L G Lim 1 , M Bajbouj , S von Delius , A Meining . Show Affiliations »
Abstract
BACKGROUND: Autofluorescence imaging (AFI) is sensitive but not specific for differentiating neoplastic from non-neoplastic colorectal polyps. We aimed to determine the sensitivity and specificity of fluorescein-enhanced AFI (FAFI) in differentiating neoplastic from non-neoplastic colorectal polyps. METHODS: All patients with colorectal polyps detected during AFI colonoscopy received intravenous fluorescein followed by AFI (FAFI). The video sequences were recorded and divided into a learning group and a test group. AFI and FAFI criteria for neoplastic and non-neoplastic lesions were determined after viewing videos in the learning group unblinded to histology. Videos in the test group were viewed blinded to histology, and diagnoses of neoplastic versus non-neoplastic were made for AFI and FAFI using the predetermined criteria. Still frames were objectively measured for red:green ratio (AFI) and green contrast (FAFI). RESULTS: Eight videos (four neoplastic, four non-neoplastic) were used for the learning group. Criteria for neoplasia when using FAFI were determined as the presence of a patchy or granular pattern which appeared more fluorescent green compared with the background. For AFI, purple or pink represented neoplasia; green represented non-neoplasia. In the test group (13 neoplastic, 12 non-neoplastic), for differentiating between neoplasia and non-neoplasia, subjective analysis of video sequences yielded a sensitivity of 100 % for AFI and 100 % for FAFI ( P = 1.000), and a specificity of 16.7 % for AFI and 91.7 % for FAFI ( P = 0.004). Using objective color analysis, the area under the receiver operating characteristics curve was 0.647 for AFI using the red:green ratio to distinguish between neoplasia and non-neoplasia, and 0.994 for FAFI using green contrast. CONCLUSIONS: FAFI accurately differentiated between neoplastic and non-neoplastic colorectal polyps. © Georg Thieme Verlag KG Stuttgart · New York.
BACKGROUND: Autofluorescence imaging (AFI) is sensitive but not specific for differentiating neoplastic from non-neoplastic colorectal polyps . We aimed to determine the sensitivity and specificity of fluorescein -enhanced AFI (FAFI) in differentiating neoplastic from non-neoplastic colorectal polyps . METHODS: All patients with colorectal polyps detected during AFI colonoscopy received intravenous fluorescein followed by AFI (FAFI). The video sequences were recorded and divided into a learning group and a test group. AFI and FAFI criteria for neoplastic and non-neoplastic lesions were determined after viewing videos in the learning group unblinded to histology. Videos in the test group were viewed blinded to histology, and diagnoses of neoplastic versus non-neoplastic were made for AFI and FAFI using the predetermined criteria. Still frames were objectively measured for red:green ratio (AFI) and green contrast (FAFI). RESULTS: Eight videos (four neoplastic, four non-neoplastic) were used for the learning group. Criteria for neoplasia when using FAFI were determined as the presence of a patchy or granular pattern which appeared more fluorescent green compared with the background. For AFI, purple or pink represented neoplasia ; green represented non-neoplasia . In the test group (13 neoplastic, 12 non-neoplastic), for differentiating between neoplasia and non-neoplasia , subjective analysis of video sequences yielded a sensitivity of 100 % for AFI and 100 % for FAFI ( P = 1.000), and a specificity of 16.7 % for AFI and 91.7 % for FAFI ( P = 0.004). Using objective color analysis, the area under the receiver operating characteristics curve was 0.647 for AFI using the red:green ratio to distinguish between neoplasia and non-neoplasia , and 0.994 for FAFI using green contrast. CONCLUSIONS: FAFI accurately differentiated between neoplastic and non-neoplastic colorectal polyps . © Georg Thieme Verlag KG Stuttgart · New York.
Entities: Chemical
Disease
Species
Mesh: See more »
Substances: See more »
Year: 2011
PMID: 21360422 DOI: 10.1055/s-0030-1256215
Source DB: PubMed Journal: Endoscopy ISSN: 0013-726X Impact factor: 10.093