£149,500 over 36 months
Awarded in 2015
An optimal computer model for the diagnosis and prognosis of focal epilepsies
Professor John Terry
Dr Marc Goodfellow, Prof Christoph Michel, Prof Mark Richardson
University of Exeter
The misdiagnosis of epilepsy remains a significant clinical problem, which can lead to delayed or unnecessary treatment. Our computer-based tool offers real potential to bring personalised medicine into epilepsy, which will ensure that people presenting at clinic receive an effective diagnosis and rapid access to the most appropriate treatment.
Professor John Terry
Why is this research needed?
The diagnosis of epilepsy currently relies on a description of the event(s) supported by the results of hospital tests, particularly EEG. Conventional EEG too frequently ‘misses’ epilepsy cases or falsely shows epilepsy to be present, and it is therefore is an unreliable diagnostic test. EEG recorded during a seizure is reliable, but it is inconvenient and expensive to obtain.
What are the aims?
Professor Terry and colleagues plan to find out if it is possible to identify whether a person has focal epilepsy, and whether or not they have responded to treatment, without having to observe seizures via clinical recordings.
How will the research be carried out?
The team’s ongoing research, funded by Epilepsy Research UK, has developed computer models that can analyse short ‘resting state’ periods of routine clinical EEGs and successfully identify whether a person has epilepsy in almost 80% of cases. In this study the group will validate and optimise their computer models in order to progress them towards a fully automated system that can be used clinically. To achieve this, they will work with an experimental epilepsy team based in Switzerland to repeatedly refine their models and test their ability to identify different treatment outcomes in rodents. They will then apply the optimised models to human data collected at King’s College London.
What difference will it make?
If successful it could, within 5 years, start to improve the accuracy of focal epilepsy diagnoses and allow more prompt identification of the most effective treatment(s).