Hiroaki Saito1, Tomonori Aoki2, Kazuharu Aoyama3, Yusuke Kato3, Akiyoshi Tsuboi4, Atsuo Yamada2, Mitsuhiro Fujishiro5, Shiro Oka4, Soichiro Ishihara6, Tomoki Matsuda7, Masato Nakahori7, Shinji Tanaka4, Kazuhiko Koike2, Tomohiro Tada8. 1. Department of Gastroenterology, Sendai Kousei Hospital, Miyagi, Japan. Electronic address: h.saito0515@gmail.com. 2. Department of Gastroenterology, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan. 3. AI Medical Service Inc., Tokyo, Japan. 4. Department of Endoscopy, Hiroshima University Hospital, Hiroshima, Japan. 5. Department of Gastroenterology and Hepatology, Nagoya University Graduate School of Medicine, Aichi, Japan. 6. Department of Surgical Oncology, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan; Tada Tomohiro Institute of Gastroenterology and Proctology, Saitama, Japan. 7. Department of Gastroenterology, Sendai Kousei Hospital, Miyagi, Japan. 8. AI Medical Service Inc., Tokyo, Japan; Department of Surgical Oncology, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan; Tada Tomohiro Institute of Gastroenterology and Proctology, Saitama, Japan.
Abstract
BACKGROUND AND AIMS: Protruding lesions of the small bowel vary in wireless capsule endoscopy (WCE) images, and their automatic detection may be difficult. We aimed to develop and test a deep learning-based system to automatically detect protruding lesions of various types in WCE images. METHODS: We trained a deep convolutional neural network (CNN), using 30,584 WCE images of protruding lesions from 292 patients. We evaluated CNN performance by calculating the area under the receiver operating characteristic curve (AUC), sensitivity, and specificity, using an independent set of 17,507 test images from 93 patients, including 7507 images of protruding lesions from 73 patients. RESULTS: The developed CNN analyzed 17,507 images in 530.462 seconds. The AUC for detection of protruding lesions was 0.911 (95% confidence interval [Cl], 0.9069-0.9155). The sensitivity and specificity of the CNN were 90.7% (95% CI, 90.0%-91.4%) and 79.8% (95% CI, 79.0%-80.6%), respectively, at the optimal cut-off value of 0.317 for probability score. In a subgroup analysis of the category of protruding lesions, the sensitivities were 86.5%, 92.0%, 95.8%, 77.0%, and 94.4% for the detection of polyps, nodules, epithelial tumors, submucosal tumors, and venous structures, respectively. In individual patient analyses (n = 73), the detection rate of protruding lesions was 98.6%. CONCLUSION: We developed and tested a new computer-aided system based on a CNN to automatically detect various protruding lesions in WCE images. Patient-level analyses with larger cohorts and efforts to achieve better diagnostic performance are necessary in further studies.
BACKGROUND AND AIMS: Protruding lesions of the small bowel vary in wireless capsule endoscopy (WCE) images, and their automatic detection may be difficult. We aimed to develop and test a deep learning-based system to automatically detect protruding lesions of various types in WCE images. METHODS: We trained a deep convolutional neural network (CNN), using 30,584 WCE images of protruding lesions from 292 patients. We evaluated CNN performance by calculating the area under the receiver operating characteristic curve (AUC), sensitivity, and specificity, using an independent set of 17,507 test images from 93 patients, including 7507 images of protruding lesions from 73 patients. RESULTS: The developed CNN analyzed 17,507 images in 530.462 seconds. The AUC for detection of protruding lesions was 0.911 (95% confidence interval [Cl], 0.9069-0.9155). The sensitivity and specificity of the CNN were 90.7% (95% CI, 90.0%-91.4%) and 79.8% (95% CI, 79.0%-80.6%), respectively, at the optimal cut-off value of 0.317 for probability score. In a subgroup analysis of the category of protruding lesions, the sensitivities were 86.5%, 92.0%, 95.8%, 77.0%, and 94.4% for the detection of polyps, nodules, epithelial tumors, submucosal tumors, and venous structures, respectively. In individual patient analyses (n = 73), the detection rate of protruding lesions was 98.6%. CONCLUSION: We developed and tested a new computer-aided system based on a CNN to automatically detect various protruding lesions in WCE images. Patient-level analyses with larger cohorts and efforts to achieve better diagnostic performance are necessary in further studies.