Measuring vital signs with a camera: what video and voice bring to prevention

e-sensia started with the voice, at the heart of French emergency medical dispatch (the SAMU emergency line and SOS Médecins). Its conviction, carried by Dr Jean-Baptiste Perney, emergency physician and co-founder: over the phone, the voice is a clinical data point in its own right. Breathing, tone and speaking rate reveal something about a patient’s condition before they even describe their symptoms.

Its dispatch tools analyze the voice in real time to give the dispatching physician an additional signal, presented as decision support and never as a substitute for their judgment.

In July 2026, e-sensia acquired i-Virtual’s assets, including the Saphere technology. The company thus adds a second modality to the voice: contactless measurement of vital signs from a video. Bringing voice and vision together is a step toward remote clinical assessment that combines several signals.

Saphere relies on remote photoplethysmography, or rPPG. With every heartbeat, the inflow of blood very slightly changes the color of the skin. These changes are invisible to the naked eye, but a camera can record them, and an algorithm can reconstruct a pulse signal from them.

All it takes is a video of about 30 seconds, facing the camera, with no watch or sensor. e-sensia describes Saphere as measuring heart rate, respiratory rate, heart rate variability and stress-related indicators.

i-Virtual states that its technology has been evaluated in two clinical studies involving more than 1,200 patients, including a range of skin phototypes and complex medical conditions.

rPPG is designed for spot-checks, not continuous monitoring: i-Virtual itself notes that continuous monitoring calls for dedicated medical devices. Its value lies in repetition. A measurement taken regularly, under comparable conditions, makes it possible to observe a trend.

Measurement quality also depends on conditions: lighting, movement, face position. A study published in npj Digital Medicine compared several rPPG methods with a contact reference device in participants with different skin tones. The mean heart rate difference was about 1 beat per minute, but the variability of measurements around the reference reached about 11 beats per minute. The authors highlight the influence of the environment and of individual characteristics, and the need to validate these methods in real-world conditions.

The second building block from e-sensia reverses the usual logic of health apps. The patient does not have to open the app: MyTwin calls them. If they turn the option on, they receive a short follow-up call at regular intervals, for example every day.

On the line, the assistant asks a few questions about how the person feels and about their follow-up program. Voice, breathing and cough are also analyzed to spot what changes from one call to the next. What the voice can reveal about health is covered in our article on voice biomarkers.

The main benefit is regularity: a call that arrives without having to think about it keeps follow-up going over time, where an app often ends up forgotten. It is not, however, an emergency service. In case of chest pain, difficulty breathing or signs of a stroke, call your local emergency number immediately.

At MyTwin, these two technologies join the other data sources of the patient digital twin: medical documents, lab results, connected devices, questionnaires. A single heart rate reading says little. Placed alongside sleep, activity, treatments or the answers given during a follow-up call, it helps understand how a person’s health evolves.

These repeated measurements, taken in everyday life rather than during a consultation, are part of what is known as real-world data. They do not replace a clinical examination, but they can shed light on what happens between two appointments.

It is also a simple way into prevention. Claire’s story, about not knowing where to start, shows how a first cardiovascular health score can kick off a broader approach.

To learn more about e-sensia’s work, visit the e-sensia website, and watch the full conversation with Benoît Georis in the MyTwin Inside episode on camera-based measurement.