12 noon - 1pm
Thursday 29 November 2018
Seizure prediction: A computer science challenge?
Lucas Franca will be speaking about how mathematical methods can help characterise epileptic seizures.
University of Surrey
This event has passed
- Lucas Franca (University College London)
Over the last decades, researchers have thoroughly searched for a seizure prediction method. Such a technology further became a ‘Holy Grail’ of epileptology due to its potential to revolutionise healthcare and symptoms management in patients with intractable epilepsies.
Achieving such an ambitious aim requires, nevertheless, some insight on the mechanisms of ictogenesis and how these properties manifest in brain signals. Every year, extensive recordings of brain signals of pre-ictal and ictal states are obtained during pre-surgical evaluation in patients with intractable epilepsies. Applying mathematical and/or statistical techniques to study and model such data might provide new hints on how epileptic seizures arise and evolve.
In this talk, Luca Franca will discuss putative mathematical methods to characterise epileptic seizures and the role of machine learning and pattern recognition in seizure prediction and characterisation, as well as common pitfalls in these techniques.