Pain recognition with camera photoplethysmography

In the last years a lot of effort was made in predicting the heart rate of a participant with remote Photo-plethysmography (rPPG) from the video channel but only few authors used it as a biosignal for classification of e.g. stress. In this work, we present the rPPG signal as a new modality for pain classification and evaluate the benefit of the three color channels (red, green, blue) of the rPPG signal. In short the rPPG signal is filtered in multiple frequency ranges to extract the heart rate and the respiration rate as biophysiological signals. Then the pain is classified with a Support Vector Machine (SVM) and Random Forest classifier. The performance is compared to the electrocardiogram (ECG) and the respiration from the biosignal amplifier and facial landmark features from the video. The results show that the rPPG signal can be used for pain classification, especially its low frequencies.