Patients with epilepsy or prostate cancer now benefit from safer or reduced surgical interventions thanks to pioneering imaging techniques based on computational models developed in UCL’s Centre for Medical Image Computing. The same technology helps bring pioneering drug treatments for neurological disorders to patients.
UCL researchers led by Professor Daniel Alexander (UCL Computer Science) developed biophysical models in the early 2000s that link signals measured by magnetic resonance imaging (MRI) scanners to the size, shape, and density of cells in tissue.
Such histological information is used universally in medicine to identify and classify disease. It is normally accessible only through invasive biopsy (extracting a tissue sample) and viewing it under a microscope, while the MRI scans are entirely harmless.
Professor Geoff Parker (UCL Medical Physics) used Alexander’s models to visualise brain connections often damaged during surgery to cure drug-resistant epilepsy. Alexander’s models reveal directions of the brain fibres that form the brain’s communication network: bundles of nerve cells, like the fibre-optic cables supporting the internet. “Tractography” algorithms use fibre directions to piece together maps of the brain’s connectivity or “wiring diagram”. Parker’s uniquely sensitive tractography enabled the BRC-supported EpiNavTM (pioneered by Professor John Duncan, at our National Hospital for Neurology and Neurosurgery, NHNN) surgical planning system to allow surgeons to avoid damaging vital brain connections while removing brain tissue that causes epilepsy. EpiNav dramatically reduces brain damage from surgery (e.g. vision problems preventing driving reduced from 15-20% to 0% of patients 1). 400 patients at NHNN have benefitted, and EpiNav is now deployed internationally.
Alexander’s biophysical models also underpin “microstructure imaging” techniques that map tissue properties using MRI. Dr Laura Panagiotaki (UCL Computer Science) developed VERDICT, which maps cell density in the prostate. Professor Shonit Punwani (UCLH) led BRC-supported clinical trials demonstrating VERDICT is the first non-invasive exam that identifies clinically significant prostate cancer, reducing by 50% the need for invasive biopsy and side-effects 2. Dr Gary Zhang (UCL Computer Science) developed NODDI 3, providing brain-cell density maps. NODDI is used internationally in neuroimaging studies and supports clinical trials (e.g. emerging treatments for Huntington’s disease, led by UCL Neurology’s Professor Sarah Tabrizi with BRC support).
These examples illustrate the BRC’s crucial translational role in enabling basic engineering/computer-science research to benefit patients in diverse ways.
1. Winston Annals Neurol. 2012; 2. Singh European Urology 2019; 3. Zhang Neuroimage 2012
