AI Innovation Benefits Hospitals in 44 Countries Worldwide

Background

Digestive endoscopy is a primary diagnostic method for detecting early gastrointestinal cancers and precancerous lesions. Variations in endoscopist skill, fatigue and procedural differences can lead to missed lesions, incomplete coverage of anatomical areas or inadequate inspection time.

At Renmin Hospital of Wuhan University, Professor Yu Honggang's team developed the ENDOANGEL artificial intelligence system following years of research involving deep learning models. The system has been evaluated in multiple randomised controlled trials published in journals such as The Lancet Gastroenterology & Hepatology, Gut and Endoscopy, demonstrating improvements in detection rates and reductions in miss rates for colorectal adenomas and gastric neoplasms.

What was achieved

According to a July 2026 report, the ENDOANGEL system is the world's first applied across all digestive endoscopy scenarios, including colonoscopy, upper gastrointestinal endoscopy and endoscopic ultrasound. It has been deployed in over 1,000 hospitals in 44 countries.

The system has assisted in the detection of more than 130,000 early cancers and precancerous lesions. The official EndoAngel platform reports assistance in more than 1,000,000 procedures worldwide. Studies associated with the development team showed it could increase adenoma detection rates (for example from 8% to 16% in one randomised trial) and reduce gastric neoplasm miss rates (from 27.3% to 6.1% in a tandem randomised trial). It received NMPA Class III certification in 2023.

How it works

During an endoscopy, a flexible tube equipped with a camera and light is passed through the mouth or anus to examine the lining of the digestive tract. The endoscopist must systematically view all relevant areas while advancing and withdrawing the instrument. ENDOANGEL connects to the endoscope video feed via an HDMI pass-through and analyses the live images in real time using deep convolutional neural networks trained on large datasets of prior endoscopic procedures.

For upper gastrointestinal endoscopy it monitors 26 anatomical stations in the stomach to identify blind spots, tracks withdrawal speed and inspection time with visual and auditory alerts, and highlights suspicious lesions with coloured boxes on screen (such as blue for focal lesions and red for those suspected to be neoplastic). In colonoscopy it supports polyp detection and characterisation. During endoscopic ultrasound it assists with anatomical landmark recognition such as pancreatic structures, bile ducts and major vessels. The system is designed to provide assistive alerts without replacing clinical judgement.

What it means for international patients

The global deployment of the technology means AI-assisted digestive endoscopy is available in numerous countries. Patients from abroad considering care at the originating hospital in Wuhan may inquire about its integration into gastroenterology services.

Renmin Hospital of Wuhan University has experience treating international patients, including cases involving English-language communication. International patients should contact the hospital directly to confirm current availability and eligibility for procedures involving the ENDOANGEL system or related gastroenterology care, as well as any arrangements for foreign visitors including language support.

About the hospital

Renmin Hospital of Wuhan University, also known as Hubei General Hospital or Hubei Provincial People's Hospital, is a large comprehensive Grade A tertiary hospital located in Wuhan, Hubei Province. Founded in 1923, it is affiliated with Wuhan University and functions as the university's First Clinical College.

The hospital operates multiple campuses and is noted for its departments of gastroenterology (where the ENDOANGEL system was developed), cardiovascular medicine, ophthalmology, minimally invasive techniques, organ transplantation and oncology. It engages in international collaborations with institutions in Europe, the United States and elsewhere and participates in medical education and training programs.