Computer Modeling May Improve Surgical Treatment of Temporal Lobe Epilepsy Interview with:
Dr. Frances Hutchings

Interdisciplinary Computing and Complex BioSystems
School of Computing Science, Newcastle University
Newcastle upon Tyne, United Kingdom

Medical Research: What is the background for this study?

Response: Temporal lobe epilepsy is a prevalent and disruptive disorder, which is often treated with surgical removal of epileptic tissue when medication fails to repress seizures. In around a third of cases surgery is unsuccessful at preventing seizures. The aim of this study is to seek ways to improve surgery success rates by giving a better prediction of where seizures are starting and spreading on an individual patient basis, using an individual’s brain imaging data. Surgery is simulated in the model to provide a prediction of a procedures effectiveness at reducing seizures.

Medical Research: What are the main findings?

Response: Using patient Diffusion Tensor Imaging data to reconstruct the brain as a network, locations commonly implicated in temporal lobe epilepsy were found by the model to be most vulnerable to seizures. Simulations of surgery (following a commonly used surgery procedure) in patient models predicted a surgery success rate close to 70%, matching clinical observations. Subject specific surgery simulations were also tried, following individual predictions from the model of which regions to remove for which person. These showed far greater improvements in seizure likelihood than regular surgery. This is a preliminary study and there is much to be done to improve the predictive success, and also to ensure that model predicted subject specific surgery regions are safe to remove. Nonetheless it is a significant move towards computer aided patient specific surgery planning to improve outcomes.

Medical Research: What should clinicians and patients take away from your report?

Response: Computational models using individualised patient data have the potential to be the way forward for more accurate patient specific surgery planning. Such models will hopefully soon be able to give extra information and guidance to clinicians with the aim of improving outcomes for patients. This study is a first step towards this practical application of computer modeling.

Medical Research: What recommendations do you have for future research as a result of this study?

Response: The most important next step is to validate the predictions of this model with patient outcomes from surgery. After that, the model and prediction success rates need to be improved to the point where it is providing a reliable estimate of whether a procedure will be successful or not. From there the model can be used to predict alternative surgery sites and inform surgery procedures planned by clinicians.


Predicting Surgery Targets in Temporal Lobe Epilepsy through Structural Connectome Based Simulations

Frances Hutchings ,Cheol E. Han,Simon S. Keller,Bernd Weber,Peter N. Taylor ,Marcus Kaiser

PLOS Computational Biology

[wysija_form id=”5″]

Dr. Frances Hutchings (2015). Predicting Surgery Targets in Temporal Lobe Epilepsy through Structural Connectome Based Simulations