Abstracto

Pulse based sensor design for wrist pulse signal analysis and health diagnosis

Krittika Goyal, Ravinder Agarwal

Pulse diagnosis is a common non-invasive method for measurement of health of person in Ayurveda. According to Ayurveda, the root cause of the disease is imbalance of three doshas i.e. vat, pit and kaph. The health status of the person is examined by Ayurvedic physician by feeling palpation from three fingers (index, middle and ring) placed on the radial artery for vat, pit and kaph respectively. Examination of the pulse (Nadi Parikshan) requires a lot of experience in pulse reading. It depends upon the perception of the practitioner, hence there is a need to develop pulse diagnosis system to obtain accurate diagnosis of disease. In present research work the design and development of the pulse diagnosis system was carried out, in which sensor MPXM2053D sensor from FREESCALE was used to sense the wrist pulse signals from 42 subjects. Thereafter, the signal conditioning circuit was designed using instrumentation amplifier. Real time monitoring was performed using myRIO DAQ card in LabVIEW myRIO 2014 and filtering was performed in Matlab. Frequency domain analysis using power spectrum was carried out and features such as band energy ratio were extracted. An interpretation of useful features to distinguish between healthy and unhealthy subject using statistical ANOVA analysis along with t test was carried out. Different classifiers were used to classify healthy and unhealthy subject and results were obtained with high accuracy, sensitivity and specificity. Thus wrist pulse acquisition system developed for pulse diagnosis will be a helpful tool for diagnosis of healthy and unhealthy person.