Reflective PPG Sensor Measures Heart Rate Through Clothing During Driving

With heart rate as a key physiological marker for well-being, recent advances in in-vehicle health monitoring systems aim to improve driver safety. Photoplethysmography (PPG) sensors provide a non-intrusive optical method for continuous heart rate tracking. The sensors use Light Emitting Diode (LED) light to detect blood volume changes in the skin, producing a waveform that helps calculate heart rate and other cardiovascular signals.

Integrating these sensors easily into vehicles opens up possibilities for real-time in-vehicle health monitoring, making them a promising tool for mobile health applications beyond traditional clinical settings. Earlier methods, such as steering wheel sensors and camera-based systems using red, green, blue (RGB), or infrared imaging, often faced issues with motion artifacts and changing lighting conditions. Embedding these sensors in car seat backrests has become a more reliable solution, ensuring stable and consistent signal quality even during movement.

The proposed reflective PPG sensor, integrated into the seat backrest, is uniquely designed to monitor heart rate accurately, providing tracking even through clothing. It uses the Beer-Lambert law to detect blood volume changes through light reflection with a design that minimizes motion artifacts, allowing reliable monitoring during driving.

The sensor features a novel topology with a central LED surrounded by photodiodes in a 4.2 cm ring, maximizing light capture from all angles to improve signal quality through clothing and reduce motion artifacts. Data analysis using MATLAB 2023a involved filtering the PPG signal and comparing it to electrocardiogram (ECG) data by detecting peaks and calculating the heart rate over 60-second periods.

Sensor performance was evaluated using different statistical methods like Bland–Altman plots, coverage rates, and gender comparisons. The researchers used distribution functions for environmental changes to identify outliers and improve accuracy and reliability.

Both in controlled and real-world conditions, performance validation showed a strong correlation with reference ECG signals. The sensor achieved a concordance correlation coefficient of 0.89 and a mean absolute percentage error of 3.07% with minimal gender differences in performance.

The Signal-to-Noise Ratio parameter was higher for males (5.68 dB) than for females (3.29 dB), but performance remained consistent across both groups. During a 560-minute drive, the sensor consistently delivered accurate readings, maintaining high alignment with the reference signal, and showed stability and reliability over extended use.

In urban environments, the sensor successfully managed varying noise and lighting conditions, while in rural settings, it showed even more accurate heart rate tracking. Although motion artifacts remain challenging for traditional PPG sensors, this sensor demonstrated robustness, maintaining periodicity during stable driving segments and quickly stabilizing after the motion artifacts subside.

This research confirms the sensor's accuracy and robustness across varied driving conditions, making it well-suited for real-time, seamless heart rate tracking in mobile health apps. The sensor is designed for comfort during long-duration use and can help monitor stress and fatigue. Its reflective, non-intrusive design enables heart rate tracking through clothing, supporting continuous health monitoring in vehicles and other settings like public transport and smart furniture. It shows opportunities for future research to optimize the sensors' potential.