Startups & Business News

Alibaba’s GRAPE AI: A Game Changer in Stomach Cancer Detection

futureTEKnow
July 2, 2025

Stomach cancer has long been a silent threat—often symptomless until it’s too late. In China alone, it claims around 260,000 lives each year, and globally it remains the fourth leading cause of cancer-related deaths. Early detection is vital: catching gastric cancer early can boost five-year survival rates from under 30% to over 90%. Yet, the gold-standard diagnostic tool—endoscopy—is invasive and unpopular, with less than 30% of at-risk patients in China opting for the procedure.

Enter GRAPE, the world’s first AI-powered tool for early stomach cancer detection, developed by Alibaba’s DAMO Academy in partnership with Zhejiang Cancer Hospital. What sets GRAPE apart isn’t just its cutting-edge technology, but its promise to democratize and simplify screening for a disease that’s notoriously hard to catch early.

How GRAPE Works

GRAPE (Gastric Cancer Risk Assessment Procedure with Artificial Intelligence) leverages deep learning to analyze routine, non-contrast CT scans—something previously considered inadequate for early gastric cancer detection. The AI model was trained on the world’s largest dataset of gastric cancer CT images, allowing it to spot subtle patterns and early lesions that even experienced radiologists might miss.

The process is straightforward:

The AI first identifies the stomach region in a CT image.
It then segments the area and scans for tumors, simultaneously classifying whether the patient has gastric cancer.
GRAPE’s predictions are visually mapped, enabling radiologists to quickly interpret results and take action.

Outperforming Human Experts

In clinical trials involving over 70,000 patients, GRAPE achieved 85.1% sensitivity and 96.8% specificity—outperforming radiologists by 21.8% and 14%, respectively. These numbers are significant: higher sensitivity means more cancers caught early, and higher specificity means fewer false alarms.

Perhaps most impressively, in real-world hospital deployments, GRAPE identified high-risk individuals at rates up to 24.5% in some regions, with about 40% of detected cases showing no previous abdominal symptoms. In some instances, the AI flagged tumors months before clinical diagnosis, highlighting its potential to save lives by catching cancer earlier than ever before.

Why This Matters

Traditional endoscopic screening is invasive, costly, and often avoided. GRAPE’s non-invasive, CT-based approach could be a breakthrough for countries and regions where endoscopic resources are limited or where patient compliance is low. By offering a scalable, less daunting screening method, GRAPE could drive up early detection rates not just in China, but worldwide.

Limitations and Next Steps

While GRAPE’s performance is remarkable, it’s not flawless. Its sensitivity for the earliest-stage cancers (T1) is about 50%, reflecting the ongoing challenge of detecting very small or subtle lesions on CT scans. Endoscopy remains superior for picking up these early changes, but GRAPE’s role as a first-line, non-invasive screening tool is clear.

Researchers are calling for larger, prospective trials to further validate GRAPE’s real-world impact and refine its capabilities, especially for those elusive early-stage cancers.

Alibaba’s GRAPE AI represents a leap forward in the fight against stomach cancer, offering a non-invasive, highly accurate screening alternative that could transform global cancer detection strategies. As the technology matures and expands, it’s poised to make early diagnosis—and better outcomes—accessible to millions more patients.

futureTEKnow is a leading source for Technology, Startups, and Business News, spotlighting the most innovative companies and breakthrough trends in emerging tech sectors like Artificial Intelligence (AI), Robotics, and the Space Industry.

Become a contributor

Find out how

Discover the companies and startups shaping tomorrow — explore the future of technology today.