Real-time assessment of video images for esophageal squamous cell carcinoma invasion depth using artificial intelligence
- PDF / 10,739,027 Bytes
- 9 Pages / 595.276 x 790.866 pts Page_size
- 10 Downloads / 197 Views
ORIGINAL ARTICLE–ALIMENTARY TRACT
Real-time assessment of video images for esophageal squamous cell carcinoma invasion depth using artificial intelligence Yusaku Shimamoto1 • Ryu Ishihara1 • Yusuke Kato2 • Ayaka Shoji1 • Takahiro Inoue1 • Katsunori Matsueda1 • Muneaki Miyake1 • Kotaro Waki1 Mitsuhiro Kono1 • Hiromu Fukuda1 • Noriko Matsuura1 • Koji Nagaike3 • Kenji Aoi4 • Katsumi Yamamoto5 • Takuya Inoue6 • Masanori Nakahara7 • Akihiro Nishihara8 • Tomohiro Tada2,9,10
•
Received: 13 May 2020 / Accepted: 2 August 2020 Ó Japanese Society of Gastroenterology 2020
Abstract Background Although optimal treatment of superficial esophageal squamous cell carcinoma (SCC) requires accurate evaluation of cancer invasion depth, the current process is rather subjective and may vary by observer. We, therefore, aimed to develop an AI system to calculate cancer invasion depth. Methods We gathered and selected 23,977 images (6857 WLI and 17,120 NBI/BLI images) of pathologically proven superficial esophageal SCC from endoscopic videos and still images of superficial esophageal SCC taken in our & Ryu Ishihara [email protected] 1
Department of Gastrointestinal Oncology, Osaka International Cancer Institute, 3-1-69 Otemae, Chuo-ku, Osaka 541-8567, Japan
2
AI Medical Service Inc., Tokyo, Japan
3
Department of Gastroenterology, Suita Municipal Hospital, Osaka, Japan
4
Department of Gastroenterology, Kaizuka City Hospital, Osaka, Japan
5
Department of Gastroenterology, Japan Community Health Care Organization Osaka Hospital, Osaka, Japan
6
Department of Gastroenterology, Osaka General Medical Center, Osaka, Japan
7
Department of Gastroenterology, Ikeda City Hospital, Osaka, Japan
8
Department of Gastroenterology, Minoh City Hospital, Osaka, Japan
9
Tada Tomohiro Institute of Gastroenterology and Proctology, Saitama, Japan
10
Department of Surgical Oncology, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan
facility, to use as a learning dataset. We annotated the images with information [such as magnified endoscopy (ME) or non-ME, pEP-LPM, pMM, pSM1, and pSM2-3 cancers] based on pathologic diagnosis of the resected specimens. We created a model using a convolutional neural network. Performance of the AI system was compared with that of invited experts who used the same validation video set, independent of the learning dataset. Results Accuracy, sensitivity, and specificity with nonmagnified endoscopy (ME) were 87%, 50%, and 99% for the AI system and 85%, 45%, 97% for the experts. Accuracy, sensitivity, and specificity with ME were 89%, 71%, and 95% for the AI system and 84%, 42%, 97% for the experts. Conclusions Most diagnostic parameters were higher when done by the AI system than by the experts. These results suggest that our AI system could potentially provide useful support during endoscopies. Keywords Esophageal cancer Squamous cell carcinoma Artificial intelligence Convolutional neural network Video image
Introduction Worldwide, esophageal cancer (EC) is the seventh most comm
Data Loading...
