AI in neuroimaging diagnosis

In silico gliomas: an exploration of artificial intelligence in neuroimaging diagnosis

Published on March 1, 2025

Gliomas are the most common type of intraparenchymal tumour, arising from the glial cells in the central nervous system (CNS)1. The World Health Organisation classifies them into four grades: grade I and II are considered low-grade gliomas (LGG) while III and IV are considered high-grade gliomas (HGG)2 due to their pathological and genetic properties3. While the diagnostic gold standard remains brain biopsies4, 5, allowing for comprehensive molecular and genetic profiling6, this invasive surgery is associated with a notable 6% incidence of procedure-related risks and complications7 as well as comorbidities patients may have that would limit them from undergoing surgery. Moreover, it often fails to capture the spatial heterogeneity inherent within these tumours.

Deep learning (DL), a subclass of machine learning (ML) within the broader field of artificial intelligence (AI), offers the potential for complex...

Back to top