The aim of this project is to develop a radiomics-based deep learning model to predict treatment outcomes and disease progression in glioblastoma multiforme (GBM). The model will be tested in pre-clinical GBM mouse models exposed to drug administration alone or drug combined with FUS treatments to disrupt the blood-tumour barrier. The trained model will be applied to historical data from GBM patients to evaluate its precision in a clinical setting. The aims of the PhD project will be the following: 1) Establish a pre-clinical GBM mouse model and characterise tumour features using MRI and PET imaging; 2) Utilise pre-clinical radiomics data to develop a deep learning model capable of predicting tumour response to treatment; and 3) Validate the deep learning model on historical GBM patient data to test model efficacy and predictive power.
Back to projects
Integrating deep learning and radiomics to predict glioblastoma response to treatment
