Evaluating the precision of contactless heart rate monitoring

This pilot study evaluates the accuracy of facial video photoplethysmography (vPPG) used in the Shen AI’s technology in measuring heart rate (HR) using only a smartphone camera.

As contactless monitoring becomes essential in telemedicine and digital health, assessing vPPG's performance across different time windows is critical, especially its ability to provide reliable, quasi-instantaneous (fast-responding) measurements. The research assessed Shen AI's vPPG results against ECG, the clinical gold standard, in controlled conditions.

‍

Key insights: Unprecedented short-window accuracy

The study validates Shen AI’s vPPG as a highly accurate tool for contactless heart rate monitoring, demonstrating exceptional performance across long and short measurement windows:

• Benchmark 60-second heart rate: Achieved near-perfect accuracy with a Mean Absolute Error (MAE) of just 0.1 bpm. 100% of measurements were within 1 bpm of the reference ECG.
• Reliable 10-second heart rate: For short-window averages, the MAE was only 0.2 bpm, with 99.8% of values falling within 1 bpm.
• Validation of quasi-instantaneous (4-second) HR: This is the first vPPG study to validate performance at this speed. Even for these fast-responding values, the MAE was 0.4 bpm, with 94.5% of results within 1 bpm.
• High correlation: Correlation coefficients exceeded 0.99 across all time windows.

‍

Research summary

The following is an excerpt of the “Facial video photoplethysmography for measuring average and quasi-instantaneous heart rate: a pilot validation study” paper.

‍

Background: Video photoplethysmography (vPPG) is a contactless optical technique for recording blood pulsations in the blood vessels of the skin using a digital camera that is increasingly used to measure or estimate various physiological parameters. In this study, we evaluated the accuracy of average and quasi-instantaneous heart rate (HR) measurements performed via facial vPPG technology Shen.AI Vitals and a smartphone camera.

Methods: We studied 35 healthy volunteers in a seated position (median age 25 years, 17 females). Video recordings of participants’ faces were obtained using the front camera of a smartphone mounted on a tripod. In parallel, a 1-lead chest electrocardiogram (ECG) was recorded to obtain reference HR values (average value from the entire 60-s measurement and multiple values averaged over 10-s or 4-s periods during the measurement).

Results: The mean absolute errors were 0.1, 0.2, and 0.4 beats per minute (bpm) for HR averaged over 60-s, 10-s, and 4-s periods, respectively. The errors did not exceed 1 bpm in 100.0%, 99.8%, and 94.5% of the cases, respectively. For the latter, our sample included almost 1,900 HR values from a relatively wide range (46–117 bpm). Regardless of the HR averaging time, the correlation between the vPPG-based and reference values was very strong (r > 0.99, P < 0.001).

Conclusion: In predominantly young, white, seated subjects, the tested vPPG technology provided highly accurate HR measurements, both when the values were averaged over 60 s and in the case of short-term values averaged over 10 s or quasi-instantaneous values averaged over 4 s. To our knowledge, this is the first study on vPPG technology to examine quasi-instantaneous HR measurements (averaged over periods shorter than 5 s). The results should be confirmed in a larger study with greater diversity in age, skin tone, and lighting conditions.

‍

Research details

Title: Facial video photoplethysmography for measuring average and quasi-instantaneous heart rate: a pilot validation study
Authors: Leszek Pstras, Tymoteusz Okupnik, Beata Ponikowska, Bartlomiej Paleczny
Institutions: Nalecz Institute of Biocybernetics and Biomedical Engineering, Polish Academy of Sciences & Wroclaw Medical University
Published September 24, 2025

‍

Read the full study here.

‍