Abstract:Major depressive disorder (MDD) recognition using multi-channel electroencephalographic (EEG) signals has profound clinical value with its richness and accessibility of temporal information, but such signals may suffer from non-stationarity and redundant characteristics. To cope with these problems, the WS-BiLSTM-MA network is proposed with three core modules: 1) wavelet scattering network is applied to build a wavelet scattering matrix which captures the deformation stability information from multi-channel cleaned EEG signals, 2) bidirectional long short-term memory network is utilized to obtain a relevant scattering matrix which learns the potential temporal relationship from the wavelet scattering matrix, 3) mixed attention block has two self-attention blocks which gain the ability to further redistribute the weights of features and integrate the key comprehensive information from scattering coefficients and relevant matrices, respectively. The performance of our proposed WS-BiLSTM-MA network is evaluated on two MDD datasets in both eyes closed and eyes open conditions with 19-channel EEG signals: the Hospital University Sains Malaysia (HUSM) dataset and Zhongda Hospital Southeast University (ZHSU) dataset. To ensure subject independence, the leave-one-subject-out validation experiment and blind-test validation experiment are conducted. In leave-one-subject-out experiment, the WS-BILSTM-MA network, with the zeroth-order, first-order and second-order scattering coefficients, presents high performance in terms of classification accuracy, precision, recall and F 1-score whatever conditions in both datasets. The blind-test experiment demonstrates the excellent ability of our framework with only zeroth-order and first-order scattering coefficients, where the classification accuracy, precision, recall and F 1-score can reach 98.80%, 99.90%, 99.71%, 99.81% and 99.81%, 99.78%, 99.17%, 99.47% in eyes closed and eyes open conditions in HUSM dataset, 83.29%, 87.97%, 75.65%, 83.60% and 83.41%, 90.59%, 74.20%, 81.58% in eyes closed and eyes open conditions in ZHSU dataset, respectively. Compared with some state-of-the-art methods in HUSM dataset, the WS-BILSTM-MA network can improve the performance of MDD recognition proving its clinical interest.

Bipolar Disorder Recognition via Multi-scale Discriminative Audio Temporal Representation.

Bipolar Disorder Classification Based on Multimodal Recordings

Audio Visual Multimodal Classification of Bipolar Disorder Episodes

Multi-instance Learning for Bipolar Disorder Diagnosis Using Weakly Labelled Speech Data

DISTINGUISHING BIPOLAR DEPRESSION FROM MAJOR DEPRESSIVE DISORDER USING FNIRS AND DEEP NEURAL NETWORK

Bipolar Disorder Recognition with Histogram Features of Arousal and Body Gestures.

Bipolar Population Threshold Encoding for Audio Recognition with Deep Spiking Neural Networks

Multi-modality Hierarchical Recall based on GBDTs for Bipolar Disorder Classification.

Multimodal temporal machine learning for Bipolar Disorder and Depression Recognition

SE-1DCNN-LSTM: A Deep Learning Framework for EEG-Based Automatic Diagnosis of Major Depressive Disorder and Bipolar Disorder

A Hybrid Model for Bipolar Disorder Classification from Visual Information

Multimodal Depression Recognition with Dynamic Visual and Audio Cues

Modelling Paralinguistic Properties in Conversational Speech to Detect Bipolar Disorder and Borderline Personality Disorder

Multi-Head Attention-Based Long Short-Term Memory for Depression Detection From Speech

Leveraging Linguistic Characteristics for Bipolar Disorder Recognition with Gender Differences

Towards Suicide Prevention from Bipolar Disorder with Temporal Symptom-Aware Multitask Learning

Machine learning with multiple modalities of brain magnetic resonance imaging data to identify the presence of bipolar disorder

Bi-Emotional Siamese Network for MDD Recognition.

A Multimodal Approach for Automatic Mania Assessment in Bipolar Disorder

WS-BiLSTM-MA: Wavelet Scattering-Based BiLSTM with Mixed Attention Block for MDD Recognition Using Multi-Channel EEG Signals

The Deep Learning Method Differentiates Patients with Bipolar Disorder from Controls with High Accuracy Using EEG Data