Objective To explore the auxillary diagnosis value of artificial intelligence colonoscopy (colonoscopy AI) in intestinal adenoma. Methods A total of 512 patients with suspected intestinal adenoma were selected as the study subjects and underwent general colonoscopy, colonoscopy AI-aided diagnosis, and pathological tissue biopsy. The efficacy of general colonoscopy and colonoscopy AI combined with general colonoscopy in the diagnosis of intestinal adenoma was compared. Results Among the 512 patients with suspected intestinal adenoma, 445 patients were found to have intestinal adenoma by biopsy. The sensitivity, specificity, and accuracy of colonoscopy AI combined with general colonoscopy in the diagnosis of intestinal adenoma were higher than those of general colonoscopy (97.98% vs. 90.34%, 92.54% vs. 61.19%, 97.27% vs. 86.52%, all P<0.05). A total of 428 patients′ pathological classifications of intestinal adenoma by colonoscopy AI combined with general colonoscopy were consistent with those by pathological tissue biopsy, 390 patients′ by general colonoscopy were consistent with those by pathological tissue biopsy (96.18% vs. 87.64%, P<0.05). Among the 445 patients with intestinal adenoma, 19 cases had malignant intestinal adenoma according to pathological tissue biopsy, and the specificity and accuracy of colonoscopy AI combined with general colonoscopy in diagnosing malignant intestinal adenoma were higher than those of general colonoscopy (96.95% vs. 93.66%, 95.96% vs. 91.91%, all P<0.05). Conclusion In the diagnosis of intestinal adenoma, colonoscopy AI combined with general colonoscopy has higher sensitivity, specificity, and accuracy, and its ability to identify the pathological classification of intestinal adenoma is greater when compared with general colonoscopy, which can improve the detection rate of malignant intestinal adenoma lesions and have high clinical application value.

无