Automated Artifact-Detection Procedure for Vaginal Photoplethysmography

Vaginal photoplethysmography is the most commonly used method of assessing women's genital sexual arousal. Raw photoplethysmograph data consist of a series of peaks and troughs, and movement by the participant results in artifacts indicated by unusually high or low peak-to-trough amplitudes. The gold-standard approach to artifact detection involves visual inspection by a trained experimenter and manual removal of artifacts from the data set, however, this process is time-consuming and subject to human error. We present an automated data-processing procedure that uses a series of smoothing regression splines to model the data and identify outliers. The automated procedure was applied to a set of neutral data and sexual-arousal response data, and artifacts identified were compared to artifacts identified by the standard approach of visual inspection. The automated method showed acceptable accuracy in terms of sensitivity and specificity comparable to the manual-processing method. The automated procedure could reduce human error and data-processing time for studies using vaginal photoplethysmography.