In China, 57-year-old worker Qiu Sijun returned home after what he believed was a routine diabetes check-up. However, three days later, a doctor from the hospital called him back for additional tests. Naturally, the unexpected call caused anxiety, and his fears soon proved justified.
AI Detects Cancer Before Symptoms Appear
Further investigations revealed that Qiu Sijun had pancreatic cancer. Crucially, doctors detected the disease at a very early stage—an uncommon occurrence with this type of cancer. This early diagnosis was possible because the hospital used an artificial intelligence (AI)-based disease detection system. Acting swiftly, surgeons removed the tumour, giving the patient a strong chance of recovery. Notably, the AI tool identified the cancer even before any symptoms appeared.
Why Early Detection of Pancreatic Cancer Matters
Pancreatic cancer remains one of the deadliest forms of cancer worldwide. Only about 10% of patients survive beyond five years, largely because doctors often diagnose the disease at an advanced stage. Symptoms typically emerge late, when treatment options become limited. Even Apple co-founder Steve Jobs succumbed to this aggressive cancer.
While researchers in the US, the UK, and China have been developing AI tools to detect pancreatic cancer using CT scans, MRI images, blood test patterns, and medical records, this case stands out. Detecting pancreatic cancer through routine diabetes-related data represents a significant breakthrough.
What Made This Case Remarkable
Several key factors set this case apart. First, the patient visited the hospital only for a diabetes check-up and showed no cancer-related symptoms. Next, the AI system identified unusual patterns in his earlier medical data. As a result, doctors detected the cancer at a very early stage, making surgical treatment possible and effective.
AI Succeeds Where Routine Tests Fall Short
Today, Qiu Sijun has fully recovered and spends his time growing vegetables on his farm. Although he admits he knows little about artificial intelligence, he firmly believes the early warning saved his life.
This case clearly demonstrates how Chinese hospitals and technology companies are using AI to address some of the most difficult challenges in cancer detection and treatment.
Limitations of Conventional Pancreatic Cancer Screening
Typically, pancreatic cancer symptoms appear only when the disease has already progressed. Doctors rely on contrast CT scans to diagnose it, but these scans involve high radiation exposure. Due to the associated risks, large-scale screening is not recommended. On the other hand, low-radiation tests such as non-contrast CT scans often fail to detect early abnormalities, making diagnosis even more challenging.
How AI Is Transforming Diagnosis
AI is now changing this landscape by identifying subtle changes in imaging data that the human eye often misses. At China’s People’s Hospital, doctors use an AI tool called PANDA—short for Pancreatic Cancer Detection with Artificial Intelligence. This system detects pancreatic cancer using non-contrast CT scans, which doctors traditionally overlook.
Clinical Trials Show Promising Results
Since November 2024, PANDA has been under clinical trials at the People’s Hospital affiliated with Ningbo University. So far, the system has analysed more than 180,000 abdominal and chest CT scans. Using this data, it detected around 24 pancreatic cancer cases, including 14 at an early stage. In addition, the tool identified intraductal adenocarcinoma—the most aggressive form of pancreatic cancer—in 20 patients.
Interestingly, many of these patients visited the hospital for common complaints such as bloating or nausea and had never consulted a pancreas specialist.
Strong Accuracy Backed by Research
As reported by Bhaskarenglish, scientists from Damo Academy, a research institute linked to Chinese technology company Alibaba, developed the PANDA system. To train the AI, researchers studied thousands of past CT scans. They first marked tumour locations on contrast CT images and then matched them with non-contrast scans. This process enabled the AI to recognise cancer even in less detailed images.
Research published in 2023 showed that PANDA correctly identified pancreatic cancer in 93% of cases. According to Alibaba, the US Food and Drug Administration has granted the tool “Breakthrough Device” status, fast-tracking its approval process. Meanwhile, several additional clinical trials are underway in China as efforts continue to bring the technology into wider clinical use.
A Glimpse Into the Future of Cancer Care
Although doctors emphasise that AI cannot replace experienced clinicians, this case clearly illustrates its potential as a powerful support tool. By enabling early detection, AI-driven systems like PANDA may significantly improve survival outcomes for one of the world’s most lethal cancers.