광용적맥파 미분 기반 수술 후 통증 평가 지표

The purpose of this study was to develop pain assessment indices based on derivative photoplethysmogram for objective postoperative pain assessment. We obtained photoplethysmogram and surgical pleth index from 78 surgical patients in preoperative and postoperative conditions. Then, we extracted a total of 48 time and amplitude related candidate features from derivative photoplethysmogram. As a statistical result, 35 of the 48 basic features were significantly different (p<0.05) preoperation and postoperation. After normalization, the coefficient of variation of all candidate features decreased by 5.7%, time related features increased by 3%, amplitude related features decreased by 12.7%. In postoperative pain classification using logistic regression model, top 20 classification accuracy of 48 features including had an average of 64.9 ± 1.6% accuracy of pain assessment and the accuracy showed 0.9% higher than surgical pleth index.