An investigation of linear correlation for a new alternative hardware approach to process the medical wearable sensor signals

Until now, the three-axis accelerometer (ACC) has been the major way of reflecting the motion artifact (MA) that is normally adopted in the Adaptive Noise Cancellation (ANC); however, this technique has failed to achieve the expected level of MA required for accurate processing. In this study, a covered photodetector (CPD) was investigated for possible usage in MA detection; this approach was considered as a way of unifying the sources of the generated desired & MA signals such that electronic noise differences can be avoided. At first, the two devices (ACC & CPD) were tied together, subjected to various forms of motions, and monitored for amplitude fluctuations (AFs) that respond in the time domain; the linear correlation coefficient (r) was also observed. The evaluation results showed the validity of the suggested method via pointing a rapprochement in AFs. In a steady-state, the estimated r value was 0.999 while in motion states, the r-value ranged from 0.998 - 0.794 in relation to changes in movements. Replacing the ACC by the CPD, such a concept can be implemented for processing Photoplethysmograph signals within ANC to achieve a cost-efficient accurate medical diagnosis.