Contactless Measurement of Heart Rate, Heart Rate Variability and Breathing Rate Using Remote Photoplethysmography

Contactless Measurement of Heart Rate, Heart Rate Variability and Breathing Rate Using Face Video Analysis (Remote Photoplethysmography) - Validation of Shen.AI Vitals Software.

The purpose of this study it to evaluate the accuracy of heart rate, heart rate variability and breathing rate measurement with the use of Shen.AI Vitals software developed by MX Labs.

Study Overview

• Status: Completed
• Conditions: Heart Rate Variability; Heart Rate; Respiratory Rate
• Intervention / Treatment: Diagnostic test: measurement of physiological parameters wih Shen.AI Vitals software on various camera-equipped devices (smartphones, tablets, laptops)

Detailed Description

Resting heart rate (HR), heart rate variability (HRV) and breathing rate (BR) will be measured (estimated) with the use of remote photoplethysmography (1-minute face video analysis with Shen.AI Vitals software).

The measurements will be taken with various camera-equipped devices (smartphones, tablets, laptops) using either front camera (self-measurement) or rear camera (measurement made by someone else), in two different lighting conditions (artificial light or natural daylight), with various levels of device stability (fully stable or held in hand/s), and in two different body positions (sitting or supine). In total, each participant will have 8 measurements taken.

Standard deviation of normal heartbeat intervals (SDNN) will be used as the main HRV index.

Apart from the average HR (averaged over the whole 1-minute measurement), instantaneous values of HR (averaged over 10 s) will also be determined and displayed every 1 s by the Shen.AI VItals software.

In all cases, reference measurements will be taken simultaneously (chest ECG for the calculation of average and instantaneous HR as well as HRV; impedance pneumography for the determination of BR).

The reference measurements will be taken with Finapres Nova with the ECG+Resp module (Finapres Medical Systems B.V.).

For each measured parameter, the results from both methods will be compared using the Bland-Altman analysis or non-parametric limits of agreement.

• Study Type: Interventional
• Enrollment (Actual): 136
• Phase: Not Applicable

Contacts and Locations

• Study Contact: Name: Przemysław Jaworski; Phone Number: +48691033779; Email: pjaworski@mxlabs.ai

Study Locations

• Poland
  • Dolnośląskie
    • Wrocław, Dolnośląskie, Poland, 50-556
      • Uniwersyteckie Centrum Wsparcia Badań Klinicznych we Wrocławiu

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Adult; Older Adult
• Accepts Healthy Volunteers: Yes

Description

Inclusion Criteria:

• age ≥ 18 years
• ability to read and write in Polish or English
• ability to operate a smartphone, tablet and laptop without the need for glasses
• ability to understand the study procedure
• ability to communicate and follow the instructions of the study staff
• written consent to participate in the study and to the processing of personal data by the research team and the sponsor of the study

Exclusion Criteria:

• failure to provide the required information in the participant's information survey
• failure to comply with the study procedure or the instructions of the study stuff
• extensive facial skin damage (including abrasions, wounds, burns)
• a disease process involving a large part of the face that makes the video- based measurement impossible
• extensive face dressing
• extensive facial skin tattoo or face painting
• marked deformation of the face (e.g. a tumor)
• inability to keep the head stable in the required position during the measurement
• persistent cough making it impossible to remain still during the measurement
• respiratory dysfunction in the form of shortness of breath or irregular or shallow breathing
• diagnosed arrhythmia (except sinus bradycardia or tachycardia)
• moderate or severe anemia
• diagnosed: heart failure, left ventricular systolic dysfunction, aortic stenosis or other diseases or disorders within the heart vascular system leading to decreased stroke volume, low pressure amplitude or the so-called paradoxical pulse

Study Plan

How is the study designed?

Design Details

• Primary Purpose: Diagnostic
• Allocation: N/A
• Interventional Model: Single Group Assignment
• Masking: None (Open Label)

Number of Arms

1

Arms and Interventions

Participant Group / Arm

Intervention / Treatment

Experimental: HR, HRV and BR measured with Shen.AI and reference method

Diagnostic test: measurement of physiological parameters wih Shen.AI Vitals software on various camera-equipped devices (smartphones, tablets, laptops); Resting heart rate (HR), heart rate variability (HRV) and breathing rate (BR) will be measured (estimated) with the use of remote photoplethysmography (1-minute face video analysis with Shen.AI Vitals software). Reference measurements will be taken simultaneously (chest ECG for the calculation of HR and HRV; impedance pneumography for BR).

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Accuracy of heart rate (HR) measurement with the software under test compared with the ECG-based measurement	In each participant, average HR will be estimated from 1-minute video-based remote photoplethysmography of the facial skin (software under test) and calculated from simultaneously recorded ECG. Measurements will be made in resting conditions (seated position), with artificial lighting and a fully stabilized camera device. The results from both methods will be compared using non-parametric limits of agreement.	day 1 (single visit, approximately 30 minutes)
Accuracy of heart rate variability index (SDNN) measurement with the software under test compared with the ECG-based measurement	In each participant, the standard deviation of normal heartbeat intervals (SDNN) will be estimated from 1-minute video-based remote photoplethysmography of the facial skin (software under test) and calculated from simultaneously recorded ECG. Measurements will be made in resting conditions (seated position), with artificial lighting and a fully stabilized camera device. The results from both methods will be compared using the Bland-Altman analysis.	day 1 (single visit, approximately 30 minutes)
Accuracy of breathing rate (BR) measurement with the software under test compared with the impedance pneumography-based measurement	In each participant, the breathing rate will be estimated from 1-minute video-based remote photoplethysmography of the facial skin (software under test) and calculated from simultaneously recorded impedance pneumography signal. Measurements will be made in resting conditions (seated position), with artificial lighting and a fully stabilized camera device. The results from both methods will be compared using the Bland-Altman analysis.	day 1 (single visit, approximately 30 minutes)
Accuracy of instantaneous heart rate measurement with the software under test compared with the ECG-based measurement	In each participant, multiple instantaneous heart rate values (averaged over 10 s) will be estimated from the software under test and calculated from simultaneously recorded ECG. Measurements will be made in resting conditions (seated position), with artificial lighting and a fully stabilized camera device. The results from both methods will be compared using non-parametric limits of agreement.	day 1 (single visit, approximately 30 minutes)

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Accuracy of measurement of heart rate (HR) with the software under test on a smartphone held by someone else (compared with reference measurements).	Each participant will have one video-based measurement performed with the software under test on a smartphone held by someone else (facial video measurement of HR using the rear camera). For reference, HR will be calculated from the simultaneously recorded ECG signal. The results from both methods will be compared using non-parametric limits of agreement.	day 1 (single visit, approximately 30 minutes)
Accuracy of measurement of heart rate variability index (SDNN - the standard deviation of NN intervals) with the software under test on a smartphone held by someone else (compared with reference measurements).	Each participant will have one video-based measurement performed with the software under test on a smartphone held by someone else (facial video measurement of SDNN using the rear camera). For reference, SDNN will be calculated from the simultaneously recorded ECG signal. The results from both methods will be compared using the Bland-Altman analysis.	day 1 (single visit, approximately 30 minutes)
Accuracy of measurement of breathing rate (BR) with the software under test on a smartphone held by someone else (compared with reference measurements).	Each participant will have one video-based measurement performed with the software under test on a smartphone held by someone else (facial video measurement of BR using the rear camera). For reference, BR will be obtained from the simultaneously recorded impedance pneumography signal. The results from both methods will be compared using the Bland-Altman analysis.	day 1 (single visit, approximately 30 minutes)
Accuracy of measurement of heart rate (HR) with the software under test on a smartphone held in hand/s (compared with reference measurements).	Each participant will have one video measurement performed with the software under test on a smartphone held in hand/s (facial video measurement of HR using the rear camera). For reference, HR will be calculated from the simultaneously recorded ECG signal. The results from both methods will be compared using non-parametric limits of agreement.	day 1 (single visit, approximately 30 minutes)
Accuracy of measurement of heart rate variability index (SDNN - the standard deviation of NN intervals) with the software under test on a smartphone held in hand/s (compared with reference measurements).	Each participant will have one video measurement performed with the software under test on a smartphone held in hand/s (facial video measurement of SDNN using the rear camera). For reference, SDNN will be calculated from the simultaneously recorded ECG signal. The results from both methods will be compared using the Bland-Altman analysis.	day 1 (single visit, approximately 30 minutes)
Accuracy of measurement of breathing rate (BR) with the software under test on a smartphone held in hand/s (compared with reference measurements).	Each participant will have one video measurement performed with the software under test on a smartphone held in hand/s (facial video measurement of BR using the rear camera). For reference, BR will be obtained from the simultaneously recorded impedance pneumography signal. The results from both methods will be compared using the Bland-Altman analysis.	day 1 (single visit, approximately 30 minutes)
Accuracy of measurement of heart rate (HR) with the software under test in daylight conditions (compared with reference measurements).	Each participant will have one facial video measurement performed in daylight conditions (1-minute measurement of HR using the front smartphone camera). For reference, HR will be calculated from the simultaneously recorded ECG signal. The results from both methods will be compared using non-parametric limits of agreement.	day 1 (single visit, approximately 30 minutes)
Accuracy of measurement of heart rate variability index (SDNN - the standard deviation of NN intervals) with the software under test in daylight conditions (compared with reference measurements).	Each participant will have one facial video measurement performed in daylight conditions (1-minute measurement of SDNN using the front smartphone camera). For reference, SDNN will be calculated from the simultaneously recorded ECG signal. The results from both methods will be compared using the Bland-Altman analysis.	day 1 (single visit, approximately 30 minutes)
Accuracy of measurement of breathing rate (BR) with the software under test in daylight conditions (compared with reference measurements).	Each participant will have one facial video measurement performed in daylight conditions (1-minute measurement of BR using the front smartphone camera). For reference, BR will be obtained from the simultaneously recorded impedance pneumography signal. The results from both methods will be compared using the Bland-Altman analysis.	day 1 (single visit, approximately 30 minutes)
Accuracy of measurement of heart rate (HR) with the software under test in the supine position (compared with reference measurements).	Each participant will have one facial video measurement performed in the supine position (1-minute measurement of HR using the front smartphone camera). For reference, HR will be calculated from the simultaneously recorded ECG signal. The results from both methods will be compared using non-parametric limits of agreement.	day 1 (single visit, approximately 30 minutes)
Accuracy of measurement of heart rate variability index (SDNN - the standard deviation of NN intervals) with the software under test in the supine position (compared with reference measurements).	Each participant will have one facial video measurement performed in the supine position (1-minute measurement of SDNN using the front smartphone camera). For reference, SDNN will be calculated from the simultaneously recorded ECG signal. The results from both methods will be compared using the Bland-Altman analysis.	day 1 (single visit, approximately 30 minutes)
Accuracy of measurement breathing rate (BR) with the software under test in the supine position (compared with reference measurements).	Each participant will have one facial video measurement performed in the supine position (1-minute measurement of BR using the front smartphone camera). For reference BR will be obtained from the simultaneously recorded impedance pneumography signal. The results from both methods will be compared using the Bland-Altman analysis.	day 1 (single visit, approximately 30 minutes)
Utility of the software under test for self-measurement of physiological parameters (assessed with a questionnaire).	The utility of the software under test for self-measurement of physiological parameters (HR, SDNN, and BR) will be assessed based on of the average score in the participant's questionnaire completed by each participant after the test procedure (i.e. after all measurements in the given participant). The scale will be 1-5, with higher scores meaning better outcome.	day 1 (after a single visit, approximately 30 minutes)
Utility of the software under test for the measurement of someone's physiological parameters (assessed with a questionnaire).	The utility of the software under test for the video-based measurement of someone's physiological parameters will be assessed based on the average score in the participant's questionnaire (the part of the questionnaire related to the measurement with the rear camera of a smartphone) completed by each participant after the test procedure(i.e. after all measurements in the given participant). The scale will be 1-5, with higher scores meaning better outcome.	day 1 (after a single visit, approximately 30 minutes)

Collaborators and Investigators

• Sponsor: MX LABS spółka z ograniczoną odpowiedzialnością oddział w Polsce
• Investigators: Principal Investigator: Bartłomiej Paleczny, Wrocław Medical University

Study record dates

Study Major Dates

• Study Start (Actual): April 26, 2023
• Primary Completion (Actual): June 6, 2023
• Study Completion (Actual): June 6, 2023

Study Registration Dates

• First Submitted: April 17, 2023
• First Submitted That Met QC Criteria: June 2, 2023
• First Posted (Actual): June 5, 2023

Study Record Updates

• Last Update Posted (Actual): March 28, 2024
• Last Update Submitted That Met QC Criteria: March 27, 2024
• Last Verified: June 1, 2023

More Information

Terms related to this study

• Other Study ID Numbers: MDF-01-008-01

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No