Researchers at the University of Cambridge have created an artificial intelligence (AI) tool that can diagnose coeliac disease as accurately as human doctors. The key difference is speed. While pathologists take several minutes to examine a biopsy, this AI can deliver results in less than a minute. This breakthrough could make testing faster and easier for patients.
Understanding Coeliac Disease
Coeliac disease is an autoimmune disorder triggered by gluten, a protein found in wheat, barley, and rye. When people with coeliac disease consume gluten, their immune system attacks the small intestine, damaging the villi—tiny structures that absorb nutrients. Over time, this can lead to serious health problems.
The disease affects around 700,000 people in the UK and millions worldwide. Symptoms can include diarrhoea, stomach pain, fatigue, skin rashes, and anaemia. If left untreated, it can cause malnutrition, infertility, and even increase the risk of certain cancers.
Why Current Diagnosis Is Slow
Doctors use a two-step process to diagnose coeliac disease. First, a blood test checks for antibodies linked to the disease. If the test is positive, the next step is a biopsy of the small intestine. A pathologist then examines the tissue sample under a microscope to check for damage to the villi.
This process can take time. Pathologists often take 5–10 minutes per case, and since coeliac disease is not life-threatening, testing may be delayed while urgent cases take priority. Many patients wait weeks or months for a diagnosis.
AI Reduces Waiting Time
The new AI tool aims to change that. Researchers tested it on more than 4,000 biopsy images from five different hospitals. The results showed that the AI could match human pathologists in accuracy but work much faster. Instead of waiting weeks for results, patients could get answers on the same day.
This advancement could make a huge difference for people who suspect they have coeliac disease. A quicker diagnosis means they can start a gluten-free diet sooner, reducing symptoms and preventing long-term health problems.
Potential to Help the NHS
The research was led by Professor Elizabeth Soilleux and Dr Florian Jaeckle. They believe this AI tool could ease the pressure on the UK’s National Health Service (NHS). With faster diagnosis, pathologists would have more time to focus on urgent cases like cancer detection. This could improve overall efficiency in hospitals and benefit many patients.
The study was published in The New England Journal of Medicine AI, a leading journal focused on artificial intelligence in healthcare.
Challenges and Next Steps
While this discovery is promising, experts say there are challenges before AI can be used widely in hospitals. The NHS would need to invest in better digital infrastructure, including high-quality scanners to digitize biopsy samples. Doctors and lab staff would also need training to use the AI tool effectively.
Another concern is ensuring that AI is used safely. Even though it performs well in tests, doctors will still need to oversee the process to catch any errors and ensure the tool works as expected in real-world settings.
The Future of AI in Healthcare
AI is already making an impact in medicine. In recent years, AI tools have helped detect diseases like cancer, diabetic eye disease, and heart conditions. With continued research and investment, AI could improve diagnosis and treatment for many conditions.
For coeliac disease, the new AI tool could be a game changer. Faster diagnosis means less suffering, quicker treatment, and a better quality of life for patients. If implemented correctly, it could transform how the disease is detected and managed worldwide.