蓝光成像放大内镜下JNET分型对结直肠息肉样病变的诊断价值

The value of JNET classification under blue-laser imaging magnifying endoscope in the diagnosis of colorectal polypoid lesions

目的:评估蓝光成像放大内镜(ME-BLI)下JNET分型对结直肠息肉样病变诊断价值。方法:选择2020年5月—2021年4月接受结肠镜检查的141例患者共检出173处结直肠息肉样病变,行ME-BLI观察并留取清晰图像,基于JNET分型给出病理预测,以病理检查为金标准评估JNET分型诊断效能,根据医师的BLI内镜经验分为经验组和普通组,并比较不同经验内镜医师对JNET分型诊断效能的影响。结果:ME-BLI下JNET分型诊断结直肠息肉样病变与病理诊断一致性检验Kappa值为0.75(P<0.01),一致性较好;经验组与普通组1型诊断准确率分别为95.4%、91.3%(P=0.131);2A型诊断准确率分别为86.7%,81.5%(P=0.186);2B型诊断准确率分别为89.0%、87.9%(P=0.737);3型诊断准确率分别97.8%、98.8%(P=0.410)。经验组与普通组总体诊断准确率比较,差异无统计学意义(P=0.262)。结论:ME-BLI基于JNET分型对结直肠息肉样病变的病理诊断有较好的预测价值,医师经验不是影响JNET分型对结直肠息肉样病变诊断准确性的因素。

Objective: To evaluate the value of blue-laser imaging magnifying endoscope(ME-BLI) based on Japan NBI Expert Team(JNET) classification in the diagnosis of colorectal polypoid lesions. Methods: A total of 173 colorectal polypoid lesions were detected in 141 patients who underwent colonoscopy from May 2020 to April 2021. All lesions were observed under ME-BLI and clear images were taken. The pathological prediction was given based on JNET classification, and the diagnostic efficiency of JNET classification was evaluated according to the gold standard of pathological examination. According to the experience of BLI endoscopy, the patients were divided into experience group and common group, and the effects of different experience endoscopes on the diagnostic efficiency of JNET classification were compared. Results: The consistency test between JNET classification and pathological diagnosis of colorectal polypoid lesions under ME-BLI showed that the Kappa value was 0.75(P<0.01). The diagnostic accuracy of type 1 in the experience group and the general group was 95.4% and 91.3% respectively(P=0.131). The diagnostic accuracy of type 2 A was 86.7% and 81.5% respectively(P=0.186). The diagnostic accuracy of type 2 B was 89.0% and 87.9% respectively(P=0.737). The diagnostic accuracy of type 3 was 97.8% and 98.8% respectively(P=0.410). Conclusion: ME-BLI based on JNET classification has good predictive value in pathological diagnosis of colorectal polypoid lesions, and physician experience is not a factor affecting the accuracy of JNET classification in the diagnosis of colorectal Polypoid lesions.