Abstracts

Identifying Seizure Onset Localization Using Spatial Activation: Applying a Novel Tool into Practice

Abstract number : 1.195
Submission category : 3. Neurophysiology / 3G. Computational Analysis & Modeling of EEG
Year : 2021
Submission ID : 1826185
Source : www.aesnet.org
Presentation date : 12/4/2021 12:00:00 PM
Published date : Nov 22, 2021, 06:52 AM

Authors :
Mariam khawaja, MD, MSc - Clinic Institute of Neuroscience, Epilepsy program, Neurology Department, Hospital Clinic of Barcelona.; Estefanía Conde-Blanco - Epilepsy Program, Neurology Department, Hospital Clinic of Barcelona - Clinic Institute of Neuroscience (ICN); Maria Centeno - Epilepsy Program, Neurology Department, Hospital Clinic of Barcelona - Clinic Institute of Neuroscience (ICN); Jordi Rumià - Epilepsy Program, Neurosurgery Department, Hospital Clinic of Barcelona, Barcelona, 08036, Spain - Clinic Institute of Neuroscience (ICN); Pedro Roldan - Epilepsy Program, Neurosurgery Department, Hospital Clinic of Barcelona, Barcelona, 08036, Spain - Clinic Institute of Neuroscience (ICN); Luis Pintor - Epilepsy Program, Psychiatry Department, Hospital Clinic of Barcelona, Barcelona, 08036, Spain - Clinic Institute of Neuroscience (ICN); Xavier Setoain - Epilepsy Program, Nuclear Medicine Department, Hospital Clinic of Barcelona, Barcelona, 08036, Spain - Clinic Institute of Neuroscience (ICN); Nuria Bargalló - Epilepsy Program, Neuroradiology Department, Magnetic Resonance Imaging Core Facility, Hospital Clinic of Barcelona, Barcelona, 08036, Spain - Clinic Institute of Neuroscience (ICN); Teresa Boget - Epilepsy Program, Neuropsychology Department, Hospital Clinic of Barcelona, Barcelona, 08036, Spain - Clinic Institute of Neuroscience (ICN); Manel Vila-Vidal - Universitat Pompeu Fabra (UPF); Gustavo Deco - Universitat Pompeu Fabra (UPF); Mar Carreño - Epilepsy Program, Neurology Department, Hospital Clinic of Barcelona - Clinic Institute of Neuroscience (ICN); Adrià Tauste - Universitat Pompeu Fabra (UPF), Universitat Politècnica de Catalunya (UPC); Antonio Donaire - Epilepsy Program, Neurology Department, Hospital Clinic of Barcelona - Clinic Institute of Neuroscience (ICN)

Rationale: The precise localization of the seizure onset zone is crucial for epilepsy surgery planning, prognosis and quality of life in people with drug-resistant epilepsy eligible for epilepsy surgery. The conventional qualitative visual analysis is the most commonly used method for reporting seizure type, location and classification. However, in many cases, the determination of the ictal onset zone remains unclear due to the fast propagation, EEG-artifacts and the mixture of different frequencies. Such results would be improved by using quantitative methods that help providing a more precise analysis. (Vila-Vidal et al. Clinical Neurophysiology 2017, 128, 6, p. 977-985).

Methods: We used a novel data-driven method, named "Brainfocus," for the spatial mapping of seizure-specific onset patterns in 10 patients evaluated with SEEG in our Epilepsy Surgery Program. This method quantifies the magnitude of seizure-related spectral changes with respect to a predefined baseline (Global activation) and the spread of these activations across recording sites.(Vila-Vidal et al. Neuroimage 2020, 208, 116410). Seizures were analyzed by two epileptologists independently, using visual inspection and Brainfocus quantitative analysis, respectively. We compared the number of electrodes/contacts involved in the first 5 seconds of seizure onset and determined the degree of spatial concordance using both methods.

Results: Patients' average age was 40.9 years, 60% males, mean age at seizure onset was 18.1 years and mean seizure duration was 22.8 years. The average number of AEDS was 3.4. More than a half (80%) showed a structural abnormality on MRI. The most frequent type of epilepsy was temporal lobe epilepsy (60%). The average number of implanted electrodes was 13. A total of 347 epileptic seizures were recorded from which 72 representative seizures were selected for further analysis.

The degree of agreement between qualitative visual inspection and Brainfocus global activation was substantial, Lin’s Concordance Correlation coefficient was 0.97. The sensitivity of Brainfocus for seizure onset localization was 0.93 ± 0.13 (mean ± std across patients). In 1 patient, the multi-band spectral frequency analysis allowed the detection of an additional electrode involved at seizure onset which had been overlooked by visual inspection.

Conclusions: Brainfocus is a quantitative tool that could be very useful to determine the characteristic spatio-temporal patterns at seizure onset. In addition, it allows the visualization of the epileptic activity propagation at different frequency bands, which greatly facilitates the precise interpretation of the SEEG results and consequently a better diagnosis.

Funding: Please list any funding that was received in support of this abstract.: Nothing to declare.

Neurophysiology