TY - JOUR
T1 - Enhancing detection of various pancreatic lesions on endoscopic ultrasound through artificial intelligence
T2 - a basis for computer-aided detection systems
AU - Konikoff, Tom
AU - Loebl, Nadav
AU - Benson, Ariel A.
AU - Green, Orr
AU - Sandler, Hunter
AU - Gingold-Belfer, Rachel
AU - Levi, Zohar
AU - Perl, Leor
AU - Dotan, Iris
AU - Shamah, Steven
N1 - Publisher Copyright:
© 2024 Journal of Gastroenterology and Hepatology Foundation and John Wiley & Sons Australia, Ltd.
PY - 2025/1
Y1 - 2025/1
N2 - Background and Aim: Endoscopic ultrasound (EUS) is the most sensitive method for evaluation of pancreatic lesions but is limited by significant operator dependency. Artificial intelligence (AI), in the form of computer-aided detection (CADe) systems, has shown potential in increasing accuracy and bridging operator dependency in several endoscopic domains. However, the complexity of integrating AI into EUS is far more challenging. This aims to develop and test the basis for a CADe system for real-time detection and segmentation of all pancreatic lesions. Methods: In this single-center study EUS studies of pancreatic findings were included. Lesions were outlined by two expert (>5 years performing EUS) endoscopists, and the two leading types of models were benchmarked. The models' performance was evaluated through per-pixel intersection over union (IoU). Results: A total of 1497 EUS images from 165 patients were evaluated. The dataset included malignancies, neuroendocrine tumors, benign cysts, chronic and acute pancreatitis, normal fatty pancreas, and benign lesions. The best model demonstrated detection and segmentation on the test set with a mean IoU of 0.73, achieving a PPV, NPV, total accuracy, and ROC of 0.82, 0.96, 0.95, and 0.95, respectively. The algorithm is adaptable for real-time processing. Conclusions: We developed and tested deep learning models for real-time detection and segmentation of pancreatic lesions during EUS with promising results. This constitutes the basis for a CADe system for EUS, which could be valuable in future detection and evaluation of pancreatic lesions. Further studies for validation and generalization are underway.
AB - Background and Aim: Endoscopic ultrasound (EUS) is the most sensitive method for evaluation of pancreatic lesions but is limited by significant operator dependency. Artificial intelligence (AI), in the form of computer-aided detection (CADe) systems, has shown potential in increasing accuracy and bridging operator dependency in several endoscopic domains. However, the complexity of integrating AI into EUS is far more challenging. This aims to develop and test the basis for a CADe system for real-time detection and segmentation of all pancreatic lesions. Methods: In this single-center study EUS studies of pancreatic findings were included. Lesions were outlined by two expert (>5 years performing EUS) endoscopists, and the two leading types of models were benchmarked. The models' performance was evaluated through per-pixel intersection over union (IoU). Results: A total of 1497 EUS images from 165 patients were evaluated. The dataset included malignancies, neuroendocrine tumors, benign cysts, chronic and acute pancreatitis, normal fatty pancreas, and benign lesions. The best model demonstrated detection and segmentation on the test set with a mean IoU of 0.73, achieving a PPV, NPV, total accuracy, and ROC of 0.82, 0.96, 0.95, and 0.95, respectively. The algorithm is adaptable for real-time processing. Conclusions: We developed and tested deep learning models for real-time detection and segmentation of pancreatic lesions during EUS with promising results. This constitutes the basis for a CADe system for EUS, which could be valuable in future detection and evaluation of pancreatic lesions. Further studies for validation and generalization are underway.
KW - AI in EUS
KW - CADe
KW - machine learning in EUS
KW - pancreatic cancer
UR - http://www.scopus.com/inward/record.url?scp=85209068450&partnerID=8YFLogxK
U2 - 10.1111/jgh.16814
DO - 10.1111/jgh.16814
M3 - ???researchoutput.researchoutputtypes.contributiontojournal.article???
C2 - 39538430
AN - SCOPUS:85209068450
SN - 0815-9319
VL - 40
SP - 235
EP - 240
JO - Journal of Gastroenterology and Hepatology (Australia)
JF - Journal of Gastroenterology and Hepatology (Australia)
IS - 1
ER -