A Convolutional Recurrent Neural Network With Attention For Response Prediction To Repetitive Transcranial Magnetic Stimulation In Major Depressive Disorder

This article proposes a deep learning model powered by state-of-the-art methods to classify responders (R) and non-responders (NR) to Repetitive Transcranial Magnetic Stimulation (rTMS) treatment for Major Depressive Disorder (MDD). Pre-treatment Electro-Encephalogram (EEG) signals from both public TDBRAIN dataset and 46 proprietary MDD subjects were utilized to create time-frequency representations using Continuous Wavelet Transform (CWT). Two powerful pre-trained Convolutional Neural Networks (CNN) named VGG16 and EfficientNetB0 were equipped with Bidirectional Long Short-Term Memory (BLSTM) and attention mechanism for the extraction of most discriminative spatiotemporal features from input images. The highest evaluated performance in 46 proprietary MDD subjects was acquired for the Frontal region using the TL-LSTM-Attention model based on EfficientNetB0 with accuracy, sensitivity, specificity, and Area Under the Curve (AUC) of 97.1%, 97.3%, 97.0%, and 0.96 respectively. Additionally, the TL-BLSTM-Attention models were evaluated on a public dataset called TDBRAIN and the highest accuracy of 82.3%, the sensitivity of 80.2%, the specificity of 81.9% and the AUC of 0.83 were obtained.