การพัฒนาเครื่องวัดแอลกอฮอล์โดยใช้โฟโตเพลทิสมอกราฟี

Abstract

This research investigates and develops a non-invasive alcohol meter using a PPG signal, a signal indicator for the detection of changes in blood volume. There were two goals in this study: 1) Classifying PPG's signal into groups at the signal before and after alcohol consumption. The PPG signal differentiation was analyzed in a cohort of 25 healthy subjects aged 20 to 45 years using the datum line rotation technique. The best result was 90% based on Machine Learning classification. 2) The test compared the PPG data set with a fuel cell breath alcohol analyzer to determine blood alcohol level (Blood Alcohol Concentration; BACs). The study could draw using a simple regression equation of only 36.18% in the alcohol level range of 15 to 80 mg%. the result of the study provides a preliminary result of discrimination between consumed and non-consumed classes. It is suggested to use PPG as an initial screening and it is suitable for a portable-measurable device. However, the result yet cannot be claimed in the evidential test.