Impact of yoga asanas on primary dysmenorrhea-EEG band power analysis and deep learning based classification

Background and objective

Women of all ages, starting from puberty, often encounter challenges related to premenstrual syndrome (PMS) and menstrual symptoms. These challenges can include abdominal pain and psychological effects such as stress and anxiety, which can significantly impact brain function. However, yoga has emerged as a promising method to alleviate these menstrual-related issues.

Aim

To study the impact of specific yoga asanas on primary dysmenorrhea in engineering college girl students with Electroencephalogram (EEG) signals.

Method

This interventional study was conducted from July 2023 to December 2023 with a group of 34 young female engineering college students aged 18 to 25, lasting for six months. Participants were selected for the control and intervention groups after getting informed consent. The intervention group underwent a supervised 24-week yoga program, focusing on five specific asanas designed to alleviate menstrual symptoms. This program was conducted five days a week, with each session lasting 40 min.EEG recordings were carried out during the 1st week, 12th week, and 24th week of the program using an 16-channel OpenBCI system for both groups. Power spectral density (PSD) analysis was performed for different frequency bands across various time bins in different brain lobes, and the classification of the signals was conducted using deep learning techniques. Participants’ subjective responses to the asana practice were also collected, and suitable statistical tests were performed as needed. The control group did not participate in any yoga practice.

Results

Practicing yoga asanas in the sequence of Nadi Shudhana, followed by Marjaryasana, Ustrasana, Paschimottanasana, Virasana, and Savasana over six months has led to significant changes. The yoga group experienced decreased pain severity compared to the control group. The Age and EEG power levels between both groups (N = 34) were statistically comparable. Power Spectral Density (PSD) analysis indicates an increase in alpha brainwave activity after six months of practicing yoga asanas. Bidirectional Long Short-Term Memory (BI-LSTM) a deep learning model, demonstrated an impressive ability to accurately classify the signals as pain, normal, or yoga, achieving a high accuracy rate of 93.5 %, which surpassed previous studies. Additionally, 91 % of the participants reported a reduction in menstrual abdominal pain as well as associated psychological symptoms like anxiety and stress through the survey conducted after the intervention study.

Conclusion

This study’s findings indicate that yoga asanas can effectively reduce menstrual pain and enhance cognitive functions, as demonstrated through EEG PSD analysis and participant feedback. This suggests that yoga may serve as a viable alternative treatment for primary dysmenorrhea in young women.