The Usefulness of a Suspected Blood Identification System (SBIS) in Capsule Endoscopy according to Various Small Bowel Bleeding Lesions.
- Author:
Ju Young KIM
1
;
Hoon Jai CHUN
;
Chul Young KIM
;
Jin Su JANG
;
Yong Dae KWON
;
Sanghoon PARK
;
Bora KEUM
;
Yeon Seok SEO
;
Yong Sik KIM
;
Yoon Tae JEEN
;
Hong Sik LEE
;
Soon Ho UM
;
Sang Woo LEE
;
Jai Hyun CHOI
;
Chang Duck KIM
;
Ho Sang RYU
Author Information
1. Department of Internal Medicine, Institute of Digestive Disease and Nutrition, Korea University College of Medicine, Seoul, Korea. kumcge@chol.com
- Publication Type:Original Article
- Keywords:
Capsule endoscopy;
Suspected blood identification system;
Obscure overt gastrointestinal bleeding
- MeSH:
Angiodysplasia;
Capsule Endoscopy;
Hemorrhage;
Humans;
Mass Screening;
Ulcer
- From:Korean Journal of Gastrointestinal Endoscopy
2008;37(4):253-258
- CountryRepublic of Korea
- Language:Korean
-
Abstract:
BACKGROUND/AIMS: Substantial time and attention are required to read and interpret the recordings of capsule endoscopic images. A suspected blood identification system (SBIS) has been developed to assist in the reading of capsule images. This software automatically marks "red tags" that correlate with suspected blood or red areas. However, the sensitivity and accuracy of the system have not been well characterized. We investigated the usefulness of the SBIS in capsule endoscopy according to various small bowel bleeding lesions. METHODS: Two expert endoscopists reviewed the capsule images. Angiodysplasias, ulcers and erosion were considered as significant lesions, and active bleeding lesions were considered when bleeding or blood clots were seen in the capsule images. The red tags that were automatically marked by the use of the rapid software were compared to the significant lesions reviewed by the endoscopists. RESULTS: A total of 95 patients were enrolled in the study. The endoscopists identified 159 significant lesions and 71 lesions marked by red tags were identified by the SBIS. Among the 71 lesions, 31 lesions correctly coincided with the significant lesions. The overall sensitivity and positive predictive value of the use of the SBIS were 20% and 44%, respectively. The sensitivities of active ulcers and active bleeding lesions were 83% and 93%, respectively. CONCLUSIONS: The SBIS should be considered as a rapid screening tool to identify active bleeding lesions, and a complete review of capsule images by a physician is still needed.