Neural Computing and Applications
March 5, 2026
Daisy Das, Nabamita Deb
This review systematically links EEG oscillations to relaxation achieved through mantras, yoga, and meditation. It defines EEG activity and examines individual brainwave bands associated with relaxation, while identifying gaps in the existing EEG literature. The work integrates traditional healing practices with modern EEG research, offering a comprehensive overview of methodologies, analysis techniques, and potential applications for understanding relaxation and well-being. It also highlights the relationship between mantras and EEG parameters, along with classification techniques, preprocessing steps, relaxation features, and accuracy outcomes used in EEG studies.
Biomedical physics & engineering express
March 31, 2026
Daisy Das, Nabamita Deb, Saswati Sanyal Choudhury
Brief audio meditation measurably alters brain activity in pregnant women. EEG signals recorded during resting, meditation, and post-meditation states were analyzed using Variational Mode Decomposition and a hybrid deep learning model combining CNNs and Bidirectional LSTM networks. The model classified the three mental states with 94.79% test accuracy, demonstrating that short-term cognitive modulation can be detected reliably. The findings provide an objective framework for evaluating meditation's neural effects during pregnancy, supporting its potential as a non-pharmacological intervention for stress, anxiety, and depression in maternal care.
Annals of Neurosciences
January 23, 2026
Tony Bayan, Daisy Das, Nabamita Deb
A deep hybrid model combining convolutional neural networks, bidirectional long short-term memory networks, and attention mechanisms achieved 99.46% accuracy in classifying EEG recordings from experienced meditation practitioners into resting or mantra-listening brain states. Wavelet-based time-frequency features were extracted from EEG data recorded before and during auditory mantra stimulation. The model outperformed simpler architectures including CNN alone (76.92%), LSTM alone (75.30%), CNN+LSTM (84.62%), and CNN+BiLSTM (88.65%). Receiver operating characteristic analysis confirmed high discriminative capability with an AUC near 1.0. The approach demonstrates that combining convolutional, recurrent, and attention methods substantially improves spatial-temporal feature learning for distinguishing brain states.
Annals of Neurosciences
October 17, 2025
Daisy Das, Nabamita Deb, Rita Rani Talukdar et al.
Short audio interventions can induce relaxation in pregnant women, as measured by EEG brain activity. A new classification model combining EEG data with an artificial neural network achieved 100% accuracy in identifying meditative states during mantra meditation, particularly using signals from the frontal midline, right frontal lobe, and parietal lobe. The study emphasizes that EEG readings during meditation are more significant than those after meditation, and it offers insights into trimester-specific responses to brief audio stimuli, suggesting such interventions may be useful for stress reduction in prenatal care.
2025 3rd International Conference on Intelligent Systems, Advanced Computing and Communication
February 27, 2025
Daisy Das, Bhabesh Kalita, Nabamita Deb et al.
Brief meditation increases theta brainwave power in the frontal region, which is linked to attention and relaxation. Analyzing EEG data from the PEM-43 dataset, theta power rose during meditation (52.4 µV2·Hz) and dropped afterward (45 µV2·Hz) compared to resting state. Frontal channels showed the highest mean theta power (Fp1: 59.5 µV2·Hz, Fz: 59.4 µV2·Hz), while central and posterior regions had lower activity. Statistical tests confirmed a significant increase during meditation, suggesting brief meditation can measurably modulate cognitive engagement and relaxation.