Automatic detection of abnormal clinical EEG: comparison of a finetuned foundation model with two deep learning models

TL;DR

This study compares a foundation model and two deep learning models for classifying EEG recordings as normal or abnormal, demonstrating that the foundation model achieves superior accuracy across multiple datasets, highlighting the potential of pre-trained models in EEG analysis.

Contribution

It introduces and evaluates BioSerenity-E1, a foundation model, showing its advantages over traditional deep learning models in EEG abnormality detection.

Findings

BioSerenity-E1 achieved up to 89.19% balanced accuracy on large datasets.

The foundation model outperformed other models on smaller and external datasets.

Pre-trained models enhance EEG classification robustness and efficiency.

Abstract

Electroencephalography (EEG) is commonly used by physicians for the diagnosis of numerous neurological disorders. Due to the large volume of EEGs requiring interpretation and the specific expertise involved, artificial intelligence-based tools are being developed to assist in their visual analysis. In this paper, we compare two deep learning models (CNN-LSTM and Transformer-based) with BioSerenity-E1, a recently proposed foundation model, in the task of classifying entire EEG recordings as normal or abnormal. The three models were trained or finetuned on 2,500 EEG recordings and their performances were evaluated on two private and one public datasets: a large multicenter dataset annotated by a single specialist (dataset A composed of n = 4,480 recordings), a small multicenter dataset annotated by three specialists (dataset B, n = 198), and the Temple University Abnormal (TUAB) EEG corpus evaluation dataset (n = 276). On dataset A, the three models achieved at least 86% balanced accuracy, with BioSerenity-E1 finetuned achieving the highest balanced accuracy (89.19% [88.36-90.41]). BioSerenity-E1 finetuned also achieved the best performance on dataset B, with 94.63% [92.32-98.12] balanced accuracy. The models were then validated on TUAB evaluation dataset, whose corresponding training set was not used during training, where they achieved at least 76% accuracy. Specifically, BioSerenity-E1 finetuned outperformed the other two models, reaching an accuracy of 82.25% [78.27-87.48]. Our results highlight the usefulness of leveraging pre-trained models for automatic EEG classification: enabling robust and efficient interpretation of EEG data with fewer resources and broader applicability.