RESUMO
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.
RESUMO
BACKGROUND/OBJECTIVES: Pancreas volume derived from imaging may objectively reveal volume loss relevant to identifying sequelae of acute pancreatitis (AP) and ultimately diagnosing chronic pancreatitis (CP). The purposes of this study were to: (1) quantify pancreas volume by imaging in children with either (a) a single episode of AP or (b) acute recurrent pancreatitis (ARP), and (2) compare these volumes to normative volumes. METHODS: This retrospective study was institutional review board approved. A single observer segmented the pancreas (3D Slicer; slicer.org) on n = 30 CT and MRI exams for 23 children selected from a prospective registry of patients with either an index attack of AP or with ARP after a known index attack date. Patients with CP were excluded. Segmented pancreas volumes were compared to published normal values. RESULTS: Mean pancreas volumes normalized to body surface area (BSA) in the index AP and ARP groups were 38.2 mL/m2 (range: 11.8-73.5 mL/m2) and 27.9 mL/m2 (range: 8.0-69.2 mL/m2) respectively. 43 % (6/14) of patients post-AP had volumes below the 25th percentile, 1 (17 %) of which was below the 5th percentile (p = 0.3027 vs. a normal distribution). Post-ARP, 44 % (7/16) of patients had volumes below the 5th percentile (p < 0.001). CONCLUSIONS: A significant fraction (40 %) of children with ARP have pancreas volumes <5th percentile for BSA even in the absence of CP. A similar, but not statistically significant, fraction have pancreas volumes <25th percentile after an index attack of AP. Pancreatic parenchymal volume deserves additional investigation as an objective marker of parenchymal damage from acute pancreatitis and of progressive pancreatitis in children.
