Feature Extraction from Finger Pulse Profile for High Pitta Detection

Abstract

Ancient Indian system of medicine popularly known as Ayurveda was heavily dependent on diagnosis through nadi parikshan i.e. pulse diagnosis. Through nadi parikshan, the tridosha (vata, pitta and kapha) level of body was studied and checked for imbalances, and the latter if present, was considered the very root of the ailment or illness in the body. This analysis required no sophisticated instrument and was totally dependent on the expertise of the expert. In the modern times we have highly sophisticated instruments but the expertise available earlier for diagnosis is missing. Pulse Plethysmograph is one modern instrument which can be used to overcome this lacking expertise and hence to revive this dying art of nadi parikshan. According to Ayurveda, the level of pitta increases immediately after lunch. In this thesis it has been proposed that a substantial relation may lie between the finger pulse features and high pitta level in human body. In this work the finger PPG waveform has been studied and some prominent features have been extracted from the pulse wave. The data has been recorded using BIOPAC MP System and AcqKnowledge software. The features thus extracted from the finger pulse plethysmogram have been used to find a relation between increased pitta and the peripheral finger pulse. For this pilot study data was acquired from three subjects and three parameters were extracted from both right and left hands, before and immediately after lunch. On calculating and comparing the mean and variance of these parameters, it has been found that the variance of A2/A1 in middle finger, and TP2/TPT in ring finger, decreases consistently for all the three subjects, in both left and right hand. This result may be further validated used as an indication of high pitta level in human body.