Studies on the Correlation and Pattern Classification between Pulse and PPG Signal

碩士 === 逢甲大學 === 生醫資訊暨生醫工程碩士學位學程 === 97 === PPG (Photoplethysmography) is a noninvasive bio-signal measurement equipment which not only provide the peripheral hemal circulatory circumstance but also the heart rate (HR) signal can be obtained by the appropriate algorithm. However, the relationship bet...

Full description

Bibliographic Details
Main Authors: WEN-HSIU CHEN, 陳雯琇
Other Authors: Yue-Der Lin
Format: Others
Language:zh-TW
Published: 2009
Online Access:http://ndltd.ncl.edu.tw/handle/66345820684623840676
Description
Summary:碩士 === 逢甲大學 === 生醫資訊暨生醫工程碩士學位學程 === 97 === PPG (Photoplethysmography) is a noninvasive bio-signal measurement equipment which not only provide the peripheral hemal circulatory circumstance but also the heart rate (HR) signal can be obtained by the appropriate algorithm. However, the relationship between PPG signal and the pulse signal which is morphological equivalent to PPG signal have not been discussed and researched before. In this study, the linear and nonlinear methods were used on the PPI (peak-to-peak interval) of PPG and pulse signal which acquired simultaneously to assess the correlation between the two signals. The coupling degree can be extracted by correlation coefficient, the linear method, and Complexity C, the nonlinear method. Then the study will use the method of pattern recognition to differentiate the pulse condition. Hence, the assessment result will indicate the method can be surrogated by the other. Twelve healthy people enrolled and the average value of correlation coefficient is 0.94±0.05, and the LF/HF of correlation coefficient is 0.95. The coupling degree index (I) value is 0.1±0.05. The value of agreement is 0.97±0.02. Although the physical meanings of PPG and pulse are different, HRPPG and HRPulse have high correlation. In the section of pulse condition recognition by pattern recognition, we discovered that the results cannot correspond to the expected pulse condition. Even though there are high correlation between these two signals in linear, nonlinear and frequency-domain. However, this method cannot replace the diagnosis of pulse conditions.