Wave Detection of Acceleration Plethysmogram Based on a Closed-Loop Feedback Mechanism
Jae Mok Ahn1, Jeom Keun Kim2

1Jse Mok Ahn*, School of Software, Hallym University, Chuncheon-si, Gangwon-do, South Korea.
2Jeom Keun Kim, School of Software, Hallym University, Chuncheon-si, Gangwon-do, South Korea.
Manuscript received on April 20, 2020. | Revised Manuscript received on May 01, 2020. | Manuscript published on May 10, 2020. | PP: 215-220 | Volume-9 Issue-7, May 2020. | Retrieval Number: G5180059720/2020©BEIESP | DOI: 10.35940/ijitee.G5180.059720
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: Acceleration plethysmogram (APG), which corresponds to the second derivative of the photoplethysmogram (PPG) is a noninvasive method for investigating arterial wall thickness and predicting cardiovascular diseases. To perform an APG wave analysis, five inflection points of the APG waveform (a, b, c, d, and e waves) must be successfully obtained. However, an abrupt change in PPG amplitude due to various physiological conditions and patient’s movement has made it very difficult to detect the five waves of the APG waveform in real time. Therefore, to resolve this problem, two stabilization methods for PPG and APG amplitudes were proposed based on a closed-loop feedback amplitude control mechanism. The regulation of PPG amplitude was rapidly carried out in four cardiac cycles by controlling the driving current to a light emitting diode (LED) through pulse width modulation (PWM). Two predetermined amplitude levels were applied to adjust the 1st and the 2nd derivatives of the PPG simultaneously when the wave detection algorithm failed to detect even one of five waves. Forty measurements of the APG signal on an index fingertip were performed to verify a closed-loop feedback amplitude control mechanism (CLFACM). The values of the t statistic (statistical significance) for the a, b, c, d, and e wave groups were 1.08292 (p=0.2855), 0.19607 (p=0.8456), 0.28955 (p=0.7737), 0.39467 (p=0.69467), and 0.50973 (p=0.6131), respectively. To identify waves of extreme values away from normality, the coefficient of kurtosis was obtained; the smallest value was obtained for the d wave (0.07335), and the largest value was obtained for the e wave (3.9456). The results suggested that the in-group waves did not significantly differ. The CLFACM played an important role in increasing the success rate of accurately detecting five waves from the APG signal. 
Keywords: Acceleration Plethysmogram, arterial wall thickness, Photoplethysmogram, Closed-loop feedback control, Pulse width modulation, Physiological condition.
Scope of the Article: Systems and Software Engineering