Clinic says AI scans spot inflamed heart arteries in 43% of low-risk adults

A US preventive health clinic says that an AI tool reading routine heart scans found inflamed heart arteries in a large share of adults who looked healthy by the usual checks. Biograph announced on 22 October that 42.9% of the adults it studied had raised inflammation around their heart arteries, even though cholesterol tests and risk calculators had put them in the low-risk group.¹ The release, distributed through Business Wire, described this as inflammation that “standard blood tests failed to identify.”² The clinic has not released a scientific paper or said how many people were studied.¹

What happened

Biograph called this its first US study using CaRi-Heart, an AI technology from the company Caristo Diagnostics.¹ The people studied had no symptoms and no known heart disease, and they would have been considered low risk by traditional screening.¹ All of them were members of Biograph’s top membership level, which the clinic calls its Black tier.¹

The tool works on a coronary CT angiography scan, often shortened to CCTA. This is a routine X-ray scan that shows the arteries feeding the heart. Instead of only looking for blockages, the AI measures inflammation in the tissue around those arteries.¹ Biograph combined this with its own scans and blood work, plus a second Caristo tool called CaRi-Plaque that looks at fatty buildup (plaque) and narrowing.¹

The numbers the clinic reported

• 42.9% of participants had raised inflammation around their heart arteries.¹
• 17.9% had very high inflammation in at least one major heart artery.¹
• About 47% had no detectable plaque in their heart arteries yet still showed raised inflammation around them, according to Longevity.Technology’s report.¹ The coverage does not spell out exactly how this figure fits with the 42.9%.

Biograph also compared the scan results with the standard checks doctors use. These included two cholesterol measures (ApoB and non-HDL cholesterol), a long-term blood sugar test (HbA1c), a general blood test for inflammation called hs-CRP, and a widely used heart risk calculator from the American Heart Association, called PREVENT-ASCVD.¹ The clinic reported that none of these matched up with who had inflamed arteries, and none could pick those people out.¹

Dr Michael Doney, Biograph’s executive medical director, said in the announcement:

“Crucially, no routine blood biomarker, lipid metric, or clinical risk score could identify these individuals with inflamed coronary arteries.”

He added that “many individuals with the highest inflammatory levels had completely normal coronary anatomy on CT.”¹

What it does not show yet

The announcement leaves important gaps. It does not give the number of people studied.¹ All the participants came from one clinic’s paying membership, so they may not look like the general public.¹ And the results come from the same clinic that sells the testing, not from an independent team or a peer-reviewed journal.¹

Raised inflammation around an artery is a warning sign that needs interpreting. It is not a diagnosis of heart disease.¹ This study took a single snapshot. It did not follow people over time to see who went on to have heart problems, and it did not test any treatment.

Doney said so himself: “More research is needed to validate this finding and establish the prognostic value and benefit of care directed at coronary perivascular inflammation.”¹ In plain terms, no one yet knows whether finding this inflammation, and acting on it, prevents heart attacks.¹

Who is behind it

Biograph describes itself as a preventive health clinic. It says its mission is to drive advances in human healthspan and lifespan through detailed testing and earlier insight into long-term health risks.¹ The release was issued from San Mateo, California, and New York.² The tools come from Caristo Diagnostics, so both companies have a business interest in the results.

On the regulatory side, Longevity.Technology reports that CaRi-Heart has FDA De Novo marketing authorization. That is a US route for new kinds of medical devices. The tool is described as the first and only product that measures heart artery inflammation from a routine CCTA scan.¹ CaRi-Plaque has FDA clearance for AI-assisted analysis of plaque and narrowing.¹ A February 2025 summary had described the Caristo technology as still pending FDA approval at that time.³

How it fits the field

The method traces back to researchers at the University of Oxford. They developed AI that reads CCTA scans to detect inflammation in the heart arteries.³ The idea is that inflamed arteries release molecules that change the fat around them. The AI measures those changes and turns them into a score called the Fat Attenuation Index.³ Supporters argue this gives a more specific view than hs-CRP, which picks up inflammation anywhere in the body and not just in the heart arteries.³

The earlier Oxford work was different from Biograph’s study in an important way: it tracked what happened to patients. In research presented at the American Heart Association’s 2023 meeting, Professor Charalambos Antoniades’s team looked at more than 40,000 people who had routine heart CT scans at eight UK hospitals. These were mostly people being checked for chest pain, and they were followed for a median of 2.7 years.⁴ Twice as many heart attacks and heart deaths happened in people with no significant narrowing as in people with narrowing.⁴ When an AI risk tool was tested in another 3,393 patients over 7.7 years, people with no blockages but the highest artery inflammation had more than 10 times the risk of heart death of those with lower inflammation.⁴ In a pilot with 744 patients, doctors changed treatment plans in up to 45% of cases after seeing the AI risk scores.⁴

Those results came from a hospital population with symptoms. Biograph’s study looked at a different group: people with no symptoms who were rated low-risk. So the earlier findings on heart attacks and deaths cannot simply be carried over to Biograph’s members.

Other teams also use AI on heart CT scans for different purposes. Cedars-Sinai researchers, for example, built a tool that reads plaque features on these scans to predict reduced blood flow to the heart. It was developed from data on 203 patients.⁵

What to watch next

Doney framed the result as a possible early warning stage: “This study suggests we may be identifying an upstream phase of disease and a window for proactive intervention.”¹ Behind this is the idea that heart disease, whether plaque or narrowing, is “likely a late consequence of an active inflammatory process in the vessel wall,” as he put it.¹ That remains a hypothesis.¹

Several things would make the finding easier to judge:

• A full write-up with the number of participants, their ages and how “raised” inflammation was defined.
• Peer review by outside experts.
• Studies that follow low-risk people with no symptoms over years, to see whether those with inflamed arteries actually have more heart attacks.¹
• Trials showing whether treating the inflammation changes what happens to the heart.¹

Until then, the result is a clinic’s report that many of its apparently low-risk members had something their blood work did not show. What that means for their future health is not yet known.¹