Computational Oncology

What we do

We focus on mathematical and computational approaches to improving healthcare, particularly translational and clinical applications of these approaches. These include using clinical "big data" to understand clinical outcomes, novel reasoning techniques to better understand clinical trials, computer-enhanced interpretation of imaging and using data from patient-worn sensors and online data collection.

Why it is important

Computational medicine offers a chance to provide mass-personalisation of medicine, improving healthcare and preventing harm. However, many of the underlying techniques and theory are often poorly developed in the medical arena, where we face problems of complex, longitudinal datasets with conflicting and incomplete data. Our work aims to offer theoretically sound, well-tested applications of computational approaches to real clinical problems. For that reason, we are based in the radiotherapy department at Charing Cross, and our Clinical Staff run several clinics each week in addition to their research work.

How it can benefit patients

We have a strong translational aspect to our work. We are currently developing patient-focused extensions to our previous work to develop patient and doctor-enabling tools to reason with clinical trials. We are also developing novel uses for patient-worn sensors in order to better understand toxicity and quality of life in patients being treated for their cancer, with the ultimate aim of improving patient outcomes.

Summary of current research

Computational Oncology

• Clinical Big Data: Using national-scale datasets to understand and improve healthcare at large scale. We are working with colleagues in Leeds on the Parachute project to look at outcomes after treatment in a national radiotherapy dataset; this builds on our earlier work with PHE on developing computational tools to help analyse such data and we are leading the Gliocova project looking at using linked national cancer datasets to understand patterns of care, treatment and outcomes in patients with glioma. We previously invented the Simulacrum - the world's largest publicly available individual patient-level dataset.

• Computational Argumentation: Using novel computational methods to understand and reason with clinical trial data. We are leading a Cochrane Systematic Review and Network Meta-analysis into first-line treatments for patients with brain metastases, and doing parallel work applying conventional and novel analytical approaches.
• Patient-generated Electronic Data: Using patient wearables and online platforms to gather high-resolution patient data at low cost. We are running the BrainWear trial that uses a patient-worn wrist-mounted accelerometer and are leading the TRIGGER project that is exploring the feasibility of collecting patient-reported outcome measures in patients who have had pelvic radiotherapy.

Clinical Oncology

• Data and information
• Accelerator development
• Technical breast radiotherapy
• Clinical trial development and leadership