Smartphone PPG Validation for a Depression Assessment Protocol

A new smartphone application for pulse photoplethysmography (PPG) recording is validated for a depression assessment protocol, consisting of four stages that includes recordings of movement-related tasks, talking-related tasks, and stillness. The separation of these stages allows to evaluate the agreement between the PPG acquired using the smartphone camera and a commercial pulse oximeter, used as reference, in different conditions. Several time-domain and frequency-domain heart rate variability (HRV) metrics have been studied. Results suggest a strong agreement during stillness phases
$(\mathbf{r}\geq \mathbf{0}.\mathbf{96}\ (\mathbf{p} < \mathbf{0}.\mathbf{05})$
for all the HRV metrics), as well as a moderate downgrade during movement and talking. The agreement is also different between low-frequency related metrics
$(\mathbf{r}\geq \mathbf{0}.\mathbf{92}\ (\mathbf{p} < \mathbf{0}.\mathbf{05})$
and high-frequency related metrics
$(\mathbf{r} < \mathbf{0}.\mathbf{78}\ (\boldsymbol{p} < \mathbf{0}.\mathbf{05}))$
. The overestimation of the high-frequency component of the HRV may be a limitation for smartphone PPG in depression monitoring.