Improved Accuracy in Depression Detection Using EEG Signals with CNN and LSTM Algorithms in Comparison to the CNN Algorithm

B. Latha, M. Dharani, R. Ravichandran, L. Meganathan, S. Pooja, S. Sasi Rekha

2025

Abstract

Aim: The main aim of a hybrid-station of CNN with LSTM algorithms has been developed for high-accuracy depression detection from EEG signals. There are two groups in this study. Group 1 is the model of detecting high-accuracy depression using EEG signals by applying the CNN and LSTM algorithms compared to Group 2, which only applies the CNN algorithm. Both models were tested with Google Co-lab. The G Power value is set at 80% with a threshold of 0.05% and a confidence interval at 95%. Performance evaluation was performed in terms of the accuracy, precision, and F1 score, showing the superiority of hybrid CNN-LSTM over CNN in depression detection. The hybrid model obtained an accuracy of 92% with an F1 score of 0.91 while significantly outperforming the CNN model, which only reached 85% in terms of accuracy and an F1 score of 0.87. The optimal performance for the hybrid model was also noted with a significance level of 0.001. Based on the findings, it is found that the hybrid CNN-LSTM model provides a more effective framework for possible detection of depression from EEG signals.

Paper Citation

in Harvard Style

Latha B., Dharani M., Ravichandran R., Meganathan L., Pooja S. and Rekha S. (2025). Improved Accuracy in Depression Detection Using EEG Signals with CNN and LSTM Algorithms in Comparison to the CNN Algorithm. In Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - ICRDICCT`25; ISBN 978-989-758-777-1, SciTePress, pages 794-801. DOI: 10.5220/0013905900004919

in Bibtex Style

@conference{icrdicct`2525,
author={B. Latha and M. Dharani and R. Ravichandran and L. Meganathan and S. Pooja and S. Rekha},
title={Improved Accuracy in Depression Detection Using EEG Signals with CNN and LSTM Algorithms in Comparison to the CNN Algorithm},
booktitle={Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - ICRDICCT`25},
year={2025},
pages={794-801},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013905900004919},
isbn={978-989-758-777-1},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - ICRDICCT`25
TI - Improved Accuracy in Depression Detection Using EEG Signals with CNN and LSTM Algorithms in Comparison to the CNN Algorithm
SN - 978-989-758-777-1
AU - Latha B.
AU - Dharani M.
AU - Ravichandran R.
AU - Meganathan L.
AU - Pooja S.
AU - Rekha S.
PY - 2025
SP - 794
EP - 801
DO - 10.5220/0013905900004919
PB - SciTePress