women). Compared with the FOM for routine-dose CT (size-specific dose estimate, 6.5 mGy ± 1.8; FOM, 0.86 [95% confidence interval: 0.80, 0.91]), FOM was noninferior for all lower-dose configurations except for 2.5 QRM with IR. The sensitivity for subsolid nodules at 70 QRM was 60% (range, 48%-72%) and was significantly worse at a dose of 5 QRM and lower, whether or not IR was used (P < .05). Diagnostic image quality decreased with decreasing dose (P < .001) and was better with IR at 5 QRM (P < .05). Conclusion CT images reconstructed at dose levels down to 10 quality reference mAs (size-specific dose estimate, 0.9 mGy) had noninferior performance compared with routine dose in depicting pulmonary nodules. Iterative reconstruction improved subjective image quality but not performance at low dose levels. © RSNA, 2020 Online supplemental material is available for this article. See also the editorial by White and Kazerooni in this issue.
... According to another experiment without any artificial objects, model-based iterative reconstruction (MBIR) was able to detect GGN of -630 HU at 20% of the standard dose, and -800 HU at 40% without any significant difference. [14] Our data support those previous results and also found that FIRST is useful even under a pacemaker at low dose. Meanwhile, SEMAR turned out to be very useful for detecting GGNs at low dose in our results. ...
The aim was to compare the effects of metal artifacts from a pacemaker on pulmonary nodule detection among computed tomography (CT) images reconstructed using filtered back projection (FBP), single-energy metal artifact reduction (SEMAR), and forward-projected model-based iterative reconstruction solution (FIRST).
Nine simulated nodules were placed inside a chest phantom with a pacemaker. CT images reconstructed using FBP, SEMAR, and FIRST were acquired at low and standard dose, and were evaluated by 2 independent radiologists.
FIRST demonstrated the most significantly improved metal artifact and nodule detection on low dose CT (P < .0032), except at 10 mA and 5-mm thickness. At standard-dose CT, SEMAR showed the most significant metal artifact reduction (P < .00001). In terms of nodule detection, no significant differences were observed between FIRST and SEMAR (P = .161).
With a pacemaker present, FIRST showed the best nodule detection ability at low-dose CT and SEMAR is comparable to FIRST at standard dose CT.
