Influencia de las estructuras profundas sobre el EEG y su estudio invasivo y no invasivo

Resumen

The EEG consists in a graphical representation of the summated postsynaptic potentials generated in the pyramidal neurons. One of the main challenges of EEG is to decipher the relation between the recorded EEG activity and the activity in the neuronal networks. To find the source of EEG activity, complex non-linear and linear mechanisms as well as volume conduction effect and influence of the shape and electrical properties of the brain and skull need to be taken in consideration. In addition, brain regions are profusely interconnected and modulate each other adding additional complexity. In epilepsy it is particularly relevant to find out the initiation of the epileptiform activity and seizures as well as their propagation to connected areas. This is particularly difficult for generalised activity which involve large cortical regions bilaterally by the time it is recorded on the scalp, thus hiding any contributing deeper source. Invasive techniques are often employed to record intracranial fields and better define the epileptogenic zone. Advances in signal analysis have contributed to improve the detection of deep sources using different mathematical methods. Our group has developed an alternative approach employing single pulse electrical stimulation to investigate both, functional connectivity and increased epileptogenicity.To address objective 1, we developed an algorithm capable of detecting on the scalp EEG epileptiform discharges generated in the mesial temporal region visible with foramen ovale electrodes but not visually detectable in the scalp EEG. For objective 2, we reviewed patients assessed with intracranial electrodes in which SPES elicited responses similar to spontaneous K-complexes. For objective 3, we studied ictal and interictal thalamic and surface recordings from 3 patients assessed for deep brain stimulation of the centromedian nucleus of the thalamus. For objective 4, we developed a model based in control systems to explain the morphology of early responses to SPES. It is well known that scalp EEG activity emerges from the interaction of superficial and deep cortical and subcortical structures within a conductive medium. In first place the present work we show that It is possible to detect epileptiform activities generated in deep areas of the mesial temporal lobe at the scalp level. Secondly, we have assessed with SPES the ability to produce cortically-generated phasic events similar to those physiologically generated by the cortex (K-complexes). The role of deep mesial structures in generalised seizures was assessed with thalamic centromedian electrodes. Our analysis suggested that seizure onset in generalised seizures can be complex but once the thalamus is involved, it becomes the leading structure over the cortex acting as a pacemaker. Finally, we developed a model to characterise responses to SPES employing a set of control systems, able to predict intrinsic properties of the cortex such as background rhythms, and the capability to generate epileptiform activity and possibly seizures. The aim of this thesis is to tease out the contributions of deep and superficial structures to the EEG in order to determine whether scalp EEG contains enough information to reliably identify focal hippocampal epileptiform discharge sources in temporal lobe epilepsy (objective 1); identify the structures originating K-complexes (objective 2); estimate the role of centromedian thalamic nucleus in the initiation and maintenance of seizures in humans (objective 3) and develop a method to characterise the morphology of responses to single pulse electrical stimulation (SPES) that could explain the oscillatory behaviour of the spontaneous EEG (objective 4).This work shows how quantitative methods can help to improve our understanding of the relative contribution of deep structures to the EEG and the interplay between different areas leading to physiological and pathological scenarios.