Decentralized EEG-Based Detection of Major Depressive Disorder via Transformer Architectures and Split Learning

Muhammad Umair ¹ Jawad Ahmad ² Nada Alasbali ³ Oumaima Saidani ⁴ Muhammad Hanif ^5* Aizaz Ahmad Khattak ⁶ Muhammad Shahbaz Khan ⁶

• ¹ Multimedia University, Cyberjaya, Selangor Darul Ehsan, Malaysia
• ² Prince Mohammad bin Fahd University, Khobar, Saudi Arabia
• ³ King Khalid University, Abha, Saudi Arabia
• ⁴ Princess Nourah bint Abdulrahman University, Riyadh, Saudi Arabia
• ⁵ Örebro University, Örebro, Örebro, Sweden
• ⁶ Edinburgh Napier University, Edinburgh, United Kingdom

Major Depressive Disorder (MDD) remains a critical mental health concern, necessitating accurate detection. This study utilizes EEG signals to classify MDD and healthy individuals through a combination of machine learning, deep learning, and split learning approaches. Multiple classifiers, including Random Forest, Support Vector Machine, and Gradient Boosting, as well as advanced architectures such as Transformers and Autoencoders, are evaluated. Results demonstrate a commendable classification performance with certain ensemble methods, and a Transformer-Random Forest combination achieved 99% accuracy. In addition, to address data-sharing constraints, a split learning framework is implemented across three clients, yielding high accuracy (over 95%) while preserving privacy. The best client recorded 96.23% accuracy, underscoring the robustness of combining Transformers with Random Forest under resourceconstrained conditions. These findings demonstrate that distributed deep learning pipelines can deliver precise MDD detection from EEG data without compromising data security.