Comprehensive end-to-end intelligent platform for brain tumor diagnosis, treatment and recurrence monitoring using multimodal datasets
Brain tumour
This study employs a multimodal artificial intelligence (AI) approach integrating radiological and pathological imaging data for brain tumour diagnosis and treatment support. MRI scans will be collected and anonymised for data cleaning, preprocessing, and expert annotation. Other data collected include Whole Slide Image (WSI) reports, Isocitrate Dehydrogenase (IDH) reports, and clinical, demographic, radiological, pathological, and treatment-related data available in the medical records of patients diagnosed with brain tumours. All data will be anonymised before analysis. Deep learning models will be trained using public datasets for pre-operative tumour diagnosis and postoperative recurrence prediction and validated using local clinical datasets from PPUM. Segmentation models will also be developed for tumour and blood vessel identification and integrated with a surgical navigation platform to enhance precision and safety. The outcome will be an end-to-end intelligent platform for multimodal analysis, recurrence warning, and surgical planning assistance.
1. Patients diagnosed with primary brain tumour 2. Patients who underwent standard of care procedure at University of Malaya Medical Centre (UMMC)
1. Patients diagnosed with other types of brain disease