Glioblastoma is an aggressive, incurable type of brain tumour that strikes around 2,500 people every year in the UK. Only a quarter of patients survive more than a year from diagnosis.
In this project, Dr Gooya is exploring whether a new combination treatment is more effective than the current standard for patients with glioblastoma. He will also explore whether his team can use AI and MRI brain scans to create personalised tumour progress maps to forecast how each patient’s tumour may evolve.
Following rigorous assessment as part of our competitive grant round, this project was recommended for its promising approach to personalised glioblastoma treatment.
Glioblastoma is an aggressive, invasive tumour that grows and spreads quickly and infiltrates the brain. Despite many years of research, the treatments developed in the laboratory have failed to improve patient outcomes and glioblastoma remains incurable.
The current treatment strategy includes surgery to ‘debulk’ the tumour, followed by radiotherapy and chemotherapy to destroy remaining tumour. This prolongs survival but is not curative; the tumour always grows back. Only a quarter of patients survive more than a year from diagnosis, and just 5% survive five years.
A major limitation in current care is the inability to assess accurately how individual patients are responding to treatment. There is therefore an urgent need for more advanced, personalised tools that can track tumour behaviour more accurately over time, identify non-responders earlier, and help guide adaptive, patient-specific treatment strategies. This project aims to address that gap by developing a computational framework to model and understand glioblastoma dynamics at the individual level.
The results of this research will help people like Poppy and Mia's dad Richard.
The combination of radiotherapy with the PARP inhibitor ‘olaparib’ has shown promising early results in glioblastoma treatment. PARP is a protein that helps damaged cells to repair themselves, and some cancer cells rely on PARP to keep their DNA healthy. Olaparib stops PARP from repairing DNA damage, which means that the cancer cells then die. In the PARADIGM-2 phase 1 clinical trial, led by Dr Gooya’s team, this approach was found to be extremely well tolerated by patients, with early indications of improved overall survival and ‘progression free’ survival compared to historical controls receiving standard chemoradiation treatment. Building on the success of PARADIGM-2, the current project will develop advanced computational models to quantify tumour behaviour more precisely, forecast patient-specific outcomes, and identify areas of likely recurrence. This will provide a more accurate and personalised assessment of treatment efficacy, helping to guide future clinical decisions and optimise the use of olaparib in glioblastoma care.
Dr Gooya and his team will address these research questions by developing an advanced AI-based model that simulates how glioblastoma tumours grow and respond to treatment. This model is built on biophysical principles, meaning it uses knowledge of how tumours actually behave in the body—how they spread, grow, and react to therapy. By applying this model to MRI scans and clinical data, the team can create personalised “tumour density maps” for each patient. These maps show where the tumour is most active and how it changes over time. This detailed, patient-specific mapping goes beyond standard methods and will allow the team to detect early signs of the success or failure of treatment. Ultimately, this approach is intended to offer a more accurate and interpretable way to evaluate and personalise treatment for brain tumour patients.
Glioblastoma is a devastating brain tumour, with average survival of just 12 to 18 months.
In this project the team will add to existing knowledge by providing a new, reliable way to assess early and accurately how glioblastoma patients respond to treatment. This is vital for a cancer with poor outcomes and high recurrence rates. Longer term, by deploying the model on the Tessa Jowell BRAIN MATRIX platform, the project could benefit many more patients across the UK.
The BIOGRAPH project brings together a multidisciplinary team with the expertise needed to successfully deliver its aims.
Dr Ali Gooya is a Senior Lecturer in Machine Learning at the University of Glasgow. He will lead the technical development of the AI model, drawing on his extensive experience in medical image analysis and tumour growth modelling.
Dr Gooya will be supported by three experienced co-PIs: Professor Anthony Chalmers is a leading clinical oncologist and Chief Investigator of the PARADIGM-2 trial. He brings essential clinical insight and has secured all necessary approvals for data access. Dr Gerry Thompson is a Neuroradiologist at the University of Edinburgh, and will ensure the imaging outputs are clinically meaningful and valid. Mrs Aoife Williamson is Consultant Therapy Radiographer, and will coordinate secure access to the imaging data.
Finally, a dedicated post-doctoral researcher will be appointed to support the technical implementation and delivery of the project under Dr Gooya’s supervision.
The team is further supported by collaboration with NVIDIA, who will advise on best practices for high-performance AI development.
Find out about our other research in this area:
