i-Virtual brings camera-based vital signs to MyTwin

Room light and the light from the screen reflect off the skin of the forehead. With each heartbeat, the amount of blood under the skin changes slightly, and so does the light it reflects. The camera records that signal, and algorithms rebuild the pulse wave and derive the measurements from it. Benoît Georis, CEO of i-Virtual, compares it to the finger pulse oximeter used in hospitals, which works on the same principle with light passing through the finger instead of reflected by the face.

Conditions matter: the phone set down and still, the user seated and calm, good light on the face. The technology is designed for spot checks, not continuous monitoring.

Cardiovascular diseases are the world’s leading cause of death, with an estimated 19.8 million deaths in 2022, about 32% of all deaths, according to the World Health Organization. Simple indicators that anyone can check often are one way to bring the heart into everyday prevention. With Saphere, MyTwin shows:

• heart rate;
• heart rate variability, which reflects the balance between the nervous systems of action and rest;
• breathing rate;
• a stress level;
• a cardiovascular health score, computed by Saphere.

Benoît Georis gives the right way to read them: low heart rate variability during exercise is normal; at rest, and repeatedly, it is worth checking with a professional. A score estimates, it doesn’t predict: our article on what a risk score really means explains the difference.

Saphere is not presented as a medical device, and the measurements it gives in MyTwin are indicators, not medical measurements. i-Virtual also develops Caducy, a separate product intended as a medical device for measuring heart rate and respiratory rate, for use under the supervision of healthcare professionals.

The technology has been tested in hospital. In two studies at Nancy University Hospital, funded by i-Virtual and run on 963 patients with its Caducy software, heart rate measured from video agreed with an electrocardiogram for 94.6% of patients, and breathing rate agreed with a reference chest belt for 96% of them.

Reviews of the field are more cautious. Consumer contactless monitors measure heart rate accurately compared with medical devices, but most studies are small and run in laboratory conditions, and motion or poor lighting degrade them. Skin tone matters too: the datasets used to develop rPPG methods often lack diversity in skin tones, and the Nancy study included too few patients with the two darkest skin types to draw conclusions for them. Caducy’s instructions for use contraindicate the measurement for those skin types.

For professionals, the question is where such indicators fit in a care pathway: our guide to remote patient monitoring covers how to build one that works.