Robust In-Vehicle Respiratory Rate Detection Using Multimodal Signal Fusion

Early detection of health issues is critical to ensuring timely and effective medical care. Our study aimed to improve the assessment of a patient's health by utilizing an in-vehicle health monitoring system that incorporates primary vital signs such as respiratory rate. To achieve this, we designed a redundant sensor system composed of an accelerometer, a piezoelectric sensor, and a camera for image photoplethysmogram (iPPG). The sensor system was tested on 15 subjects under four different conditions: rest, city driving, highway driving, and rural driving, with a total recording time of 5 minutes for rest and 15 minutes for each of the driving conditions. The recorded signals and ground truth data were analyzed and are now publicly available, offering a valuable resource for the reproduction of our results and the improvement of existing algorithms for health assessments.