Introduction
For early gastrointestinal lesions, size is one of the most important factors in the selection of treatment. Lesion size is generally estimated visually, but there are errors in the size estimated visually, which may cause problems in reproducibility and objectivity. Virtual Scale Endoscope (VSE) is newly developed endoscope that displays a virtual scale using the principle of triangulation, and has been reported to measure the size of polyps in a simulated colorectal phantom more accurately than visual estimate. However, it is unknown whether VSE can accurately measure the size of the actual gastrointestinal lesions with various sizes and macroscopic types.
Aims & Methods
The aim of this study was to investigate whether the size of the resected lesion measured by VSE is close to the true size. We prospectively enrolled patients who had early gastrointestinal lesions of 20 mm or less visually at the preoperative endoscopic examination. Finally, all lesions were performed en bloc resection endoscopically. "The size of the lesion visually estimated in vivo through endoscopy before treatment in real time (visual size)", "The size of the resected lesion measured by VSE (VSE size)", and "The size of the resected lesion measured by ruler (ruler size)" were evaluated for each lesion. The primary endpoint was the difference between "VSE size" and the true size, which was defined as “VSE error”. The secondary endpoints were the difference between "visual size" and the true size, which was defined as “visual error”, and the comparison of the variation between “VSE error” and “visual error”. The difference between the two groups was calculated as [100 x (estimated or measured size - true size) / true size] (%). The true size was defined as "ruler size". Variation between “VSE error” and “visual error” was analyzed by the F-test.
Results
60 lesions were enrolled at two centers from April 2022 to December 2022. There were 20 lesions each in the esophagus, stomach, and colon. The mean lesion size ± standard deviation (SD) was 14.0 ± 6.3 mm. Macroscopic types were protruded type in 8 (13.3%) lesions, slightly elevated type in 24 (40.0%), and flat or slightly depressed type in 28(46.7%). 21 (35%) lesions were measured by expert and 39 (65%) were measured by non-expert before treatment. The mean “VSE error” ± SD was 0.3 ± 8.8 % and the mean “visual error” ± SD was -1.7 ± 29.3 %. And, “VSE error” were significantly less variable than “visual error” (p<0.001).
Conclusion
The measurement of lesion size by VSE is accurate and acceptable because it is closer to the true size and has less variation than visual estimation by endoscopists.
Disclosure
Virtual Scale Endoscope is provided by FUJIFILM for this study.
