Classification of Electrophotonic Images of Yogic Practice of Mudra through Neural Networks

BACKGROUND: Mudras signify a gesture with hands, eyes, and the body. Different configurations of the joining of fingertips are also termed mudra and are used by yoga practitioners for energy manipulation and for therapeutic applications. Electrophotonic imaging (EPI) captures the coronal discharge around the fingers as a result of electron capture from the ten fingers. The coronal discharge around each fingertip is studied to understand the effect of mudra on EPI parameters. METHODS: The participants were from Swami Vivekananda Yoga Anusandhana Samsthana and Sushrutha Ayurvedic Medical College, in Bengaluru, India. There were 29 volunteers in the mudra group and 32 in the control group. There were two designs: one was a pre-post design with control the other was pre-post with repeated measures with 18 individuals practicing mudra for 3 days. The duration of intervention for the pre-post design was 10 min on the 1(st) day, 15 min on the 2(nd) day, and 20 min on the 3(rd) day. A neural network classifier was used for classifying mudra and control samples. RESULTS: The EPI parameters, normalized area and average intensity, passed the test of normality Shapiro–Wilk. The Cohen's d, effect size was 0.988 and 0.974 for the mudra and control groups, respectively. Neural network-based analysis showed the classification accuracy of the post-intervention samples for mudra and control varied from 85% to 100% while the classification accuracy varied from 55% to 70% for the pre-intervention samples. The result of the mudra intervention showed statistically significant changes in the mean values on the 3(rd) day compared to the 1(st) day. CONCLUSIONS: The effect size of the variations in mudra was more than that of the control group. Mudra practice of a longer duration showed statistically significant change in the EPI parameter, average intensity in comparison to the practice on the 1(st) day.