Theme |
Pit Pattern Classification in colorectal Lesion |
Title |
Microvascular Architecture of Localized Colorectal Lesions Observed with Narrow Band Imaging (NBI) |
Author |
Yoshiki Wada |
Digestive Disease Center, Showa University Northern Yokohama Hospital |
Author |
Hiroshi Kashida |
Digestive Disease Center, Showa University Northern Yokohama Hospital |
Author |
Shin-ei Kudo |
Digestive Disease Center, Showa University Northern Yokohama Hospital |
Author |
Masashi Misawa |
Digestive Disease Center, Showa University Northern Yokohama Hospital |
Author |
Toshihisa Hosoya |
Digestive Disease Center, Showa University Northern Yokohama Hospital |
Author |
Kunihiko Wakamura |
Digestive Disease Center, Showa University Northern Yokohama Hospital |
Author |
Hiroshi Kanie |
Digestive Disease Center, Showa University Northern Yokohama Hospital |
Author |
Nobunao Ikehara |
Digestive Disease Center, Showa University Northern Yokohama Hospital |
Author |
Fuyuhiko Yamamura |
Digestive Disease Center, Showa University Northern Yokohama Hospital |
Author |
Kazuo Ohtsuka |
Digestive Disease Center, Showa University Northern Yokohama Hospital |
Author |
Shigeharu Hamatani |
Department of Pathology, Showa University Northern Yokohama Hospital |
[ Summary ] |
Recent developments in NBI enable us to observe microvascular structural changes. The diagnosis of colorectal tumors through magnified colonoscopic findings of microvascular architecture is simple and usefuL We classified colonic microvascular patterns into six categories : normal, faint, network, dense, irregular and sparse. If we assume that the faint pattern is a diagnostic indicator for hyperplastic polyps, we are able to differentiate between neoplastic and non-neoplastic lesions within a sensitivity and specificity range of 89.6 % and 98.4 %. Assuming that either irregular or sparse patterns represent massively invasive cancers, the sensitivity and specificity are 94.1 % and 88.4 %. Utilizing the NBI system is useful for distinguishing neoplastic from non-neoplastic lesions, and also for distinguishing cancers from adenomas. |