Loading use case index…
Loading use case index…
AI use case
Zhejiang Cancer Hospital and Alibaba DAMO Academy developed DAMO GRAPE, the first non-contrast CT AI for gastric cancer screening, achieving 85.1% sensitivity and 96.8% specific…
Core facts from this catalog record. Primary narrative lives in the hero above; full raw fields follow in the next section.
Every column from the source row, in stable order. URLs open in a new tab.
Title
DAMO GRAPE AI screens for gastric cancer on non-contrast CT across 20 centres, detecting cancer 2-10 months earlier
Content
On June 24, 2025, a related paper was published in Nature Medicine, announcing DAMO GRAPE, the world’s first AI model to use non-contrast CT to identify early gastric cancer. In a multi-centre clinical study across 20 centres involving close to 100,000 people, it showed meaningful improvement in gastric cancer detection rates and assisted imaging physicians in improving sensitivity. Zhejiang Cancer Hospital and Alibaba DAMO Academy jointly developed this work. The jointly developed team overcame challenges of complex gastric morphology, internal content interference, and early-stage lesions limited to the mucosal layer, building the world’s largest multi-centre dataset of non-contrast CT gastric cancer imaging (6,720 cases), from which DAMO GRAPE was developed. DAMO GRAPE achieved sensitivity and specificity of 85.1% and 96.8%, representing a 21.8% improvement in sensitivity and 14.0% improvement in specificity over human radiologists. In a specific case shared by first co-author Hu Can, a gastric surgery PhD at Zhejiang Cancer Hospital, a patient was diagnosed in April 2024 with locally advanced gastric cancer. When the team retrospectively fed that patient’s earlier non-contrast CT (taken for other reasons) into DAMO GRAPE, the AI flagged signs of early gastric cancer six months earlier. Had AI-based automatic detection been available then, the patient would have been diagnosed and treated six months earlier. The retrospective study reviewed CT images from 11 patients taken before their diagnoses and found that the AI detected gastric cancer 2 to 10 months earlier than radiologists. DAMO GRAPE has been deployed in Zhejiang, Anhui, and other provinces. In simulated screening trials at two regional hospitals, the non-contrast CT + AI gastric cancer detection rate reached as high as 24.5%. Approximately 40% of detected gastric cancer patients were originally asymptomatic. DAMO GRAPE follows DAMO PANDA (pancreatic cancer screening AI on non-contrast CT), which was published in Nature Medicine in November 2023 and received FDA Breakthrough Device designation. PANDA tested on 20,000+ CT examinations found 31 cases of pancreatic cancer previously missed clinically, with 2 patients receiving successful early surgery.
Continue exploring AI deployments in the catalog.
Back to use casesCity
Hangzhou
Company/Organization
Alibaba DAMO Academy
Continent
Asia
Country
China
Category
Health Care Providers & Services
Type
Deployment
Id
1dd5dc3a-2bb4-434b-8e7f-1b9d5c1ca001
Created At
2026-08-30T01:43:09.43326+00:00