Abstracts

Rationale: Epileptic syndromes and seizures are the expression of brain complex systems. Measurements of complexity and graph analysis have been used to track connectivity and circuitry from MRI and EEG data. However, no similar analysis has been applied to epileptic seizures semiology, which is the expression of those altered brain networks. Therefore, our goal was to apply neuroethology and graph analysis to the study the behavioral manifestations of epileptic seizures in human frontal lobe epilepsy (FLE) and temporal lobe epilepsy (TLE). Methods: We analyzed the video records of 120 seizures of 18 patients (FLE), and 28 seizures of 28 patients (TLE), all patients were seizure-free > 1 year after surgical treatment (Engel I). All patients' behaviors were annotated second-by-second and analyzed by means of a behavioral glossary containing all behaviors. Behavioral data, taken as item series (txt files) were analyzed for neuroethology using the Ethomatic software and displayed as flowcharts including frequency, mean duration, and sequential statistic interaction of behavioral items (X² ≥ 3.84, P < 0.05). The same item series were used for the graph analysis where each behavior is displayed as a node (circle) connected by edges with other nodes according with their temporal sequence of appearance. We used CYTOSCAPE® and additional home-made software to create and to analyze the graphs, choosing some of the most used measurements such as: cluster coefficient, degree (in and out-degree), and an estimation of the Shannon entropy, in addition to Global and Local entropies, based on the ratio between the degree of a node and the maximum degree of the graph. Results: Analysis of the flowcharts for both FLE and TLE have shown that well-established data in the literature were confirmed. In the case of FLE, the brief and frequent seizures, the complex motor behaviors, head and eye version, unilateral and bilateral tonic posturing, speech arrest, vocalization and rapid postictal recovery. In the case of TLE, the presence of epigastric aura, the lateralization value of dystonias, impairment of consciousness and speech during ictal and postictal periods, and development of secondary generalization. Conclusions: The graph analysis of FLE and TLE confirmed the data from the flowcharts, but because of the algorithms we used, it was more powerful highlighting connectivity measurements and complex associations among behaviors which are quite selective, depending on the origin of the seizures. Worth of notice, when we used filters to zoom-in specific behavioral categories such as tonic/clonic events or automatisms, we detected complex associations of behaviors in FLE and TLE seizures, beyond those detected with the flowcharts. The fact that we used algorithms widely used in order to track brain connectivity from EEG and MRI sources makes our quantitative semiology study, although initially quite exploratory, very promising for future studies in this field. Financial support: CNPq, CINAPCE-FAPESP, PROEX-CAPES, FAEPA.