Scholarlink

Depressive Disorder (MDD) is an important medical research topic due to its high frequency and severe impact on quality of life, including disability, comorbidity, and suicide risk. MDD’s convoluted etiology, overlapping symptoms with other mental conditions, and subjective diagnostic techniques make it hard to diagnose early and accurately. DALENet (A Novel Approach for Predicting and Classifying Major Depressive Disorder Using DenseNet and Bi-LSTM with Attention Mechanism) improves MDD prediction accuracy using EEG data. Our technique uses modern deep learning models for feature extraction and selection. Start using Empirical Mode Decomposition (EMD) to extract all relevant EEG data. PSO prioritizes model performance characteristics. DenseNet, a robust convolutional neural network for spatial feature extraction, and a Bi-LSTM with an attention mechanism to capture temporal dependencies and significant signal characteristics are used in a fusion model for classification. The DALENet model evaluated on EEG data contains brain electrical activity, predicts MDD with 96.01% Accuracy, 96% Precision, 95% Recall, and 96 F1-Score. These measures show DALENet’s improved MDD EEG diagnosis.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

DALENet: A Novel Approach for Predicting and Classifying Major Depressive Disorder Using DenseNet and Bi-LSTM with Attention Mechanism

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Udutala Mahender,
  • S. Arivalagan,
  • V. Sathiyasuntharam,
  • P. Sudhakar

摘要

Depressive Disorder (MDD) is an important medical research topic due to its high frequency and severe impact on quality of life, including disability, comorbidity, and suicide risk. MDD’s convoluted etiology, overlapping symptoms with other mental conditions, and subjective diagnostic techniques make it hard to diagnose early and accurately. DALENet (A Novel Approach for Predicting and Classifying Major Depressive Disorder Using DenseNet and Bi-LSTM with Attention Mechanism) improves MDD prediction accuracy using EEG data. Our technique uses modern deep learning models for feature extraction and selection. Start using Empirical Mode Decomposition (EMD) to extract all relevant EEG data. PSO prioritizes model performance characteristics. DenseNet, a robust convolutional neural network for spatial feature extraction, and a Bi-LSTM with an attention mechanism to capture temporal dependencies and significant signal characteristics are used in a fusion model for classification. The DALENet model evaluated on EEG data contains brain electrical activity, predicts MDD with 96.01% Accuracy, 96% Precision, 95% Recall, and 96 F1-Score. These measures show DALENet’s improved MDD EEG diagnosis.