DAMO Academy, the research institute of Alibaba Group, has developed an (artificial intelligence) AI-powered screening method for the early detection of pancreatic cancer, marking a significant advance enabling large-scale screening.
DAMO Academy’s deep learning algorithm identifies pancreatic lesions not easily seen on human-eye observation in non-contrast CT scans, revolutionizing pancreatic cancer imaging-based screening.
Trained on 3,200+ image sets, this model, highlighted in a recent Nature Medicine journal, achieved notable performance metrics. With a specificity of 99.9% (meaning one false positive in every 1,000 tests) and a sensitivity of 92.9%, surpassing human radiologists by 34.1% in sensitivity and 6.3% in specificity, it demonstrated superior diagnostic capabilities.
Alibaba opens access to privacy-preserving computation technologies
Alibaba Cloud deploys cloud technology innovations at Asian Games
“Early detection of pancreatic cancer is hard to realize in conventional screening, which results in late detection and poor prognosis,” said Le Lu, head of Alibaba’s Damo Academy’s medical AI team and fellow at IEEE. “The AI plus non-contrast CT technology holds the promise to be an effective and cost-efficient tool to achieve detection of pancreatic cancer in the early stages and make large-scale pancreatic cancer screening possible to prevent the loss of lives.”
AI-based screening
Collaborating with more than 10 top medical institutions, DAMO Academy’s researchers applied this AI-based screening to over 20,000 patients, uncovering 31 previously missed cases of pathological changes. Widely adopted, it has been used over 500,000 times in Chinese hospitals and medical check-ups.
According to DAMO Academy, pancreatic cancer’s low survival rates stem from late-stage detection. As the seventh leading cause of cancer-related deaths globally, its 5% to 10% average five-year survival rate underscores the urgency for early detection methods.
This technology, when paired with non-contrast CT imaging, aids in early pancreatic cancer detection, especially given its nonspecific symptoms. It can integrate into routine medical check-ups or emergency department protocols, advancing early detection efforts.
Experts like Jörg Kleeff & Ulrich Ronellenfitsch from Martin-Luther-University Halle-Wittenberg, University Medical Center Halle (Saale) in Germany, acknowledge its effectiveness over existing screening methods for other cancers. However, they advocate for comprehensive assessment before widespread implementation, recognizing its potential clinical impact while emphasizing the need for careful evaluation before broad adoption.