Remote monitoring of vital physiological signals using smartphone during controlled and uncontrolled respiratory conditions

Indeed, photoplethysmography (PPG)-based health monitoring devices had played a purposeful role in clinical and non-clinical environments for the past few years. The produced devices which invoked this optical technique provided favorable applications over the traditional vital signals measuring devices. The former devices, such as wearable devices, were comfortable for health care monitoring during daily activities. Then, imaging PPG (iPPG) emerged as a contactless optical technique. With the aid of video cameras, the reflection of subtle changes in blood circulation on skin color was used to measure heart rate (HR). In the meantime, smartphones have widely penetrated our personal lives accompanied by the continuous renovation of embedded sensors and communication technologies. In that case, the smartphone could be an inviting low-cost technology for continuous remote health monitoring and a potential for a wide range of remote cardiovascular applications. Hence, there is a substantial need to explore the signal processing techniques and applications for reliable cardiovascular measures using a smartphone camera. In this thesis, we explored the smartphone camera to develop a processing scheme for estimating the HR information from iPPG signals extracted from a smartphone facial video sequence. We adopted the plane-orthogonal-to-skin (POS) method to compute iPPG. The proposed method is evaluated by applying it to extract the HR of subjects standing at rest and during two motion conditions. Meanwhile, they were performing several respiratory maneuvers. Automatic face detection algorithms were implemented in the proposed method with steps of filtration processes to eliminate motion artifacts and for the reconstruction of a clean iPPG signal. Then, for the same previous experimental setup, we investigated iPPG- based pulse rate variability (PRV) as a noninvasive marker to evaluate the functionality of the autonomic nervous system (ANS) and as a surrogate for electrocardiogram (ECG)- based heart rate variability (HRV). We incorporated signal quality and statistical analysis methods to evaluate the accuracy of our results because signals contained variability of linear contents and nonlinear fluctuations impact. In another part of the thesis, we proposed a remote smartphone camera-based and noninvasive blood pressure (BP) monitoring technique from subjects’ hand palms. iPPG signals were derived from pixel values of consecutive video frames. Based on the attained waveform features, we used the pulse transit time (PTT) between both palms' iPPG signals to estimate the subjects' systolic blood pressure (SysBP) and diastolic blood pressure (DysBP). Our proposed systems were validated and compared against gold-standard reference measures. All measurements were conducted using a remote smartphone that captured short videos and without any physical contact between subjects and further sensors. This work provides alternative solutions for health tracking under regular living conditions and promotes the management of cardiovascular diseases in out of clinic settings.

Restricted until 06/2023. To request the author grant access, click on the PDF link to the left.