Japan's Leadership in GI Endoscopy AI: From Polyp Detection to Early Cancer Screening

Japan is the world leader in gastrointestinal (GI) endoscopy AI research. With one of the highest rates of gastric and colorectal cancer, a mature national screening culture and decades of endoscopic expertise, Japan has produced some of the most influential studies on computer-aided detection (CAD) in gastroscopy and colonoscopy.

Why Japan leads

Three factors explain Japan's position:

• Disease burden and screening culture: gastric cancer and colorectal cancer are major health problems, and endoscopic screening is widely accepted and publicly supported;
• High-quality annotated data: experienced endoscopists and established image archives provide the large, expert-annotated datasets that deep learning models need;
• Industry integration: major endoscope manufacturers such as Olympus and Fujifilm have integrated AI assistance into their systems, accelerating clinical availability.

Key research directions

Japanese research spans the whole endoscopy workflow:

• Polyp detection in colonoscopy: real-time CAD systems that highlight suspected polyps during scope withdrawal, helping reduce missed adenomas;
• Gastric cancer detection: AI that flags suspicious gastric lesions, including early-stage cancer, and assists with image-enhanced endoscopy such as narrow-band imaging (NBI);
• Quality control: automated measurement of withdrawal time, bowel preparation scoring and examination completeness — the same QC indicators that matter in China.

Clinical reality

Despite the research volume, Japanese clinical practice emphasizes that CAD is an assistive tool: the physician remains responsible for diagnosis, and AI findings are integrated into the examination flow without disrupting it. Published studies consistently report higher sensitivity with CAD, while stressing that specificity and workflow integration determine real-world value.

Lessons for China

Japan's experience offers three lessons for China's endoscopy AI:

1. Screening culture drives data: national screening programs create the datasets and clinical scenarios that make AI meaningful;
2. AI must fit the workflow: real-time, non-disruptive assistance beats isolated tools;
3. Quality control is part of the product: QC automation turns AI from a detection aid into a management tool.

yoyolink shares this philosophy: dedicated deep-learning models for gastroscopy and colonoscopy, automatic mode switching, structured image-text reports and automated QC statistics — all without AI hallucination.

• Product overview — how yoyolink works
• Clinical applications — screening and QC scenarios
• Contact us — book a demo

Automating the quality loop

Japan's strong QC culture relies on manual registration of withdrawal time, bowel preparation scores and completeness. yoyolink automates these indicators and adds patented reporting support: CN112233752A/B covers criticality assessment, report print priority and physician matching, while CN112714233B secures image transmission with block-based encryption and quality detection — fitting the strict integration and security expectations of Japanese-style endoscopy centers.

Patents & IP

Basemed Medical protects its technology with granted invention patents:

• "Information Processing Method and Device for Endoscopic Examination Report" (CN112233752A/B, co-filed with Zhejiang University): intelligently processes endoscopy reports — detects lesions, assesses criticality, prioritizes report printing, and matches patients with suitable physicians via neural networks. View | PDF
• "Method and System for Intelligent Transmission of Endoscopic Images Based on Block Decoding" (CN112714233B): transmits endoscopic images in blocks with multi-layer key information and quality detection, improving transmission quality, security and efficiency. View | PDF

See the full portfolio: About · Patents & IP