I have an over-arching interest in the application of mathematical and computational solutions to clinical problems.
Since moving to Imperial, these have been increasingly focussed on neuro-oncology and some of the particular problems experienced by patients with brain tumours. I work closely with the John Fulcher neuro-oncology lab, and am part of the neuro-oncology research group which spans Imperial College and the associated hospitals from Imperial College NHS Trust.
In Feb 2015, I co-chaired the first e-Oncology workshop bringing together expertise from clinical and computational areas. This is being followed by the Computational Oncology conference on Tues 27th March at Imperial College (South Ken).
If you are interested in pursuing a PhD (or MD/ PhD if clinical) in one of these areas please email me.
My clinical work is based at Charing Cross Hospital, but my academic work is based in the Big Data Analysis Unit (BDAU) on the St. Mary's Campus. My ORCHID ID is 0000-0001-7096-0718
Big Data in Cancer Care
As more healthcare data becomes available electronically, there are increasing opportunities and questions about what we can do with this data.
In conjunction with colleagues at UCH and UCL, we have developed methods for using routine data to estimate recurrence-free survival in patients with head and neck cancers. This work is now being extended to provide these abilities at national level in conjunction with the NCIN. We have also applied this to brain and lung cancers.
Through the NCRAS CNS/Brain Tumour group, we are beginning to apply these approaches to patients with brain tumours. We have reported the survival of all patients with glioblastoma multiforme in England (2007 - 2012), and have recently reported the first ever analysis of the impact of neurosurgical volumes on 30 day mortality outside of the USA.
Computational Advances for Medical Reasoning
In conjunction with colleagues at UCL and Imperial (including the EPSRC-funded TRADAR project), we are developing better tools for reasoning with and understanding clinical data, especially clinical trials. Our work has been published widely and develops a series of different approaches to reasoning with data extracted from clinical trials. Most of these are based on computational argumentation, a relatively new technique in non-montonic reasoning.
A summary of our approach can be found in our papers in AI in Medicine (2012) and Lung Cancer (2015), including a tutorial appendix to the Lung Cancer paper. We gave a webinar summarising our work to the Cochrane Methods Group in the Summer of 2016, and the videos from this can be found here
New technology for Proton Therapy
Proton therapy is a technological development of radiotherapy, and may be better than conventional radiotherapy for some patients. However, the current technology for the delivery of proton-beam therapy is very large, and very expensive.
In conjunction with colleagues in High-Energy Physics at Imperial, we are exploring novel technological approaches for proton therapy. If successful, these may allow clinical proton accelerators to be manufactured that would be an order of magnitude smaller and cheaper than current technology.
Novel Technologies for Assessing Cancer Patients
In conjunction with our interest in using Cancer "Big Data" we are also interested in developing and testing new sources of data.
In particular, we are interested in the role of high-frequency/ continuous near-patient monitoring devices, and the associated computational tools needed to use and reason with this data.
For that reason, we have started the BrainWear clinical trial that is looking at adding a wrist-worn sensor to conventional measures of disease and toxicity in patients with primary and metastatic brain tumours.