Validation of a novel endoscopic program for measuring the size of gastrointestinal lesions

Abstract

Background

Valid and reliable methods for measurement of lesion sizes during endoscopy have not been established.

Methods

We developed a novel software program (ENDOVER; Poinbionics, Seoul, Republic of Korea) to measure lesions sizes and assessed its validity and reliability. To validate the program, we measured standard coin sizes and estimated measurement errors. To assess program reliability, 32 pictures of endoscopically resected specimens were used to measure tissue sizes six times in a 24-h interval by two examiners. Intraclass correlation coefficients (ICCs) were used to assess intraobserver and interobserver agreements. Agreement between the program and pathological measurements was assessed by absolute differences.

Results

The nominal standard sizes of 10-won, 100-won, and 500-won coins were 18.0, 24.0, and 26.5 mm, and the calculated sizes were 18.09, 24.48, and 26.31 mm, respectively. ICCs of the long and short diameters were, respectively, 0.92 [95% confidence interval (CI) 0.87–0.95] and 0.93 (CI 0.89–0.96) for examiner 1 and 0.88 (CI 0.81–0.93) and 0.92 (CI 0.87–0.95) for examiner 2. Interobserver ICCs of the long and short diameters were 0.97 (CI 0.94–0.99) and 0.97 (CI 0.94–0.99), respectively. The mean absolute differences between the program and pathological measurements were 4.4 and 4.7 mm for the long and short diameters, respectively.

Conclusions

Our findings indicate that the novel measurement program is a valid and reliable method for estimation of lesion sizes from endoscopic findings during and after examination.

http://ift.tt/2oHJnx6