Deep-Learning Models for Abdominal CT Organ Segmentation in Children: Development and Validation in Internal and Heterogeneous Public Datasets.

Background: Deep-learning abdominal organ segmentation algorithms have shown excellent results in adults; validation in children is sparse. Objective: To develop and validate deep-learning models for liver, spleen, and pancreas segmentation on pediatric CT examinations. Methods: This retrospective study developed and validated deep-learning models for liver, spleen, and pancreas segmentation using 1731 CT examinations (1504 training, 221 testing), derived from three internal institutional pediatric (age ≤18) datasets (n=483) and three public datasets comprising pediatric and adult examinations with various pathologies (n=1248). Three deep-learning model architectures (SegResNet, DynUNet, and SwinUNETR) from the Medical Open Network for AI (MONAI) framework underwent training using native training (NT), relying solely on institutional datasets, and transfer learning (TL), incorporating pre-training on public datasets. For comparison, TotalSegmentator (TS), a publicly available segmentation model, was applied to test data without further training. Segmentation performance was evaluated using mean Dice similarity coefficient (DSC), with manual segmentations as reference. Results: For internal pediatric data, DSC for normal liver was 0.953 (TS), 0.964-0.965 (NT models), and 0.965-0.966 (TL models); normal spleen, 0.914 (TS), 0.942-0.945 (NT models), and 0.937-0.945 (TL models); normal pancreas, 0.733 (TS), 0.774-0.785 (NT models), and 0.775-0.786 (TL models); pancreas with pancreatitis, 0.703 (TS), 0.590-0.640 (NT models), and 0.667-0.711 (TL models). For public pediatric data, DSC for liver was 0.952 (TS), 0.876-0.908 (NT models), and 0.941-0.946 (TL models); spleen, 0.905 (TS), 0.771-0.827 (NT models), and 0.897-0.926 (TL models); pancreas, 0.700 (TS), 0.577-0.648 (NT models), and 0.693-0.736 (TL models). For public primarily adult data, DSC for liver was 0.991 (TS), 0.633-0.750 (NT models), and 0.926-0.952 (TL models); spleen, 0.983 (TS), 0.569-0.604 (NT models), and 0.923-0.947 (TL models); pancreas, 0.909 (TS), 0.148-0.241 (NT models), and 0.699-0.775 (TL models). DynUNet-TL was selected as the best-performing NT or TL model and was made available as an opensource MONAI bundle (https://github.com/cchmc-dll/pediatric_abdominal_segmentation_bundle.git). Conclusion: TL models trained on heterogeneous public datasets and fine-tuned using institutional pediatric data outperformed internal NT models and TotalSegmentator across internal and external pediatric test data. Segmentation performance was better in liver and spleen than in pancreas. Clinical Impact: The selected model may be used for various volumetry applications in pediatric imaging.

Deep Learning for Automated Measurement of Total Cardiac Volume for Heart Transplantation Size Matching.

Total Cardiac Volume (TCV)-based size matching using Computed Tomography (CT) is a novel technique to compare donor and recipient heart size in pediatric heart transplant that may increase overall utilization of available grafts. TCV requires manual segmentation, which limits its widespread use due to time and specialized software and training needed for segmentation. This study aims to determine the accuracy of a Deep Learning (DL) approach using 3-dimensional Convolutional Neural Networks (3D-CNN) to calculate TCV, with the clinical aim of enabling fast and accurate TCV use at all transplant centers. Ground truth TCV was segmented on CT scans of subjects aged 0-30 years, identified retrospectively. Ground truth segmentation masks were used to train and test a custom 3D-CNN model consisting of a DenseNet architecture in combination with residual blocks of ResNet architecture. The model was trained on a cohort of 270 subjects and a validation cohort of 44 subjects (36 normal, 8 heart disease retained for model testing). The average Dice similarity coefficient of the validation cohort was 0.94 ± 0.03 (range 0.84-0.97). The mean absolute percent error of TCV estimation was 5.5%. There is no significant association between model accuracy and subject age, weight, or height. DL-TCV was on average more accurate for normal hearts than those listed for transplant (mean absolute percent error 4.5 ± 3.9 vs. 10.5 ± 8.5, p = 0.08). A deep learning-based 3D-CNN model can provide accurate automatic measurement of TCV from CT images. This initial study is limited as a single-center study, though future multicenter studies may enable generalizable and more accurate TCV measurement by inclusion of more diverse cardiac pathology and increasing the training data.

Computed tomography-based measurements of normative liver and spleen volumes in children.

BACKGROUND: Quantification of organ size has utility in clinical care and research for diagnostics, prognostics and surgical planning. Volumetry is regarded as the best measure of organ size and change in size over time. Scarce reference values exist for liver and spleen volumes in healthy children. OBJECTIVE: To report liver and spleen volumes for a sample of children defined by manual segmentation of contrast-enhanced CT images with the goal of defining normal values and thresholds that might indicate disease. MATERIALS AND METHODS: This retrospective study included clinically acquired contrast-enhanced CTs of the abdomen/pelvis for children and adolescents imaged between January 2018 and July 2021. Liver and spleen volumes were derived through manual segmentation of CTs reconstructed at 2.5-, 3- or 5-mm slice thickness. A subset of images (5%, n=16) was also segmented using 0.5-mm slice thickness reconstructions to define agreement based on image slice thickness. We used Pearson correlation and multivariable regression to assess associations between organ volumes and patient characteristics. We generated reference intervals for the 5th, 25th, 50th (median), 75th and 95th percentiles for organ volumes as a function of age and weight using quantile regression models. Finally, we calculated Bland-Altman plots and intraclass correlation coefficients (ICC) to quantify agreement. RESULTS: We included a total of 320 children (mean age ± standard deviation [SD] = 9±4.6 years; mean weight 38.1±18.8 kg; 160 female). Liver volume ranged from 340-2,002 mL, and spleen volume ranged from 28-480 mL. Patient weight (kg) (ß=12.5), age (months) (ß=1.7) and sex (female) (ß = -35.3) were independent predictors of liver volume, whereas patient weight (kg) (ß=2.4) and age (months) (ß=0.3) were independent predictors of spleen volume. There was excellent absolute agreement (ICC=0.99) and minimal absolute difference (4 mL) in organ volumes based on reconstructed slice thickness. CONCLUSION: We report reference liver and spleen volumes for children without liver or spleen disease. These results provide reference ranges and potential thresholds to identify liver and spleen size abnormalities that might reflect disease in children.

Deep Learning for Automated Measurement of Total Cardiac Volume for Heart Transplantation Size Matching.

Background: Total Cardiac Volume (TCV) based size matching using Computed Tomography (CT) is a novel technique to compare donor and recipient heart size in pediatric heart transplant that may increase overall utilization of available grafts. TCV requires manual segmentation, which limits its widespread use due to time and specialized software and training needed for segmentation. Objective: This study aims to determine the accuracy of a Deep Learning (DL) approach using 3-dimensional Convolutional Neural Networks (3D-CNN) to calculate TCV, with the clinical aim of enabling fast and accurate TCV use at all transplant centers. Materials and Methods: Ground truth TCV was segmented on CT scans of subjects aged 0-30 years, identified retrospectively. Ground truth segmentation masks were used to train and test a custom 3D-CNN model consisting of a Dense-Net architecture in combination with residual blocks of ResNet architecture. Results: The model was trained on a cohort of 270 subjects and a validation cohort of 44 subjects (36 normal, 8 heart disease retained for model testing). The average Dice similarity coefficient of the validation cohort was 0.94 ± 0.03 (range 0.84-0.97). The mean absolute percent error of TCV estimation was 5.5%. There is no significant association between model accuracy and subject age, weight, or height. DL-TCV was on average more accurate for normal hearts than those listed for transplant (mean absolute percent error 4.5 ± 3.9 vs. 10.5 ± 8.5, p = 0.08). Conclusion: A deep learning based 3D-CNN model can provide accurate automatic measurement of TCV from CT images.

Reducing Pediatric Intraoral Radiography Radiation Dose Using Reduced-Power Dental X-Ray Units: A Randomized Trial.

Purpose: To assess the diagnostic confidence of intraoral radiographic image quality while reducing the pediatric patient's radiation exposure using a longer position indicating device (PID), additional X-ray beam filtration and rectangular collimation while using modern, lower-power intraoral dental X-ray units.
Methods: A randomized prospective study scored bitewing intraoral dental images based on relevant clinical features. Observer studies with pediatric dentists and dental residents were conducted to verify whether diagnostic confidence remained unchanged after dose reduction modifications. The study involved a two-phase investigation to determine: (1) the best thickness of aluminum (Al) 2024-T3 alloy filter and (2) required increased exposure time to maintain intraoral radiographic image quality. A 30 cm PID with a rectangular collimator was used to further manage patient dose. For each phase, images from 125 patients were collected from February 2017 to September 2018 and analyzed.
Results: The results from the observer study using a 30 cm PID, 1.02 mm thick Al alloy filter, and a rectangular collimator resulted in a patient dose reduction between 64 percent (exposure time of 400 msec) to 77 percent (250 msec), without any statis- tically significant effect to the diagnostic confidence of the observers in evaluating the reduced radiation images.
Conclusion: Long recognized dose reduction methods, when implemented on a modern, low-power intraoral dental X-ray unit, do not impact confidence in bite- wing diagnostic images, but substantially reduce patient dose and should be adopted to increase patient safety, especially for children.

Application of reference air kerma alert levels for pediatric fluoroscopic examinations.

The purpose of this study was to provide an empirical model to develop reference air kerma (RAK) alert levels as a function of patient thickness or age for pediatric fluoroscopy for any institution to use in a Quality Assurance program. RAK and patient thickness were collected for 10&663 general fluoroscopic examinations and 1500 fluoroscopically guided interventions (FGIs). RAK and patient age were collected for 6137 fluoroscopic examinations with mobile-C-arms (MC). Coefficients of linear regression fits of logarithmic RAK as a function of patient thickness or age were generated for each fluoroscopy group. Regression fits of RAK for 50%, 90%, and 98% upper prediction levels were used as inputs to derive an empirical formula to estimate alert levels as a function of patient thickness. A methodology is presented to scale results from this study for any patient thickness or age for any institution, for example, the patient thickness dependent RAK alert level at the top 1% of expected RAK can be set using the 98% upper prediction interval boundary given by: RAK 98 % = e m . x avg + s 98 . c ̂ ${\rm{RAK}}_{98\% } = {e}^{m.{x}_{{\rm{avg}}} + {s}_{98}.\hat{c}}\ $ , where xavg is the institute's average patient thickness or age, and c ̂ $\hat{c}$ is the intercept based on the average RAK of the patient population calculated as c ̂ = ln ( RAK avg ) - m . x avg . RA K avg $\hat{c} = \ln ( {{\rm{RAK}}_{{\rm{avg}}}} )\ - m.{x}_{{\rm{avg}}}{\rm{.RA}}{{\rm{K}}}_{{\rm{avg}}}$ is the institution's average RAK (mGy). m and s98 are constants presented for each type of fluoroscope and RAK group and represent slope of the fit and scale factor, respectively. An empirical equation, which estimates alert levels expressed as air Kerma without backscatter at the interventional reference point as a function of patient thickness or age is provided for each fluoroscopic examination type. The empirical equations allow any facility with limited data to scale the results of this study's single facility data to model their practice's unique RAK alert levels and patient population demographics to establish pediatric alert levels for fluoroscopic procedures.

Defining Normal Ranges of Skeletal Muscle Area and Skeletal Muscle Index in Children on CT Using an Automated Deep Learning Pipeline: Implications for Sarcopenia Diagnosis.

BACKGROUND. Skeletal muscle area (SMA), representing skeletal muscle cross-sectional area at the L3 vertebral level, and skeletal muscle index (SMI), representing height-normalized SMA, can serve as markers of sarcopenia. Normal SMA and SMI values have been reported primarily in adults. OBJECTIVE. The purpose of this study was to use an automated deep learning (DL) pipeline for muscle segmentation on abdominal CT to define normative age- and sex-based values for pediatric muscle cross-sectional area as a guide for diagnosis of sarcopenia in children. METHODS. This retrospective study reviewed records of patients who underwent abdominal CT at Cincinnati Children's Hospital Medical Center from January 1, 2009, to January 3, 2019. Patients were excluded on the basis of age outside of the eligible range (2.00-18.99 years), body mass index (BMI) outside of 5-95% age-based percentiles using CDC and WHO growth charts, known medical condition, medication use, support devices, surgery, or missing axial images at the L3 level. A previously validated automated DL pipeline was used to identify an axial slice at L3 and segment skeletal muscle to generate SMA and SMI. Pearson correlation coefficients were computed. Quantile regression analysis was used to plot SMA and SMI as functions of age and sex and to determine age- and sex-based percentile values. RESULTS. Of 8817 patients who underwent abdominal CT during the study period, 2168 (mean age, 12.3 ± 4.3 [SD] years; 1125 female patients, 1043 male patients) met inclusion criteria. Mean BMI-for-age percentile based on CDC and WHO growth charts was 64.8% ± 25.3% for female patients and 61.4% ± 25.8% for male patients. SMA showed strong correlation with weight, height, age, and BMI for male (0.79-0.94) and female (0.75-0.90) patients; SMI showed weak-to-moderate correlation with weight, height, age, and BMI for male (0.25-0.57) and female (0-0.43) patients. Normal SMA and SMI ranges for age and sex were expressed as curves and as a lookup table, identifying 54 male and 59 female patients with muscle measurements below the 5th percentile regression curve. CONCLUSION. By using an automated DL pipeline in a large sample of carefully selected children, normal ranges for SMA and SMI were calculated as functions of age and sex. CLINICAL IMPACT. The normative values should aid the diagnosis of sarcopenia in children.

Detection of urinary tract calculi on CT images reconstructed with deep learning algorithms.

BACKGROUND: Deep learning Computed Tomography (CT) reconstruction (DLR) algorithms promise to improve image quality but the impact on clinical diagnostic performance remains to be demonstrated. We aimed to compare DLR to standard iterative reconstruction for detection of urolithiasis by unenhanced CT in children and young adults. METHODS: This was an IRB approved retrospective study involving post-hoc reconstruction of clinically acquired unenhanced abdomen/pelvis CT scans. Images were reconstructed with six different manufacturer-standard DLR algorithms and reformatted in 3 planes (axial, sagittal, and coronal) at 3 mm intervals. De-identified reconstructions were loaded as independent examinations for review by 3 blinded radiologists (R1, R2, R3) tasked with identifying and measuring all stones. Results were compared to the clinical iterative reconstruction images as a reference standard. IntraClass correlation coefficients and kappa (k) statistics were used to quantify agreement. RESULTS: CT data for 14 patients (mean age: 17.3 ± 3.4 years, 5 males and 9 females, weight class: 31-70 kg (n = 6), 71-100 kg (n = 7), > 100 kg (n = 1)) were reconstructed into 84 total exams. 7 patients had urinary tract calculi. Interobserver agreement on the presence of any urinary tract calculus was substantial to almost perfect (k = 0.71-1) for all DLR algorithms. Agreement with the reference standard on number of calculi was excellent (ICC = 0.78-0.96) and agreement on the size of the largest calculus was fair to excellent (ICC = 0.51-0.97) depending on reviewer and DLR algorithm. CONCLUSION: Deep learning reconstruction of unenhanced CT images allows similar renal stone detectability compared to iterative reconstruction.

Automated Segmentation of Abdominal Skeletal Muscle on Pediatric CT Scans Using Deep Learning.

PURPOSE: To automate skeletal muscle segmentation in a pediatric population using convolutional neural networks that identify and segment the L3 level at CT. MATERIALS AND METHODS: In this retrospective study, two sets of U-Net-based models were developed to identify the L3 level in the sagittal plane and segment the skeletal muscle from the corresponding axial image. For model development, 370 patients (sampled uniformly across age group from 0 to 18 years and including both sexes) were selected between January 2009 and January 2019, and ground truth L3 location and skeletal muscle segmentation were manually defined. Twenty percent (74 of 370) of the examinations were reserved for testing the L3 locator and muscle segmentation, while the remaining were used for training. For the L3 locator models, maximum intensity projections (MIPs) from a fixed number of central sections of sagittal reformats (either 12 or 18 sections) were used as input with or without transfer learning using an L3 localizer trained on an external dataset (four models total). For the skeletal muscle segmentation models, two loss functions (weighted Dice similarity coefficient [DSC] and binary cross-entropy) were used on models trained with or without data augmentation (four models total). Outputs from each model were compared with ground truth, and the mean relative error and DSC from each of the models were compared with one another. RESULTS: L3 section detection trained with an 18-section MIP model with transfer learning had a mean error of 3.23 mm ± 2.61 standard deviation, which was within the reconstructed image thickness (3 or 5 mm). Skeletal muscle segmentation trained with the weighted DSC loss model without data augmentation had a mean DSC of 0.93 ± 0.03 and mean relative error of 0.04 ± 0.04. CONCLUSION: Convolutional neural network models accurately identified the L3 level and segmented the skeletal muscle on pediatric CT scans.Supplemental material is available for this article.See also the commentary by Cadrin-Chênevert in this issue.© RSNA, 2021.

DeepLiverNet: a deep transfer learning model for classifying liver stiffness using clinical and T2-weighted magnetic resonance imaging data in children and young adults.

BACKGROUND: Although MR elastography allows for quantitative evaluation of liver stiffness to assess chronic liver diseases, it has associated drawbacks related to additional scanning time, patient discomfort, and added costs. OBJECTIVE: To develop a machine learning model that can categorically classify the severity of liver stiffness using both anatomical T2-weighted MRI and clinical data for children and young adults with known or suspected pediatric chronic liver diseases. MATERIALS AND METHODS: We included 273 subjects with known or suspected chronic liver disease. We extracted data including axial T2-weighted fast spin-echo fat-suppressed images, clinical data (e.g., demographic/anthropomorphic data, particular medical diagnoses, laboratory values) and MR elastography liver stiffness measurements. We propose DeepLiverNet (a deep transfer learning model) to classify patients into one of two groups: no/mild liver stiffening (<3 kPa) or moderate/severe liver stiffening (≥3 kPa). We conducted internal cross-validation using 178 subjects, and external validation using an independent cohort of 95 subjects. We assessed diagnostic performance using accuracy, sensitivity, specificity and area under the receiver operating characteristic curve (AuROC). RESULTS: In the internal cross-validation experiment, the combination of clinical and imaging data produced the best performance (AuROC=0.86) compared to clinical (AuROC=0.83) or imaging (AuROC=0.80) data alone. Using both clinical and imaging data, the DeepLiverNet correctly classified patients with accuracy of 88.0%, sensitivity of 74.3% and specificity of 94.6%. In our external validation experiment, this same deep learning model achieved an accuracy of 80.0%, sensitivity of 61.1%, specificity of 91.5% and AuROC of 0.79. CONCLUSION: A deep learning model that incorporates clinical data and anatomical T2-weighted MR images might provide a means of risk-stratifying liver stiffness and directing the use of MR elastography.

Improving Image Quality and Reducing Radiation Dose for Pediatric CT by Using Deep Learning Reconstruction.

Background CT deep learning reconstruction (DLR) algorithms have been developed to remove image noise. How the DLR affects image quality and radiation dose reduction has yet to be fully investigated. Purpose To investigate a DLR algorithm's dose reduction and image quality improvement for pediatric CT. Materials and Methods DLR was compared with filtered back projection (FBP), statistical-based iterative reconstruction (SBIR), and model-based iterative reconstruction (MBIR) in a retrospective study by using data from CT examinations of pediatric patients (February to December 2018). A comparison of object detectability for 15 objects (diameter, 0.5-10 mm) at four contrast difference levels (50, 150, 250, and 350 HU) was performed by using a non-prewhitening-matched mathematical observer model with eye filter (d'NPWE), task transfer function, and noise power spectrum analysis. Object detectability was assessed by using area under the curve analysis. Three pediatric radiologists performed an observer study to assess anatomic structures with low object-to-background signal and contrast to noise in the azygos vein, right hepatic vein, common bile duct, and superior mesenteric artery. Observers rated from 1 to 10 (worst to best) for edge definition, quantum noise level, and object conspicuity. Analysis of variance and Tukey honest significant difference post hoc tests were used to analyze differences between reconstruction algorithms. Results Images from 19 patients (mean age, 11 years ± 5 [standard deviation]; 10 female patients) were evaluated. Compared with FBP, SBIR, and MBIR, DLR demonstrated improved object detectability by 51% (16.5 of 10.9), 18% (16.5 of 13.9), and 11% (16.5 of 14.8), respectively. DLR reduced image noise without noise texture effects seen with MBIR. Radiologist ratings were 7 ± 1 (DLR), 6.2 ± 1 (MBIR), 6.2 ± 1 (SBIR), and 4.6 ± 1 (FBP); two-way analysis of variance showed a difference on the basis of reconstruction type (P < .001). Radiologists consistently preferred DLR images (intraclass correlation coefficient, 0.89; 95% CI: 0.83, 0.93). DLR demonstrated 52% (1 of 2.1) greater dose reduction than SBIR. Conclusion The DLR algorithm improved image quality and dose reduction without sacrificing noise texture and spatial resolution. © RSNA, 2020 Online supplemental material is available for this article.

Achievable dose reductions with gonadal shielding for children and adults during abdominal/pelvic radiographic examinations: A Monte Carlo simulation.

PURPOSE: Recently, medical professionals have reconsidered the practice of routine gonadal shielding for radiographic examinations. The objective of this study was to evaluate the gonadal dose reduction achievable with gonadal shields in the primary beam during abdominal/pelvic radiographic examinations under ideal and non-ideal shielding placement. METHODS: CT scans of CIRS anthropomorphic phantoms were used to perform voxelized Monte Carlo simulations of the photon transport during abdominal/pelvic radiographic examinations with standard filtration and 0.1 mm Cu + 1 mm Al added filtration to estimate gonadal doses for an adult, 5 yr old, and newborn phantom with and without gonadal shields. The reduction in dose when the shields were not placed at the ideal locations was also evaluated. The ratio of the number of scattered-to-primary photons (SPR) across the anteroposterior (AP) dimension of the phantoms was also reported. RESULTS: The simulated dose reduction with ideal shielding placement for the testes and ovaries ranged from 80% to 90% and 55% to 70% respectively. For children, a misalignment of the shield to the gonad of 4 cm reduced the measured dose reduction to the gonads to <10%. For adults, this effect did not occur until the misalignment increased to ~6 cm. Effects of dose reduction with and without the gonadal shields properly placed were similar for standard filtration and added filtration. SPR at the level of the testes was consistently <1 for all phantoms. SPR for ovaries was ~1.5 for the adult and 5-yr old, and ~1 for the newborn phantom. CONCLUSION: Dose reduction with ideal alignment of the simulated gonadal shield to the gonads in this study was greater for the testes than the ovaries; both reductions were substantial. However, the dose reductions were greatly reduced (to <10%) for both sexes with misalignment of the gonads to the shields by 4 cm for children and 6 cm for adults.

Automatic Detection of Inadequate Pediatric Lateral Neck Radiographs of the Airway and Soft Tissues using Deep Learning.

PURPOSE: To develop and validate a deep learning (DL) algorithm to identify poor-quality lateral airway radiographs. MATERIALS AND METHODS: A total of 1200 lateral airway radiographs obtained in emergency department patients between January 1, 2000, and July 1, 2019, were retrospectively queried from the picture archiving and communication system. Two radiologists classified each radiograph as adequate or inadequate. Disagreements were adjudicated by a third radiologist. The radiographs were used to train and test the DL classifiers. Three technologists and three different radiologists classified the images in the test dataset, and their performance was compared with that of the DL classifiers. RESULTS: The training set had 961 radiographs and the test set had 239. The best DL classifier (ResNet-50) achieved sensitivity, specificity, and area under the receiver operating characteristic curve of 0.90 (95% confidence interval [CI]: 0.86, 0.94), 0.82 (95% CI: 0.76, 0.90), and 0.86 (95% CI: 0.81, 0.91), respectively. Interrater agreement for technologists was fair (Fleiss κ, 0.36 [95% CI: 0.29, 0.43]), while that for radiologists was moderate (Fleiss κ, 0.59 [95% CI: 0.52, 0.66]). Cohen κ value comparing the consensus rating of ResNet-50 iterations from fivefold cross-validation, consensus technologists' rating, and consensus radiologists' rating to the ground truth were 0.76 (95% CI: 0.63, 0.89), 0.49 (95% CI: 0.37, 0.61), and 0.66 (95% CI: 0.54, 0.78), respectively. CONCLUSION: The development and validation of DL classifiers to distinguish between adequate and inadequate lateral airway radiographs is reported. The classifiers performed significantly better than a group of technologists and as well as the radiologists.© RSNA, 2020.

Quantification of skeletal muscle mass: sarcopenia as a marker of overall health in children and adults.

Sarcopenia is defined as the loss of muscle mass or function and has been associated with increased morbidity and mortality in a variety of diseased populations. Sarcopenia results from a higher rate of muscle protein degradation compared to protein synthesis and is an important marker of metabolic status related to nutrition and physical activity. The diagnosis of sarcopenia is accomplished by clinical assessment demonstrating decreased muscle function and radiographic confirmation of decreased muscle mass, via dual X-ray absorptiometry, bioelectric impedance or cross-sectional imaging with CT or MRI. However, normative data for skeletal muscle mass are lacking, especially for children and young adults. Additionally, studies of skeletal muscle mass by cross-sectional imaging in children are scarce. Here, we review the concept of sarcopenia with an emphasis on its relevance in the pediatric population.

Machine Learning Prediction of Liver Stiffness Using Clinical and T2-Weighted MRI Radiomic Data.

OBJECTIVE. The purpose of this study is to develop a machine learning model to categorically classify MR elastography (MRE)-derived liver stiffness using clinical and nonelastographic MRI radiomic features in pediatric and young adult patients with known or suspected liver disease. MATERIALS AND METHODS. Clinical data (27 demographic, anthropomorphic, medical history, and laboratory features), MRI presence of liver fat and chemical shift-encoded fat fraction, and MRE mean liver stiffness measurements were retrieved from electronic medical records. MRI radiomic data (105 features) were extracted from T2-weighted fast spin-echo images. Patients were categorized by mean liver stiffness (< 3 vs ≥ 3 kPa). Support vector machine (SVM) models were used to perform two-class classification using clinical features, radiomic features, and both clinical and radiomic features. Our proposed model was internally evaluated in 225 patients (mean age, 14.1 years) and externally evaluated in an independent cohort of 84 patients (mean age, 13.7 years). Diagnostic performance was assessed using ROC AUC values. RESULTS. In our internal cross-validation model, the combination of clinical and radiomic features produced the best performance (AUC = 0.84), compared with clinical (AUC = 0.77) or radiomic (AUC = 0.70) features alone. Using both clinical and radiomic features, the SVM model was able to correctly classify patients with accuracy of 81.8%, sensitivity of 72.2%, and specificity of 87.0%. In our external validation experiment, this SVM model achieved an accuracy of 75.0%, sensitivity of 63.6%, specificity of 82.4%, and AUC of 0.80. CONCLUSION. An SVM learning model incorporating clinical and T2-weighted radiomic features has fair-to-good diagnostic performance for categorically classifying liver stiffness.

Development and validation of an open source Monte Carlo dosimetry model for wide-beam CT scanners using Fluka.

PURPOSE: Development and validation of an open source Fluka-based Monte Carlo source model for diagnostic patient dose calculations. METHODS: A framework to simulate a computed tomography (CT) scanner using Fluka Monte Carlo particle transport code was developed. The General Electric (GE) Revolution scanner with the large body filter and 120 kV tube potential was characterized using measurements. The model was validated on benchmark CT test problems and on dose measurements in computed tomography dose index (CTDI) and anthropomorphic phantoms. Axial and helical operation modes with provision for tube current modulation (TCM) were implemented. The particle simulations in Fluka were accelerated by executing them on a high-performance computing cluster. RESULTS: The simulation results agreed to better than an average of 4% of the reference simulation results from the AAPM Report 195 test scenarios, namely: better than 2% for both test problems in case 4 using the PMMA phantom, and better than 5% of the reference result for 14 of 17 organs in case 5, and within 10% for the three remaining organs. The Fluka simulation results agreed to better than 2% of the air kerma measured in-air at isocenter of the GE Revolution scanner. The simulated air kerma in the center of the CTDI phantom overestimated the measurement by 7.5% and a correction factor was introduced to account for this. The simulated mean absorbed doses for a chest scan of the pediatric anthropomorphic phantom was completed in ~9 min and agreed to within the 95% CI for bone, soft tissue, and lung measurements made using MOSFET detectors for fixed current axial and helical scans as well as helical scan with TCM. CONCLUSION: A Fluka-based Monte Carlo simulation model of axial and helical acquisition techniques using a wide-beam collimation CT scanner demonstrated good agreement between measured and simulated results for both fixed current and TCM in complex and simple geometries. Code and dataset will be made available at https://github.com/chezhia/FLUKA_CT.

Radiation Dose for Pediatric CT: Comparison of Pediatric versus Adult Imaging Facilities.

Background The American College of Radiology Dose Index Registry for CT enables evaluation of radiation dose as a function of patient characteristics and examination type. The hypothesis of this study was that academic pediatric CT facilities have optimized CT protocols that may result in a lower and less variable radiation dose in children. Materials and Methods A retrospective study of doses (mean patient age, 12 years; age range, 0-21 years) was performed by using data from the National Radiology Data Registry (year range, 2016-2017) (n = 239 622). Three examination types were evaluated: brain without contrast enhancement, chest without contrast enhancement, and abdomen-pelvis with intravenous contrast enhancement. Three dose indexes-volume CT dose index (CTDIvol), size-specific dose estimate (SSDE), and dose-length product (DLP)-were analyzed by using six different size groups. The unequal variance t test and the F test were used to compare mean dose and variances, respectively, at academic pediatric facilities with those at other facility types for each size category. The Bonferroni-Holm correction factor was applied to account for the multiple comparisons. Results Pediatric radiation dose in academic pediatric facilities was significantly lower, with smaller variance for all brain, 42 of 54 (78%) chest, and 48 of 54 (89%) abdomen-pelvis examinations across all six size groups, three dose descriptors, and when compared with that at the other three facilities. For example, abdomen-pelvis SSDE for the 14.5-18-cm size group was 3.6, 5.4, 5.5, and 8.3 mGy, respectively, for academic pediatric, nonacademic pediatric, academic adult, and nonacademic adult facilities (SSDE mean and variance P < .001). Mean SSDE for the smallest patients in nonacademic adult facilities was 51% (6.1 vs 11.9 mGy) of the facility's adult dose. Conclusion Academic pediatric facilities use lower CT radiation dose with less variation than do nonacademic pediatric or adult facilities for all brain examinations and for the majority of chest and abdomen-pelvis examinations. © RSNA, 2019 See also the editorial by Strouse in this issue.

Fully automated tissue classifier for contrast-enhanced CT scans of adult and pediatric patients.

To develop a consistent, fully-automated classifier for all tissues within the trunk and to more accurately discriminate between tissues (such as bone) and contrast medium with overlapping high CT numbers. Twenty-eight contrast enhanced NCAP (neck-chest-abdomen-pelvis) CT scans (four adult and three pediatric patients) were used to train and test a tissue classification pipeline. The classifier output consisted of six tissue classes: lung, fat, muscle, solid organ, blood/bowel contrast and bone. The input features for training were selected from 28 2D image filters and 12 3D filters, and one hand crafted spatial feature. To improve differentiation between tissue and blood/bowel contrast classification, 70 additional CT images were manually classified. Two different training data sets consisting of manually classified tissues from different locations in body were used to train the models. Training used the random forest algorithm in WEKA (Waikato Environment for Knowledge Analysis); the number of trees was optimized for best out-of-bag error. Automated classification accuracy was compared with manual classification by calculating dice similarity coefficient (DSC). Model performance was tested on 21 manually classified slices (two adult and one pediatric patient). The overall DSC at image locations represented in the training dataset were-lung: 0.98, fat: 0.90, muscle: 0.85, solid organ: 0.75, blood/contrast: 0.82, and bone: 0.90. The overall DSC for slice locations that were not represented in the training dataset were-lung: 0.97, fat: 0.89, muscle: 0.76, solid organ: 0.79, blood: 0.56, and bone: 0.74. Analyzing the classification maps for the entire scan volume revealed that except for misclassifications in the trabecular bone region of the spinal column, and solid organ and blood/contrast interfaces within the abdomen, the results were acceptable. A fully-automated whole-body tissue classifier for adult and pediatric contrast-enhanced CT using random forest algorithm and intensity-based image filters was developed.