The European Organization for Nuclear Research (CERN) is applying its expertise in artificial intelligence (AI) to develop innovative medical applications. The organization has initiated collaborations with over 30 hospitals across Europe to implement and evaluate new AI technologies in clinical practice.
One promising system, Truckstroke, is already enhancing stroke treatment for approximately 10,000 patients in hospitals in Germany, Belgium, and the Vall d'Hebron Hospital in Barcelona. Luigi Serio, a CERN scientist overseeing AI applications in healthcare, explains that CERN offers hospitals its experience in managing vast amounts of data securely and in a decentralized manner.
This approach is crucial for ensuring the privacy and security of patient information used to 'train' algorithms. The CERN system processes data locally, preventing it from being sent to a central storage, which aids in protecting privacy and optimizing resources when different hospitals collaborate on reliable disease analysis and prediction models based on AI.
Truckstroke compares brain images of stroke patients with models trained by CERN. The algorithm predicts the patient's potential progression, suggests appropriate treatment, and necessary monitoring, and importantly, assesses the risk of recurrence.
Serio highlights that healthcare professionals often face overwhelming patient loads and require more tools to support their work. Truckstroke provides doctors with predictive models capable of measuring the severity of strokes and guiding treatment and monitoring decisions.
Another significant application being developed by CERN is a breast cancer detection program, promising to be 50% more accurate than the current GAIL model. This new model will assess the risk of developing breast cancer by combining multiple factors, including dietary habits, lifestyle, and age at first pregnancy or menopause.
CERN is also working to improve linear accelerators for radiotherapy (LINACs) using AI to make them more user-friendly and accessible in low- and middle-income countries. The STELLA project aims to enhance radiotherapy in certain African nations, where there is one radiotherapy device for every 3.5 million people, compared to one for every 80,000 to 100,000 in the U.S. and most European countries.
Finally, CERN has developed an application capable of detecting anomalies or pathologies in the brain, indicating to doctors the exact location where a pathology, such as a tumor, may be developing. This technology is currently undergoing clinical trials at the Kapodistrian University Hospital in Greece and could be used in the future to monitor the progression of Alzheimer’s or dementia.
Serio emphasizes that CERN's strength lies in its power, expertise, and reputation, along with the assurance that its services are supported by a non-profit, impartial status. CERN plans to complete the breast cancer detection program next year. Once the model is finalized, it will need to be tested and calibrated before replacing the current screening protocol.