Yasushi Oda

Research and development of artificial intelligence (AI) in the field of gastrointestinal endoscopy is progressing rapidly. In Japan alone, there are more than 10 AI‐assisted endoscopic medical devices that have received regulatory approval, and numerous randomized controlled trials have been published both domestically and internationally. However, the adoption of AI in clinical practice has not been smooth due to factors such as insufficient evaluation of the balance between clinical benefits and harms, unclear cost‐effectiveness, the lack of reliable guidelines, and the absence of established reimbursement systems for medical fees. Considering this situation, the Japan Gastroenterological Endoscopy Society (JGES) presents its perspective on the status of AI in endoscopic practice in the form of the following position statements. This comprises nine statements developed by the JGES AI Committee in collaboration with a diverse panel of members. These statements comprehensively address issues related to the quality of endoscopic examinations, cost‐effectiveness, clinical disadvantages, preparatory knowledge, medical safety, and legal responsibilities. They have been developed to be practical and useful in actual endoscopy settings.

INTRODUCTION: Index colonoscopy findings can affect surveillance decisions that have an impact on the future risk of colorectal cancer (CRC). However, the effectiveness of surveillance colonoscopy in reducing CRC risk remains unclear. The aim of this study was to evaluate whether the timing of the first surveillance colonoscopy affected CRC incidence rates. METHODS: We conducted a retrospective cohort study at 2 community endoscopy clinics between 2005 and 2021, including 20,397 individuals. Based on the findings at index colonoscopy, we categorized into 4 groups: advanced adenoma (AA), ≥3 non-AAs, 1–2 non-AAs, and no adenoma. We compared the cumulative incidence of CRC among these groups, focusing on whether the first surveillance interval was less than or greater than 3 years after index colonoscopy. RESULTS: Index colonoscopy findings identified 11,601 subjects with no adenoma, 5,288 with 1–2 non-AA, 1,484 with ≥3 non-AA, and 2,024 with AA. There were 50 cases of CRC: 23 in the no adenoma, 10 in the 1–2 non-AA, 6 in the ≥3 non-AA, and 11 in the AA group. First surveillance after 3 years was associated with a higher incidence of subsequent CRC in the AA group (hazard ratio 3.49, 95% confidence interval 1.06–11.46, P = 0.04). DISCUSSION: In the AA group, surveillance colonoscopy within 3 years was associated with a lower risk of CRC. These findings suggest that a first surveillance within 3 years is particularly effective in subjects with AA.

INTRODUCTION: This study investigated the detection rate of colorectal neuroendocrine neoplasms (NENs) using large-scale colonoscopy data. METHODS: This cross-sectional analysis used large-scale data from a Japanese multicenter observational study of colonoscopies performed from 2010 to 2020. RESULTS: Among 82,005 colonoscopy cases, colorectal NENs were identified in 71 (67 of which were neuroendocrine tumors), with a detection rate of 0.087% (95% confidence interval: 0.069–0.109). Most were small rectal lesions, with only 4 >10 mm in size and 3 located in the colon. DISCUSSION: The detection rate of colorectal NENs during colonoscopy is substantially higher than expected.

Background and Aim The impact of frequent colonoscopy on colorectal cancer (CRC) remains unclear. The present study aimed to determine the relationship between frequency of surveillance colonoscopy and CRC prevention. Methods From April 2010 to April 2011, patients who underwent surveillance colonoscopy after screening and polypectomy in four Japanese endoscopy centers were enrolled in this multicenter historical cohort study. Patients were classified into the following two groups according to the findings of past colonoscopy: a low‐risk group (no neoplasia or 1–2 cumulative adenomas <10 mm) and an increased‐risk group (advanced adenoma or ≥3 cumulative adenomas). The relationship between colonoscopy frequency within the previous 5 years and the prevalence of advanced neoplasia in each group was analyzed using multiple logistic regression. Results The final analysis included 2391 patients. In the low‐risk group, the odds ratios for advanced adenoma in patientsundergoing moderately frequent colonoscopy (2–3 times within the previous 5 years), and frequent colonoscopy (≥4 times within 5 years) were 0.33 (95% confidence interval [CI], 0.14–0.81) and 0.21 (95% CI, 0.02–1.60), respectively, compared with infrequent colonoscopy (once or not at all within 5 years). In the increased‐risk group, the respective odds ratios were 0.48 (95% CI, 0.28–0.83) and 0.28 (95% CI, 0.12–0.64). Conclusions Although frequent colonoscopy provides benefits against advanced adenoma, the optimal benefit was achieved at 2–3 times. With very frequent colonoscopy (i.e. ≥4 times within 5 years), the additional risk reduction for advanced adenoma was relatively small.