Bounding box-based 3D AI model for user-guided volumetric segmentation of pancreatic ductal adenocarcinoma on standard-of-care CTs

• Published in: Pancreatology : official journal of the International Association of Pancreatology (IAP) ... [et al.] 2023-08, Vol.23 (5), p.522-529
• Main Authors: Mukherjee, Sovanlal, Korfiatis, Panagiotis, Khasawneh, Hala, Rajamohan, Naveen, Patra, Anurima, Suman, Garima, Singh, Aparna, Thakkar, Jay, Patnam, Nandakumar G., Trivedi, Kamaxi H., Karbhari, Aashna, Chari, Suresh T., Truty, Mark J., Halfdanarson, Thorvardur R., Bolan, Candice W., Sandrasegaran, Kumar, Majumder, Shounak, Goenka, Ajit H.
• Format: Article
• Language: English
• Subjects: Aged; Artificial intelligence; Carcinoma, Pancreatic Ductal - diagnostic imaging; Humans; Image Processing, Computer-Assisted - methods; Male; Middle Aged; Neural Networks, Computer; Pancreas; Pancreatic ductal carcinoma; Pancreatic Ducts; Pancreatic Neoplasms - diagnostic imaging; Tomography, X-Ray Computed - methods; X-ray computed tomography
• ISSN: 1424-3903; 1424-3911; 1424-3911
• DOI: 10.1016/j.pan.2023.05.008

To develop a bounding-box-based 3D convolutional neural network (CNN) for user-guided volumetric pancreas ductal adenocarcinoma (PDA) segmentation. Reference segmentations were obtained on CTs (2006–2020) of treatment-naïve PDA. Images were algorithmically cropped using a tumor-centered bounding box for training a 3D nnUNet-based-CNN. Three radiologists independently segmented tumors on test subset, which were combined with reference segmentations using STAPLE to derive composite segmentations. Generalizability was evaluated on Cancer Imaging Archive (TCIA) (n = 41) and Medical Segmentation Decathlon (MSD) (n = 152) datasets. Total 1151 patients [667 males; age:65.3 ± 10.2 years; T1:34, T2:477, T3:237, T4:403; mean (range) tumor diameter:4.34 (1.1–12.6)-cm] were randomly divided between training/validation (n = 921) and test subsets (n = 230; 75% from other institutions). Model had a high DSC (mean ± SD) against reference segmentations (0.84 ± 0.06), which was comparable to its DSC against composite segmentations (0.84 ± 0.11, p = 0.52). Model-predicted versus reference tumor volumes were comparable (mean ± SD) (29.1 ± 42.2-cc versus 27.1 ± 32.9-cc, p = 0.69, CCC = 0.93). Inter-reader variability was high (mean DSC 0.69 ± 0.16), especially for smaller and isodense tumors. Conversely, model's high performance was comparable between tumor stages, volumes and densities (p > 0.05). Model was resilient to different tumor locations, status of pancreatic/biliary ducts, pancreatic atrophy, CT vendors and slice thicknesses, as well as to the epicenter and dimensions of the bounding-box (p > 0.05). Performance was generalizable on MSD (DSC:0.82 ± 0.06) and TCIA datasets (DSC:0.84 ± 0.08). A computationally efficient bounding box-based AI model developed on a large and diverse dataset shows high accuracy, generalizability, and robustness to clinically encountered variations for user-guided volumetric PDA segmentation including for small and isodense tumors. AI-driven bounding box-based user-guided PDA segmentation offers a discovery tool for image-based multi-omics models for applications such as risk-stratification, treatment response assessment, and prognostication, which are urgently needed to customize treatment strategies to the unique biological profile of each patient's tumor.