Artificial intelligence-assisted diagnosis of rectal neuroendocrine tumors during white-light endoscopy - Original Article

Sep 28, 2026

Uploaded by Pentious

Artificial intelligence-assisted endoscopic diagnosis is used to address the frequent misdiagnosis of rectal neuroendocrine tumors (rNETs), which show low incidence and macroscopic appearances similar to hyperplastic polyps. A neuroendocrine tumor lesion detection algorithm based on the YOLOv7 model is developed and assessed for its ability to identify rNETs from white-light endoscopic images. In total, 137,748 images are analyzed and the model’s performance is compared with endoscopists across experience levels. The algorithm demonstrates strong diagnostic accuracy and high specificity, supporting effective rNET lesion detection.

LanguageEnglish

CategoryHealthcare

FormatPDF

Pages13

Reading Time33 min

AccessFree to download

Abstract

Background

Aim

Methods

Results

Conclusion

Key Words

| W | J | G | World Journal of
Gastroenterology | | --- | --- | --- | --- | Submit a Manuscript: [https://www.f6publishing.com](https://www.f6publishing.com) World J Gastroenterol 2025 December 28; 31(48): 112683 DOI: 10.3748/wjg.v31.i48.112683 ISSN 1007-9327 (print) ISSN 2219-2840 (online) ORIGINAL ARTICLE Retrospective Cohort Study Artificial intelligence-assisted diagnosis of rectal neuroendocrine tumors during white-light endoscopy Ke Liu, Zhen-Yu Wang, Li-Zhi Yi, Feng Li, Shun-Hui He, Xi-Gang Zhang, Chun-Xiao Lai, Zhi-Jian Li, Lin Qiu, Rui-Ya Zhang, Wen Wu, Yu Lin, Hui Yang, Gui-Ming Liu, Quan-Sheng Guan, Zhi-Fang Zhao, Li-Ming Cheng, Jie Dai, Yang Bai, Fang Xie, Meng-Nan Zhang, Su-Zuan Chen, Xian-Fei Zhong Specialty type: Gastroenterology and hepatology Provenance and peer review: Unsolicited article; Externally peer reviewed. Peer-review model: Single blind Peer-review report’s classification Scientific Quality: Grade A, Grade A, Grade B, Grade B Novelty: Grade A, Grade A, Grade B, Grade C Creativity or Innovation: Grade A, Grade B, Grade B, Grade C Scientific Significance: Grade B, Grade B, Grade B, Grade B P-Reviewer: Kehagias D, MD, PhD, Greece; Zhu HC, Professor, China Received: August 4, 2025 Revised: September 17, 2025 Accepted: November 12, 2025 Published online: December 28, 2025 Processing time: 146 Days and 10.8 Hours Ke Liu, Li-Zhi Yi, Xian-Fei Zhong, Department of Gastroenterology, The People’s Hospital of Leshan, Southwest Medical University, Leshan 614000, Sichuan Province, China Zhen-Yu Wang, Su-Zuan Chen, Department of Digestive Endoscope, The First Affiliated Hospital of Shantou University Medical College, Shantou 515000, Guangdong Province, China Feng Li, Lin Qiu, Yang Bai, Fang Xie, Meng-Nan Zhang, Department of Gastroenterology, Nanfang Hospital, Southern Medical University, Guangzhou 510515, Guangdong Province, China Shun-Hui He, Zhi-Jian Li, Department of Gastroenterology, The Eighth Affiliated Hospital of Southern Medical University, Guangzhou 510515, Guangdong Province, China Xi-Gang Zhang, Department of Gastroenterology, Shenzhen Second People’s Hospital, Shenzhen 518020, Guangdong Province, China Chun-Xiao Lai, Gastrointestinal Cancer Center, Baiyun Branch, Nanfang Hospital, Southern Medical University, Guangzhou 510515, Guangdong Province, China Rui-Ya Zhang, Department of Gastroenterology, The Fifth Clinical Medical College of Shanxi Medical University, Taiyuan 030012, Shanxi Province, China Wen Wu, Department of Gastroenterology, Shanxi Academy of Traditional Chinese Medicine, Taiyuan 030024, Shanxi Province, China Yu Lin, Department of Gastroenterology, Southern Medical University Hospital of Integrated Traditional Chinese and Western Medicine, Southern Medical University, Guangzhou 510515, Guangdong Province, China Hui Yang, Department of Spleen and Stomach, Rizhao Hospital of Traditional Chinese Medicine, Shandong University of Traditional Chinese Medicine, Rizhao 276800, Shandong Province, China Gui-Ming Liu, Quan-Sheng Guan, Department of Gastroenterology, Shayang Hospital of Traditional Chinese Medicine, Jingmen 448200, Hubei Province, China Zhi-Fang Zhao, Department of Gastroenterology, National Institution of Drug Clinical Trial, Guizhou Provincial People’s Hospital, Medical College of Guizhou University, Guiyang WJG [https://www.wjgnet.com](https://www.wjgnet.com) 1 December 28, 2025 Volume 31 Issue 48 Liu K et al. AI-assisted diagnosis of rNETs 550000, Guizhou Province, China Li-Ming Cheng, Department of Internal Medicine, Taihe People’s Hospital Baiyun District, Guangzhou 510515, Guangdong Province, China Jie Dai, Suzhou Wellomen Information Technology Co. Ltd., Suzhou 215000, Jiangsu Province, China Corresponding author: Xian-Fei Zhong, Department of Gastroenterology, The People’s Hospital of Leshan, Southwest Medical University, No. 238 White Tower Road, Leshan 614000, Sichuan Province, [China.](Chi

FAQs

Why are rectal neuroendocrine tumors often misdiagnosed during white-light endoscopy?

Their incidence is much lower than colorectal polyps and their macroscopic features resemble hyperplastic polyps, leading to frequent misdiagnosis and polyp resection.

What algorithm and data were used to detect rNET lesions?

The study established a lesion detection algorithm based on the YOLOv7 model and used 137,748 white-light endoscopic images, including labeled rNET and other lesion and normal mucosa images.

How did the YOLOv7-based model perform compared with endoscopists?

The model achieved 97.8% accuracy with 99.7% specificity and showed effective rNET lesion detection, reported as better than that of most endoscopists.

Sep 28, 2026

Uploaded by Pentious