A Five-Second Facial Video Could Soon Detect High Blood Pressure and Diabetes

A Camera Glance That Could Change How Millions Are Diagnosed

No blood draw. No cuff around your arm. No clinic appointment. Just five seconds of video of your face — and an AI that can tell whether you have high blood pressure or diabetes. That's the striking promise of new research presented at the European Society of Cardiology (ESC) Congress, held August 28–31 in Munich, Germany. The findings are drawing serious attention from cardiologists and digital health experts around the world.

Researchers at the University of Tokyo and Institute of Science Tokyo investigated whether AI analysis of facial videos could be used to improve diagnosis of hypertension and diabetes. Their goal was ambitious but practical. As lead researcher Ryoko Uchida put it, "We aimed to develop an AI algorithm that enables contactless screening in everyday environments to detect common conditions earlier and at scale."

What the AI Actually Sees — and How It Works

A prospective single-centre study recruited 215 participants, involving both diagnosed patients and healthy volunteers. Each participant underwent a short, high-speed video recording of their face and palms using a spectroscopic camera. The key is what the algorithm extracts from that footage — information invisible to the naked eye.

A machine-learning algorithm analysed the individual videos and extracted data on pulse-wave dynamics (which measure the stiffness of arteries), skin blood-flow patterns and the spectral characteristics of skin colouring. In other words, the AI is reading subtle biological signals encoded in how light interacts with the skin — signals that reflect the underlying state of a person's cardiovascular and metabolic health.

The algorithm detected hypertension with 95% accuracy from a 30-second recording. Sensitivity to detect normal blood pressure was 100%, while hypertension sensitivity was 89.2%. The algorithm's accuracy from a five-second video remained high at 90.3%. For diabetes, the algorithm was able to detect the condition with 88.2% accuracy from a 30-second video and 81.2% from a five-second video.

Why This Matters on a Global Scale

Approximately 1.4 billion adults aged 30–79 years are estimated to have hypertension, and 589 million people live with diabetes worldwide. Although hypertension and diabetes are among the leading modifiable risk factors for cardiovascular disease, many cases remain undiagnosed. That diagnostic gap has enormous consequences — untreated high blood pressure and diabetes silently damage the heart, kidneys, and blood vessels for years before symptoms appear.

Conventional screening usually requires blood pressure measurements, blood tests or dedicated healthcare visits. Those barriers — cost, access, time — mean millions of people never get screened at all. If validated, this contactless approach could allow people to be screened in everyday settings — without cuffs, blood sampling or a dedicated clinic visit — helping to identify at-risk individuals who would otherwise remain undiagnosed and therefore untreated.

Excitement Tempered by the Need for More Validation

The response from the broader medical community has been enthusiastic but measured. Associate Professor Nico Bruining, programme co-chair of the ESC Digital and AI Summit, called the development remarkable, adding that "because this approach is quick, easy and contactless, it could be used in many settings beyond hospitals, giving it the potential to reach far more people than traditional screening methods."

Uchida noted that the team's machine-learning algorithm accurately detected hypertension and diabetes from facial spectroscopic video recordings as short as five seconds, and that they intend to validate these findings in larger cohorts across more diverse populations to support real-world application. Both this approach and related AI imaging efforts could eventually move screening beyond hospitals and clinics, although researchers say larger and more diverse validation studies are still needed.

The implications extend well beyond hospitals. Imagine a pharmacy kiosk, a workplace wellness station, or even a smartphone app capable of flagging cardiovascular risk in seconds. The technology processes spectroscopic facial video to detect hemodynamic and microvascular signatures linked to hypertension and diabetes — a capability that, once miniaturized and validated at scale, could fundamentally reshape how the world's most common silent killers are caught before they cause irreversible harm.