Epileptic seizure suppression: a computational approach for identification and control using real data

Abstract

AbstractEpilepsy affects millions of people worldwide every year and remains an open subject for research. Current development on this field has focused on obtaining computational models to better understand its triggering mechanisms, attain realistic descriptions and study seizure suppression. Controllers have been successfully applied to mitigate epileptiform activity in dynamic models written in state-space notation, whose applicability is, however, restricted to signatures that are accurately described by them. Alternatively, autoregressive modeling (AR), a typical data-driven tool related to system identification (SI), can be directly applied to signals to generate more realistic models, and since it is inherently convertible into state-space, it can thus be used for the artificial reconstruction and attenuation of seizures as well. Considering this, the first objective of this work is to propose an SI approach using AR models to describe real epileptiform activity. The second objective is to provide a strategy for reconstructing and mitigating such activity artificially, considering non-hybrid and hybrid controllers − designed from ictal and interictal events, respectively. The results show that AR models of relatively low order represent epileptiform activities fairly well and both controllers are effective in attenuating the undesired activity while simultaneously driving the signal to an interictal stage. These findings may lead to customized models based on each signal, brain region or patient, from which it is possible to better define shape, frequency and duration of external stimuli that are necessary to attenuate seizures.Author summaryEpilepsy is perhaps one of the most studied brain disorders and it is still not sufficiently well understood. The use of computational models is useful in this case since several simulations can be run using them, such that experience and insight about seizures can be gained without necessarily carrying out experiments. These models are usually designed with or without some knowledge about the brain region or phenomenon. Seizure attenuation approaches have been proposed for the first case, but they are limited to the type of seizure correctly described by the model. The present work proposes a similar procedure for the second one (where only the data are available and nothing else is assumed), which is regarded as more realistic due to its direct application on the signals and can lead to customized models for each activity, brain region or patient, defining important information such as the shape, frequency and duration of the external stimuli that must be applied to mitigate a seizure.