Understanding and Improving Meningioma Diagnosis with AI.

Meningiomas are the most common type of brain tumour in adults. Traditionally, doctors classify these tumours based on their appearance under a microscope, which helps predict their behavior, such as how aggressive they might be and the likelihood of recurrence after treatment.

Recent research has shown that analysing the tumour DNA provides a more accurate prediction of its behaviour. Two key features have been identified: the loss of part of chromosome 1 (known as 1p deletion) and the DNA methylation pattern, which acts like a unique barcode on the DNA. However, these molecular tests are expensive and not always available in routine medical practice.

Our project aims to explore whether artificial intelligence (AI) can analyse tumour images to predict these molecular changes without the need for costly specialised lab tests. If successful, this approach could make advanced tumour testing more accessible and affordable, helping doctors make better treatment decisions for meningioma patients.