Supervised contrastive learning enhances graph convolutional networks for predicting neurodevelopmental deficits in very preterm infants using brain structural connectome.

Very preterm (VPT) infants (born at less than 32 weeks gestational age) are at high risk for various adverse neurodevelopmental deficits. Unfortunately, most of these deficits cannot be accurately diagnosed until the age of 2-5 years old. Given the benefits of early interventions, accurate diagnosis and prediction soon after birth are urgently needed for VPT infants. Previous studies have applied deep learning models to learn the brain structural connectome (SC) to predict neurodevelopmental deficits in the preterm population. However, none of these models are specifically designed for graph-structured data, and thus may potentially miss certain topological information conveyed in the brain SC. In this study, we aim to develop deep learning models to learn the SC acquired at term-equivalent age for early prediction of neurodevelopmental deficits at 2 years corrected age in VPT infants. We directly treated the brain SC as a graph, and applied graph convolutional network (GCN) models to capture complex topological information of the SC. In addition, we applied the supervised contrastive learning (SCL) technique to mitigate the effects of the data scarcity problem, and enable robust training of GCN models. We hypothesize that SCL will enhance GCN models for early prediction of neurodevelopmental deficits in VPT infants using the SC. We used a regional prospective cohort of ∼280 VPT infants who underwent MRI examinations at term-equivalent age from the Cincinnati Infant Neurodevelopment Early Prediction Study (CINEPS). These VPT infants completed neurodevelopmental assessment at 2 years corrected age to evaluate cognition, language, and motor skills. Using the SCL technique, the GCN model achieved mean areas under the receiver operating characteristic curve (AUCs) in the range of 0.72∼0.75 for predicting three neurodevelopmental deficits, outperforming several competing models. Our results support our hypothesis that the SCL technique is able to enhance the GCN model in our prediction tasks.

Precision and Test-Retest Repeatability of Stiffness Measurement with MR Elastography: A Multicenter Phantom Study.

Background MR elastography (MRE) has been shown to have excellent performance for noninvasive liver fibrosis staging. However, there is limited knowledge regarding the precision and test-retest repeatability of stiffness measurement with MRE in the multicenter setting. Purpose To determine the precision and test-retest repeatability of stiffness measurement with MRE across multiple centers using the same phantoms. Materials and Methods In this study, three cylindrical phantoms made of polyvinyl chloride gel mimicking different degrees of liver stiffness in humans (phantoms 1-3: soft, medium, and hard stiffness, respectively) were evaluated. Between January 2021 and January 2022, phantoms were circulated between five different centers and scanned with 10 MRE-equipped clinical 1.5-T and 3-T systems from three major vendors, using two-dimensional (2D) gradient-recalled echo (GRE) imaging and/or 2D spin-echo (SE) echo-planar imaging (EPI). Similar MRE acquisition parameters, hardware, and reconstruction algorithms were used at each center. Mean stiffness was measured by a single observer for each phantom and acquisition on a single section. Stiffness measurement precision and same-session test-retest repeatability were assessed using the coefficient of variation (CV) and the repeatability coefficient (RC), respectively. Results The mean precision represented by the CV was 5.8% (95% CI: 3.8, 7.7) for all phantoms and both sequences combined. For all phantoms, 2D GRE achieved a CV of 4.5% (95% CI: 3.3, 5.7) whereas 2D SE EPI achieved a CV of 7.8% (95% CI: 3.1, 12.6). The mean RC of stiffness measurement was 5.8% (95% CI: 3.7, 7.8) for all phantoms and both sequences combined, 4.9% (95% CI: 2.7, 7.0) for 2D GRE, and 7.0% (95% CI: 2.9, 11.2) for 2D SE EPI (all phantoms). Conclusion MRE had excellent in vitro precision and same-session test-retest repeatability in the multicenter setting when similar imaging protocols, hardware, and reconstruction algorithms were used. © RSNA, 2024 Supplemental material is available for this article. See also the editorial by Tang in this issue.

Simultaneous Multiparameter Mapping of the Liver in a Single Breath-Hold or Respiratory-Triggered Acquisition Using Multi-Inversion Spin and Gradient Echo MRI.

BACKGROUND: Quantitative parametric mapping is an increasingly important tool for noninvasive assessment of chronic liver disease. Conventional parametric mapping techniques require multiple breath-held acquisitions and provide limited anatomic coverage. PURPOSE: To investigate a multi-inversion spin and gradient echo (MI-SAGE) technique for simultaneous estimation of T1, T2, and T2* of the liver. STUDY TYPE: Prospective. SUBJECTS: Sixteen research participants, both adult and pediatric (age 17.5 ± 4.6 years, eight male), with and without known liver disease (seven asymptomatic healthy controls, two fibrotic liver disease, five steatotic liver disease, and two fibrotic and steatotic liver disease). FIELD STRENGTH/SEQUENCE: 1.5 T, single breath-hold and respiratory triggered MI-SAGE, breath-hold modified Look-Locker inversion recovery (MOLLI, T1 mapping), breath-hold gradient and spin echo (GRASE, T2 mapping), and multiple gradient echo (mGRE, T2* mapping) sequences. ASSESSMENT: Agreement between hepatic T1, T2, and T2* estimated using MI-SAGE and conventional parametric mapping sequences was evaluated. Repeatability and reproducibility of MI-SAGE were evaluated using a same-session acquisition and second-session acquisition. STATISTICAL TESTS: Bland-Altman analysis with bias assessment and limits of agreement (LOA) and intraclass correlation coefficients (ICC). RESULTS: Hepatic T1, T2, and T2* estimates obtained using the MI-SAGE technique had mean biases of 72 (LOA: -22 to 166) msec, -3 (LOA: -10 to 5) msec, and 2 (LOA: -5 to 8) msec (single breath-hold) and 36 (LOA: -43 to 120) msec, -3 (LOA: -17 to 11) msec, and 4 (LOA: -3 to 11) msec (respiratory triggered), respectively, in comparison to conventional acquisitions using MOLLI, GRASE, and mGRE. All MI-SAGE estimates had strong repeatability and reproducibility (ICC > 0.72). DATA CONCLUSION: Hepatic T1, T2, and T2* estimates obtained using an MI-SAGE technique were comparable to conventional methods, although there was a 12%/6% for breath-hold/respiratory triggered underestimation of T1 values compared to MOLLI. Both respiratory triggered and breath-hold MI-SAGE parameter maps demonstrated strong repeatability and reproducibility. LEVEL OF EVIDENCE: 1 TECHNICAL EFFICACY: Stage 2.

Proposal of an Abbreviated Noncontrast MR Enterography Protocol for Patients With Crohn Disease.

MR enterography (MRE) protocols used in patients with Crohn disease are burdened by long acquisition time, high cost, and suboptimal patient experience. For several indications, highly diagnostic MRE can be performed in five or fewer sequences, without IV contrast material or antiperistaltic medication and with an examination room time of less than 12 minutes. As such, MRE could be more patient friendly, more frequently performed, and require fewer health care resources.

Associations of Liver Stiffness Measured by Ultrasound Shear-Wave Elastography With Portal Hypertension and Circulatory Failure in Individuals With Fontan Circulation.

BACKGROUND. The Fontan operation palliates single-ventricle congenital heart disease but causes hepatic congestion with associated progressive hepatic fibrosis. OBJECTIVE. The purpose of this study was to evaluate associations between liver stiffness measured using ultrasound (US) shear-wave elastography (SWE) in patients with Fontan palliation and the occurrence of portal hypertension and Fontan circulatory failure during follow-up. METHODS. This retrospective study included 119 individuals 10 years old or older (median age, 19.1 years; 61 female patients, 58 male patients) with Fontan circulation who underwent liver US with 2D SWE from January 1, 2015, to January 1, 2022, and had 1 year or more of clinical follow-up (unless experiencing earlier outcome-related events). Median liver stiffness from the initial US examination was documented. Varices, ascites, splenomegaly, and thrombocytopenia (VAST) scores (range, 0-4) were determined as a marker of portal hypertension on initial US examination and 1 year or more of follow-up imaging (US, CT, or MRI). Composite clinical outcome for Fontan circulatory failure (death, mechanical circulatory support, cardiac transplant, or unexpected Fontan circulation-related hospitalization) was assessed. Analysis included the Wilcoxon rank sum test, logistic regression analysis with stepwise variable selection, and ROC analysis. RESULTS. Median initial liver stiffness was 2.22 m/s. Median initial VAST score was 0 (IQR, 0-1); median follow-up VAST score was 1 (IQR, 0-2) (p = .004). Fontan circulatory failure occurred in 37 of 119 (31%) patients (median follow-up, 3.4 years). Initial liver stiffness was higher in patients with a follow-up VAST score of 1 or greater (2.37 m/s) than in those with a follow-up VAST score of 0 (2.08 m/s) (p = .005), and initial liver stiffness was higher in patients with (2.43 m/s) than without (2.10 m/s) Fontan circulatory failure during follow-up (p < .001). Initial liver stiffness was the only significant independent predictor of Fontan circulatory failure (OR = 3.76; p < .001); age, sex, Fontan operation type, dominant ventricular morphology, and initial VAST score were not independent predictors. Initial liver stiffness had an AUC of 0.70 (sensitivity, 79%; specificity, 57%; threshold, > 2.11 m/s) for predicting a follow-up VAST score of 1 or greater and an AUC of 0.74 (sensitivity, 84%; specificity, 52%; threshold, > 2.12 m/s) for predicting Fontan circulatory failure. CONCLUSION. In patients with Fontan circulation, increased initial liver stiffness was associated with portal hypertension and circulatory failure during follow-up, although it had moderate performance in predicting these outcomes. CLINICAL IMPACT. US SWE may play a role in post-Fontan surveillance, supporting tailored medical and surgical care.

Performance Analysis in Children of Traditional and Deep Learning CT Lung Nodule Computer-Aided Detection Systems Trained on Adults.

BACKGROUND. Although primary lung cancer is rare in children, chest CT is commonly performed to assess for lung metastases in children with cancer. Lung nodule computer-aided detection (CAD) systems have been designed and studied primarily using adult training data, and the efficacy of such systems when applied to pediatric patients is poorly understood. OBJECTIVE. The purpose of this study was to evaluate in children the diagnostic performance of traditional and deep learning CAD systems trained with adult data for the detection of lung nodules on chest CT scans and to compare the ability of such systems to generalize to children versus to other adults. METHODS. This retrospective study included pediatric and adult chest CT test sets. The pediatric test set comprised 59 CT scans in 59 patients (30 boys, 29 girls; mean age, 13.1 years; age range, 4-17 years), which were obtained from November 30, 2018, to August 31, 2020; lung nodules were annotated by fellowship-trained pediatric radiologists as the reference standard. The adult test set was the publicly available adult Lung Nodule Analysis (LUNA) 2016 subset 0, which contained 89 deidentified scans with previously annotated nodules. The test sets were processed through the traditional FlyerScan (github.com/rhardie1/FlyerScanCT) and deep learning Medical Open Network for Artificial Intelligence (MONAI; github.com/Project-MONAI/model-zoo/releases) lung nodule CAD systems, which had been trained on separate sets of CT scans in adults. Sensitivity and false-positive (FP) frequency were calculated for nodules measuring 3-30 mm; nonoverlapping 95% CIs indicated significant differences. RESULTS. Operating at two FPs per scan, on pediatric testing data FlyerScan and MONAI showed significantly lower detection sensitivities of 68.4% (197/288; 95% CI, 65.1-73.0%) and 53.1% (153/288; 95% CI, 46.7-58.4%), respectively, than on adult LUNA 2016 subset 0 testing data (83.9% [94/112; 95% CI, 79.1-88.0%] and 95.5% [107/112; 95% CI, 90.0-98.4%], respectively). Mean nodule size was smaller (p < .001) in the pediatric testing data (5.4 ± 3.1 [SD] mm) than in the adult LUNA 2016 subset 0 testing data (11.0 ± 6.2 mm). CONCLUSION. Adult-trained traditional and deep learning-based lung nodule CAD systems had significantly lower sensitivity for detection on pediatric data than on adult data at a matching FP frequency. The performance difference may relate to the smaller size of pediatric lung nodules. CLINICAL IMPACT. The results indicate a need for pediatric-specific lung nodule CAD systems trained on data specific to pediatric patients.

Implementation of a Program for Appendicitis MRI in a Pediatric Hospital.

MRI is increasingly used as an alternate to CT for the evaluation of suspected appendicitis in pediatric patients presenting to the emergency department (ED) with abdominal pain, when further imaging is needed after an initial ultrasound examination. The available literature shows a similar diagnostic performance of MRI and CT in this setting. At the authors' institution, to evaluate for appendicitis in children in the ED, MRI is performed using a rapid three-sequence free-breathing protocol without IV contrast media. Implementation of an MRI program for appendicitis in children involves multiple steps, including determination of imaging resource availability, collaboration with other services to develop imaging pathways, widespread educational efforts, and regular quality review. Such programs can face numerous practice-specific challenges, such as those involving scanner capacity, costs, and buy-in of impacted groups. Nonetheless, through careful consideration of these factors, MRI can be used to positively impact the care of children presenting to the ED with suspected appendicitis. This Clinical Perspective aims to provide guidance on the development of a program for appendicitis MRI in children, drawing on one institution's experience while highlighting the advantages of MRI and practical strategies for overcoming potential barriers.

Machine Learning Diagnosis of Small-Bowel Crohn Disease Using T2-Weighted MRI Radiomic and Clinical Data.

BACKGROUND. Radiologists have variable diagnostic performance and considerable interreader variability when interpreting MR enterography (MRE) examinations for suspected Crohn disease (CD). OBJECTIVE. The purposes of this study were to develop a machine learning method for predicting ileal CD by use of radiomic features of ileal wall and mesenteric fat from noncontrast T2-weighted MRI and to compare the performance of the method with that of expert radiologists. METHODS. This single-institution study included retrospectively identified patients who underwent MRE for suspected ileal CD from January 1, 2020, to January 31, 2021, and prospectively enrolled participants (patients with newly diagnosed ileal CD or healthy control participants) from December 2018 to October 2021. Using axial T2-weighted SSFSE images, a radiologist selected two slices showing greatest terminal ileal wall thickening. Four ROIs were segmented, and radiomic features were extracted from each ROI. After feature selection, support-vector machine models were trained to classify the presence of ileal CD. Three fellowship-trained pediatric abdominal radiologists independently classified the presence of ileal CD on SSFSE images. The reference standard was clinical diagnosis of ileal CD based on endoscopy and biopsy results. Radiomic-only, clinical-only, and radiomic-clinical ensemble models were trained and evaluated by nested cross-validation. RESULTS. The study included 135 participants (67 female, 68 male; mean age, 15.2 ± 3.2 years); 70 were diagnosed with ileal CD. The three radiologists had accuracies of 83.7% (113/135), 88.1% (119/135), and 86.7% (117/135) for diagnosing CD; consensus accuracy was 88.1%. Interradiologist agreement was substantial (κ = 0.78). The best-performing ROI was bowel core (AUC, 0.95; accuracy, 89.6%); other ROIs had worse performance (whole-bowel AUC, 0.86; fat-core AUC, 0.70; whole-fat AUC, 0.73). For the clinical-only model, AUC was 0.85 and accuracy was 80.0%. The ensemble model combining bowel-core radiomic and clinical models had AUC of 0.98 and accuracy of 93.5%. The bowel-core radiomic-only model had significantly greater accuracy than radiologist 1 (p = .009) and radiologist 2 (p = .02) but not radiologist 3 (p > .99) or the radiologists in consensus (p = .05). The ensemble model had greater accuracy than the radiologists in consensus (p = .02). CONCLUSION. A radiomic machine learning model predicted CD diagnosis with better performance than two of three expert radiologists. Model performance improved when radiomic data were ensembled with clinical data. CLINICAL IMPACT. Deployment of a radiomic-based model including T2-weighted MRI data could decrease interradiologist variability and increase diagnostic accuracy for pediatric CD.

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. The purpose of this article is 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 years) 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 Artificial Intelligence (MONAI) framework underwent training using native training (NT), relying solely on institutional datasets, and transfer learning (TL), incorporating pretraining on public datasets. For comparison, TotalSegmentator, 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, the DSC for TotalSegmentator, NT models, and TL models for normal liver was 0.953, 0.964-0.965, and 0.965-0.966, respectively; for normal spleen, 0.914, 0.942-0.945, and 0.937-0.945; for normal pancreas, 0.733, 0.774-0.785, and 0.775-0.786; and for pancreas with pancreatitis, 0.703, 0.590-0.640, and 0.667-0.711. For public pediatric data, the DSC for TotalSegmentator, NT models, and TL models for liver was 0.952, 0.871-0.908, and 0.941-0.946, respectively; for spleen, 0.905, 0.771-0.827, and 0.897-0.926; and for pancreas, 0.700, 0.577-0.648, and 0.693-0.736. For public primarily adult data, the DSC for TotalSegmentator, NT models, and TL models for liver was 0.991, 0.633-0.750, and 0.926-0.952, respectively; for spleen, 0.983, 0.569-0.604, and 0.923-0.947; and for pancreas, 0.909, 0.148-0.241, and 0.699-0.775. The DynUNet TL model was selected as the best-performing NT or TL model considering DSC values across organs and test datasets and was made available as an open-source 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 Total-Segmentator 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.

Association Between MR Elastography Liver Stiffness and Histologic Liver Fibrosis in Children and Young Adults With Autoimmune Liver Disease.

BACKGROUND. Liver fibrosis is an important clinical endpoint of the progression of autoimmune liver disease (AILD); its monitoring would benefit from noninvasive imaging tools. OBJECTIVE. The purpose of this study was to assess the relationship between MR elastography (MRE) liver stiffness measurements and histologic liver fibrosis, as well as to evaluate the performance of MRE and biochemical-based clinical markers for stratifying histologic liver fibrosis severity, in children and young adults with AILD. METHODS. This retrospective study used an existing institutional registry of children and young adults diagnosed with AILD (primary sclerosing cholangitis [PSC], autoimmune sclerosing cholangitis [ASC], or autoimmune hepatitis [AIH]). The registry was searched to identify patients who underwent both a research abdominal 1.5-T MRI examination that included liver MRE (performed for registry enrollment) and a clinically indicated liver biopsy within 6 months of that examination. MRE used a 2D gradient-recalled echo sequence. One analyst measured mean liver shear stiffness (in kilopascals) for each examination. Laboratory markers of liver fibrosis (aspartate aminotransferase-to-platelet ratio index [APRI] and fibrosis-4 [FIB-4] score) were recorded. For investigational purposes, one pathologist, blinded to clinical and MRI data, determined histologic Metavir liver fibrosis stage. The Spearman rank order correlation coefficient was calculated between MRE liver stiffness and Metavir liver fibrosis stage. ROC analysis was used to evaluate diagnostic performance for identifying advanced fibrosis (i.e., differentiating Metavir F0-F1 from F2-F4 fibrosis), and sensitivity and specificity were calculated using the Youden index. RESULTS. The study included 46 patients (median age, 16.6 years [IQR, 13.7-17.8 years]; 20 female patients, 26 male patients); 12 had PSC, 10 had ASC, and 24 had AIH. Median MRE liver stiffness was 2.9 kPa (IQR, 2.2-4.0 kPa). MRE liver stiffness and Metavir fibrosis stage showed strong positive correlation (ρ = 0.68). For identifying advanced liver fibrosis, MRE liver stiffness had an AUC of 0.81, with sensitivity of 65.4% and specificity of 90.0%; APRI had an AUC of 0.72, with sensitivity of 64.0% and specificity of 80.0%; and FIB-4 score had an AUC of 0.71, with sensitivity of 60.0% and specificity of 85.0%. CONCLUSION. MRE liver stiffness measurements were associated with histologic liver fibrosis severity. CLINICAL IMPACT. The findings support a role for MRE in noninvasive monitoring of liver stiffness, a surrogate for fibrosis, in children and young adults with AILD. TRIAL REGISTRATION. ClinicalTrials.gov NCT03175471.

Congestive Hepatopathy: Pathophysiology, Workup, and Imaging Findings with Pathologic Correlation.

Liver congestion is increasingly encountered in clinical practice and presents diagnostic pitfalls of which radiologists must be aware. The complex altered hemodynamics associated with liver congestion leads to diffuse parenchymal changes and the development of benign and malignant nodules. Distinguishing commonly encountered benign hypervascular lesions, such as focal nodular hyperplasia (FNH)-like nodules, from hepatocellular carcinoma (HCC) can be challenging due to overlapping imaging features. FNH-like lesions enhance during the hepatic arterial phase and remain isoenhancing relative to the background liver parenchyma but infrequently appear to wash out at delayed phase imaging, similar to what might be seen with HCC. Heterogeneity, presence of an enhancing capsule, washout during the portal venous phase, intermediate signal intensity at T2-weighted imaging, restricted diffusion, and lack of uptake at hepatobiliary phase imaging point toward the diagnosis of HCC, although these features are not sensitive individually. It is important to emphasize that the Liver Imaging Reporting and Data System (LI-RADS) algorithm cannot be applied in congested livers since major LI-RADS features lack specificity in distinguishing HCC from benign hypervascular lesions in this population. Also, the morphologic changes and increased liver stiffness caused by congestion make the imaging diagnosis of cirrhosis difficult. The authors discuss the complex liver macro- and microhemodynamics underlying liver congestion; propose a more inclusive approach to and conceptualization of liver congestion; describe the pathophysiology of liver congestion, hepatocellular injury, and the development of benign and malignant nodules; review the imaging findings and mimics of liver congestion and hypervascular lesions; and present a diagnostic algorithm for approaching hypervascular liver lesions. ©RSNA, 2024 Test Your Knowledge questions for this article are available in the supplemental material.

Evaluation of 3D T1-weighted spoiled gradient echo MR image quality using artificial intelligence image reconstruction techniques in the pediatric brain.

PURPOSE: To assess image quality and diagnostic confidence of 3D T1-weighted spoiled gradient echo (SPGR) MRI using artificial intelligence (AI) reconstruction. MATERIALS AND METHODS: This prospective, IRB-approved study enrolled 50 pediatric patients (mean age = 11.8 ± 3.1 years) undergoing clinical brain MRI. In addition to standard of care (SOC) compressed SENSE (CS = 2.5), 3D T1-weighted SPGR images were obtained with higher CS acceleration factors (5 and 8) to evaluate the ability of AI reconstruction to improve image quality and reduce scan time. Images were reviewed independently on dedicated research PACS workstations by two neuroradiologists. Quantitative analysis of signal intensities to calculate apparent grey and white matter signal to noise (aSNR) and grey-white matter apparent contrast to noise ratios (aCNR) was performed. RESULTS: AI improved overall image quality compared to standard CS reconstruction in 35% (35/100) of evaluations in CS = 2.5 (average scan time = 221 ± 6.9 s), 100% (46/46) of CS = 5 (average scan time = 113.3 ± 4.6 s) and 94% (47/50) of CS = 8 (average scan time = 74.1 ± 0.01 s). Quantitative analysis revealed significantly higher grey matter aSNR, white matter aSNR and grey-white matter aCNR with AI reconstruction compared to standard reconstruction for CS 5 and 8 (all p-values < 0.001), however not for CS 2.5. CONCLUSIONS: AI reconstruction improved overall image quality and gray-white matter qualitative and quantitative aSNR and aCNR in highly accelerated (CS = 5 and 8) 3D T1W SPGR images in the majority of pediatric patients.

Impact of Emerging Deep Learning-Based MR Image Reconstruction Algorithms on Abdominal MRI Radiomic Features.

OBJECTIVE: This study aims to evaluate, on one MRI vendor's platform, the impact of deep learning (DL)-based reconstruction techniques on MRI radiomic features compared to conventional image reconstruction techniques. METHODS: Under IRB approval and informed consent, we prospectively collected undersampled coronal T2-weighted MR images of the abdomen (1.5 T; Philips Healthcare) from 17 pediatric and adult subjects and reconstructed them using a conventional image reconstruction technique (compressed sensitivity encoding [C-SENSE]) and two DL-based reconstruction techniques (SmartSpeed [Philips Healthcare, US FDA cleared] and SmartSpeed with Super Resolution [SmartSpeed-SuperRes, not US FDA cleared to date]). Eight regions of interest (ROIs) across organs/tissues (liver, spleen, kidney, pancreas, fat, and muscle) were manually placed. Eighty-six MRI radiomic features were then extracted. Pearson's correlation coefficients (PCCs) and intraclass correlation coefficients (ICCs) were calculated between (A) C-SENSE versus SmartSpeed, and (B) C-SENSE versus SmartSpeed-SuperRes. To quantify the impact from the perspective of the whole MR image, cross-ROI mean PCCs and ICCs were calculated for individual radiomic features. The impact of image reconstruction on individual radiomic features in different organs/tissues was evaluated using ANOVA analyses. RESULTS: According to cross-ROI mean PCCs, 50 out of 86 radiomic features were highly correlated (PCC, ≥0.8) between SmartSpeed and C-SENSE, whereas only 15 radiomic features were highly correlated between SmartSpeed-SuperRes and C-SENSE reconstructions. According to cross-ROI mean ICCs, 58 out of 86 radiomic features had high agreements (ICC ≥0.75) between SmartSpeed and C-SENSE, whereas only 9 radiomic features had high agreements between SmartSpeed-SuperRes and C-SENSE reconstructions. For SmartSpeed reconstruction, the psoas muscle ROI appeared to be impacted most with the lowest median (IQR) correlation of 0.57 (0.25). The circular liver ROI was impacted most by SmartSpeed-SuperRes (PCC, 0.60 [0.22]). ANOVA analyses suggest that the impact of DL reconstruction algorithms on radiomic features varies significantly among different organs/tissues (P < 0.001). CONCLUSIONS: MRI radiomic features are significantly altered by DL-based reconstruction compared to a conventional reconstruction technique. The impact of DL reconstruction algorithms on radiomic features varies significantly between different organs/tissues.

Quantitative MRI in children with Crohn's disease - where do we stand?

Crohn's disease (CD) is a chronic inflammatory condition that affects the gastrointestinal tract, particularly the ileum and colon. This disease is characterized by recurrent bouts of intestinal inflammation with subsequent bowel wall damage, including scarring (i.e., fibrosis) and abnormal smooth muscle proliferation. MR enterography, an MRI examination tailored to assess the small bowel, is a first-line diagnostic tool for diagnosing CD in children, characterization and monitoring of disease severity and extent, and assessment of disease-related complications. To date, such MRI evaluations have been mostly qualitative, which can adversely impact diagnostic performance and inter-radiologist agreement. Quantitative MRI methods have been shown to aid in the evaluation of a variety of medical conditions and have been increasingly investigated in children and adults with CD. In CD, such objective techniques have been used to assist with diagnosis, assess treatment response, and characterize bowel wall histologic abnormalities. In the current work, we will review quantitative MRI methods for detecting and measuring intestinal active inflammation (MRI-based scoring systems, T1 relaxation mapping, diffusion-weighted imaging, intra-voxel incoherent motion, mesenteric phase contrast), bowel wall damage (magnetization transfer), and motility (quantitative cine imaging) in small bowel CD, with an emphasis on the pediatric population.

Body MRI in pediatrics: where we are and what the future holds.

Body magnetic resonance imaging (MRI) is increasingly used for disease diagnosis, characterization, and monitoring in children of all ages. MRI has numerous advantages when compared to other imaging modalities, including a lack of ionizing radiation, superior soft tissue image contrast, and ability to provide objective, quantitative assessments. As MRI continues to evolve, pediatric body MRI examinations of the future will certainly be different than our current and past protocols. In this review article, we will discuss the present and likely future states of pediatric body MRI, including the increasing application of quantitative MRI methods, faster imaging techniques and implementation of abbreviated targeted protocols, and the growing use of artificial intelligence methods.

Rapid abdominopelvic MR imaging in the emergency department: establishing a program and addressing the challenges.

Utilization of magnetic resonance imaging (MRI) in the pediatric emergency room or urgent care setting for abdominopelvic indications has been increasing. The creation and implementation of rapid urgent MRI programs can have various challenges. The purpose of this article is to describe a framework for the creation of a rapid urgent abdominopelvic MRI program in the pediatric emergency room setting.

Evaluation of T2W FLAIR MR image quality using artificial intelligence image reconstruction techniques in the pediatric brain.

BACKGROUND: Artificial intelligence (AI) reconstruction techniques have the potential to improve image quality and decrease imaging time. However, these techniques must be assessed for safe and effective use in clinical practice. OBJECTIVE: To assess image quality and diagnostic confidence of AI reconstruction in the pediatric brain on fluid-attenuated inversion recovery (FLAIR) imaging. MATERIALS AND METHODS: This prospective, institutional review board (IRB)-approved study enrolled 50 pediatric patients (median age=12 years, Q1=10 years, Q3=14 years) undergoing clinical brain MRI. T2-weighted (T2W) FLAIR images were reconstructed by both standard clinical and AI reconstruction algorithms (strong denoising). Images were independently rated by two neuroradiologists on a dedicated research picture archiving and communication system (PACS) to indicate whether AI increased, decreased, or had no effect on image quality compared to standard reconstruction. Quantitative analysis of signal intensities was also performed to calculate apparent signal to noise (aSNR) and apparent contrast to noise (aCNR) ratios. RESULTS: AI reconstruction was better than standard in 99% (reader 1, 49/50; reader 2, 50/50) for overall image quality, 99% (reader 1, 49/50; reader 2, 50/50) for subjective SNR, and 98% (reader 1, 49/50; reader 2, 49/50) for diagnostic preference. Quantitative analysis revealed significantly higher gray matter aSNR (30.6±6.5), white matter aSNR (21.4±5.6), and gray-white matter aCNR (7.1±1.6) in AI-reconstructed images compared to standard reconstruction (18±2.7, 14.2±2.8, 4.4±0.8, p<0.001) respectively. CONCLUSION: We conclude that AI reconstruction improved T2W FLAIR image quality in most patients when compared with standard reconstruction in pediatric patients.

MR enterography radiologic ulcers in newly diagnosed ileal Crohn disease in children: frequency, inter-radiologist agreement, and clinical correlation.

BACKGROUND: Radiologic ulcers are increasingly recognized as an imaging finding of bowel wall active inflammation in Crohn disease (CD). OBJECTIVE: To determine the frequency of ulcers at MR enterography (MRE) in children with newly diagnosed ileal CD, assess agreement between radiologists, and evaluate if their presence correlates with other imaging and clinical features of intestinal active inflammation. MATERIALS AND METHODS: This retrospective study included 108 consecutive pediatric patients (ages 6-18 years) with newly diagnosed ileal CD that underwent clinical MRE prior to treatment initiation between January 2021 and December 2022. MRE examinations were independently reviewed by three pediatric radiologists who indicated the presence vs. absence of ulcers, ulcer severity (categorical depth), and ulcer extent (categorical number of ulcers). Maximum bowel wall thickness and length of disease were measured and averaged across readers. Patient demographics and clinical inflammatory markers were documented from electronic health records. Inter-radiologist agreement was assessed using Fleiss' kappa (k) statistics. Student's t-test was used to compare continuous variables. RESULTS: Mean patient age was 13.9 years (67 [62%] boys). Radiologic ulcers were recorded in 64/108 (59.3%) cases by reader 1, 70/108 (64.8%) cases by reader 2, and 49/108 (45.4%) cases by reader 3 (k = 0.36). Based on majority consensus, radiologic ulcers were present in 60/108 (55.6%) participants. Inter-radiologist agreement for ulcer severity was k = 0.23, while ulcer extent was k = 0.66. There were significant differences in C-reactive protein, erythrocyte sedimentation rate, fecal calprotectin, albumin, maximum bowel wall thickness, and length of disease between patients without and with radiologic ulcers (P < 0.05). The sensitivity and specificity of MRE for detecting endoscopic ulcers were 66.7% (95% CI, 52.1-79.2%) and 69.2% (95% CI, 48.2-85.7%), respectively. CONCLUSION: Radiologic ulcers are visible in children with newly diagnosed ileal CD, although inter-radiologist agreement is only fair. The presence of ulcers is associated with clinical laboratory inflammatory markers as well as other MRE findings of disease activity and is an additional imaging finding that can be used to evaluate intestinal inflammation.

A novel collaborative self-supervised learning method for radiomic data.

The computer-aided disease diagnosis from radiomic data is important in many medical applications. However, developing such a technique relies on labeling radiological images, which is a time-consuming, labor-intensive, and expensive process. In this work, we present the first novel collaborative self-supervised learning method to solve the challenge of insufficient labeled radiomic data, whose characteristics are different from text and image data. To achieve this, we present two collaborative pretext tasks that explore the latent pathological or biological relationships between regions of interest and the similarity and dissimilarity of information between subjects. Our method collaboratively learns the robust latent feature representations from radiomic data in a self-supervised manner to reduce human annotation efforts, which benefits the disease diagnosis. We compared our proposed method with other state-of-the-art self-supervised learning methods on a simulation study and two independent datasets. Extensive experimental results demonstrated that our method outperforms other self-supervised learning methods on both classification and regression tasks. With further refinement, our method will have the potential advantage in automatic disease diagnosis with large-scale unlabeled data available.

Corrected T1 Mapping in Children and Young Adults With Autoimmune Liver Disease: Correlation With Histology.

The purpose of this study was to assess relationships between liver-corrected T1 (cT1) values (adjusted for T2* effect, MRI system manufacturer, and field strength) and histologic inflammation and fibrosis in 35 participants (15 women and girls, 20 boys and men; median age, 16.0 years) with autoimmune liver disease. At multivariable analysis, inflammation score (ß = 15.5) and sex (ß = 56.0 [female]) were independent predictors of cT1, and fibrosis score (ß = 32.3) and age (ß = 5.5) were independent predictors of cT1 IQR. Liver T1 may have relevance for assessing liver inflammatory activity and fibrosis stage.