Automatic Detection and Classification of Epileptic Seizures from EEG Data: Finding Optimal Acquisition Settings and Testing Interpretable Machine Learning Approach.

Deep learning (DL) is emerging as a successful technique for automatic detection and differentiation of spontaneous seizures that may otherwise be missed or misclassified. Herein, we propose a system architecture based on top-performing DL models for binary and multigroup classifications with the non-overlapping window technique, which we tested on the TUSZ dataset. The system accurately detects seizure episodes (87.7% Sn, 91.16% Sp) and carefully distinguishes eight seizure types (95-100% Acc). An increase in EEG sampling rate from 50 to 250 Hz boosted model performance: the precision of seizure detection rose by 5%, and seizure differentiation by 7%. A low sampling rate is a reasonable solution for training reliable models with EEG data. Decreasing the number of EEG electrodes from 21 to 8 did not affect seizure detection but worsened seizure differentiation significantly: 98.24 ± 0.17 vs. 85.14 ± 3.14% recall. In detecting epileptic episodes, all electrodes provided equally informative input, but in seizure differentiation, their informative value varied. We improved model explainability with interpretable ML. Activation maximization highlighted the presence of EEG patterns specific to eight seizure types. Cortical projection of epileptic sources depicted differences between generalized and focal seizures. Interpretable ML techniques confirmed that our system recognizes biologically meaningful features as indicators of epileptic activity in EEG.

Unraveling Lifelong Brain Morphometric Dynamics: A Protocol for Systematic Review and Meta-Analysis in Healthy Neurodevelopment and Ageing.

A high incidence and prevalence of neurodegenerative diseases and neurodevelopmental disorders justify the necessity of well-defined criteria for diagnosing these pathologies from brain imaging findings. No easy-to-apply quantitative markers of abnormal brain development and ageing are available. We aim to find the characteristic features of non-pathological development and degeneration in distinct brain structures and to work out a precise descriptive model of brain morphometry in age groups. We will use four biomedical databases to acquire original peer-reviewed publications on brain structural changes occurring throughout the human life-span. Selected publications will be uploaded to Covidence systematic review software for automatic deduplication and blinded screening. Afterwards, we will manually review the titles, abstracts, and full texts to identify the papers matching eligibility criteria. The relevant data will be extracted to a 'Summary of findings' table. This will allow us to calculate the annual rate of change in the volume or thickness of brain structures and to model the lifelong dynamics in the morphometry data. Finally, we will adjust the loss of weight/thickness in specific brain areas to the total intracranial volume. The systematic review will synthesise knowledge on structural brain change across the life-span.