Cardiovascular disease is responsible for an estimated 17.9 million deaths each year, making it the leading cause of death worldwide. Continuous heart health monitoring way is the best preventative for heart failure. However, standard monitoring devices, including ECG and PPG, require direct physical contact, which can be difficult for certain patient populations, particularly infants and those with skin sensitivities or injuries. Non-contact monitoring devices, including remote imaging photoplethysmography (RIPPG) and ballistocardiography (BCG) can be hindered by clothing and micromotion.
Researchers at Arizona State University have developed a novel non-contact, non-obstructive approach to extract cardiac pulses by analyzing chest motion in depth videos using time-of-flight (ToF) cameras. Utilization of a ToF cameras enables higher dynamic range and accuracy in micromotion reconstruction compared to other non-contact imaging modalities. Pre-trained computer vision models are employed to identify and track human torso landmarks and create chest ROI. Then Robust Principal Component Analysis (RPCA) is applied to isolate the cardiac signal from other body movements by suppressing unwanted motion interference from breathing motion and involuntary body motion. This method was tested on ten healthy human subjects and demonstrated high accuracy.
This novel, non-contact approach provides a sensitive and accurate means for cardiac sensing, even under variable conditions, and could be quite advantageous in remote health monitoring applications.
Potential Applications
- Remote health monitoring
- Particularly for patients with cardiovascular diseases
- Multi-subject environments such as care facilities
Benefits and Advantages
- Non-contact – safe and easy to use, even on sensitive or young patients
- Privacy preserving
- ToF depth cameras enable higher dynamic range and reconstruction of the observed motion
- Can be used for multi-subject monitoring
- Automatically identifies human torsos in the depth videos
- Capable of operating under variable conditions (crowded environment, clothing/blockage thickness, peripheral/other measurement sites, distance)
- When tested, the depth-based HR estimation accuracy achieved 100% within 5 BPM
- Provides inter-beat-interval (IBI) variations
- Valuable information for medical professionals
- The use of depth pixels directly helps with micromotion estimation
- Coherent processing techniques suppress unwanted motion interference (breathing motion and involuntary body motion)
- Able to obtain clean cardiac pulse signal
For more information about this opportunity, please see
For more information about the inventor(s) and their research, please see