| Theme | Magnifying Endoscopic Diagnosis of Gastro‒intestinal Carcinomas | |
|---|---|---|
| Title | Automatic Diagnosis System Using Endocytoscopy for Colorectal Lesions | |
| Publish Date | 2017/12 | |
| Author | Shin‒ei Kudo | Digestive Disease Center, Showa University Northern Yokohama Hospital |
| Author | Kenichi Takeda | Digestive Disease Center, Showa University Northern Yokohama Hospital |
| Author | Yuichi Mori | Digestive Disease Center, Showa University Northern Yokohama Hospital |
| Author | Masashi Misawa | Digestive Disease Center, Showa University Northern Yokohama Hospital |
| Author | Toyoki Kudo | Digestive Disease Center, Showa University Northern Yokohama Hospital |
| [ Summary ] | A precise endoscopic diagnosis is required for confirming the treatment strategy for colorectal lesions. For that reason, endoscopic diagnosis such as pit pattern classification has been established and widely performed in clinical practice. However, to properly evaluate the endoscopic findings and make a diagnosis, endoscopists need certain training. Therefore, an automated computerized system has been desired for a standardized diagnosis. High diagnostic accuracy has been reported for the computer-aided diagnosis system for endocytoscopy (EC-CAD), which is currently under development. Although EC-CAD has been reported to be useful for the diagnosis of invasive cancers, and for the differentiation between neoplasms and nonneoplasms, its diagnostic utility remains undetermined for the diagnosis of sessile serrated adenomas/polyps and inflammation of inflammatory bowel disease. In the future, study using artificial intelligence will be conducted, and it is expected that its usefulness will be demonstrated clinically. | |