Combined electrocardiogram and photoplethysmogram measurements as an indicator of objective sleepiness

There is considerable interest in unobtrusive and portable methods of monitoring sleepiness outside the laboratory setting. This study evaluates the usefulness of combined electrocardiogram (ECG) and photoplethysmogram (PPG) measurements for estimating psychomotor vigilance. The psychomotor vigilance test (PVT) was performed at various points over the course of a day, and one channel each of ECG and PPG was recorded simultaneously. Features derived from ECG and PPG were entered into multiple linear regression models to estimate PVT values. A double-loop, subject-independent validation scheme was used to develop and validate the models. We show that features obtained from the RR interval were reasonably useful for estimating absolute PVT levels, but were somewhat inadequate for estimating within-subject PVT changes. Combined ECG and PPG measurements appear to be useful for predicting PVT values, and deserve further investigation for portable sleepiness monitoring.