AI Tool Revolutionizes Celiac Disease Diagnosis, Offering Faster Results and Alleviating Healthcare Strain

According to a study, artificial intelligence (AI) has the potential to speed up the diagnostic process for celiac disease. Celiac disease is an autoimmune disorder affecting almost 700,000 individuals in the UK, and it can take years to receive an accurate diagnosis.

This condition is triggered by the consumption of gluten, a protein found in wheat, rye, and barley. Symptoms may include abdominal cramps, diarrhea, skin rashes, weight loss, fatigue, and anemia.

If left untreated, celiac disease can lead to serious complications such as malnutrition, osteoporosis, anemia, infertility, and an increased risk of certain cancers and other autoimmune disorders.

Currently, the majority of adult diagnoses are made through blood tests for gluten antibodies, followed by a biopsy of the small intestine. Pathologists then examine the biopsy sample for damage to the villi—tiny, hair-like projections that line the small intestine and are essential for nutrient absorption.

Researchers from Cambridge University have created an AI-based tool that can expedite the diagnostic process and free up pathologists to focus on more complex cases. The algorithm was trained and validated using over 4,000 images from five different hospitals, each utilizing five distinct scanners from four various manufacturers.

The study, published in the New England Journal of Medicine AI, revealed that the algorithm was as effective in diagnosing celiac disease as a human pathologist. Notably, the AI model worked significantly faster than its human counterpart.

Elizabeth Suiyo, a consultant hematologist and pathology professor at Cambridge University and the senior author of the study, stated: «The process of arriving at an accurate diagnosis can often take many years, and with healthcare systems under significant strain, these delays are likely to persist. AI has the potential to streamline this process, enabling patients to receive diagnoses more swiftly while alleviating pressure on the National Health Service.»

Co-author Dr. Florian Jaekle noted that pathologists typically take between 5 to 10 minutes to analyze each biopsy, while the AI model can diagnose celiac disease almost instantly.

«Biopsies for celiac disease (and particularly celiac tests) often end up at the bottom of the list for pathologists because they are not as urgent as potential cancer cases. This means that patients frequently have to wait weeks or even months to find out if they have celiac disease,» he explained. «With AI, they could receive results almost immediately, as the model is capable of providing outputs in under a minute once a biopsy has been scanned. This means there would be no waiting time with the AI.»

The research was funded by Coeliac UK, Innovate UK, the Cambridge Centre for Data-Driven Research, and the National Institute of Health and Medical Research.

Commenting on the findings, Dr. Bernie Croal, President of the Royal College of Pathologists, remarked that this new AI tool «could radically transform the approach to diagnosing celiac disease, speeding up diagnostics, improving patient outcomes, and reducing wait times.»

«While the emergence of artificial intelligence in pathology is incredibly exciting, and the National Health Service could lead the way in the development and application of AI in this field, additional efforts will be required to ensure that AI is fully developed and safely utilized within the NHS. Investments are needed in digital pathology, integrated functional IT systems that facilitate information sharing between organizations, and training for pathologists to understand and leverage AI effectively.»

Source