Assessment of Brain Tumours Using Machine Learning and MRI Radiomics
About the project
Brain tumours kill more children and adults under the age of 40 than any other cancer but are under-researched compared to most cancers. A priority is the need for a non-invasive, objective pre-treatment biomarker of brain tumours. Radiomics, a new and emerging field, combined with Artificial Intelligence (AI), offers the potential not only for the development of such biomarkers, but also for pre-operative prediction of tumour grading and response to treatment.
The aim of the project is to build an AI-based tool for clinicians to use to assess brain tumours to help with treatment planning and to inform clinical decision-making. This will have a significant impact on patient treatment outcome (e.g. a reduction in the number of unnecessary invasive procedures).
The project will use machine learning to correlate radiomic, radiological and histological findings in meningiomas, bench marked against the gold standard, highly predictive genomic data from meningioma tissue, in order to develop better non-invasive predictive models of tumour grade and clinical outcome. Such models would help with clinical decisions about surveillance and surveillance intervals vs surgery +/- radiotherapy.
The project will be carried out in collaboration with the Plymouth Centre of Excellence for Brain Tumour which has a biobank of brain tumour samples and associated data, including genomic and MRI data. This interdisciplinary project brings together clinical expertise in brain tumour and neuro-imaging and expertise in AI, software development, and biomedical data analysis.