The AI Tool That Can See Coronary Artery Inflammation Before Your Doctor Can

Here's a number that should stop you cold: roughly 50% of people who die from a sudden heart attack had no prior symptoms. No chest pain. No shortness of breath. No warning. Just a normal Tuesday, and then nothing. The frustrating part is that the fire had been burning for years, quietly, inside their coronary arteries — in the form of inflammation that standard cholesterol tests and even most imaging couldn't see.

That's the problem a new generation of AI-powered imaging tools is trying to solve. Specifically, a technology called pericoronary adipose tissue (PCAT) analysis — which uses artificial intelligence to detect coronary artery inflammation from a type of CT scan you may have already had. It's not science fiction. It's FDA-cleared, peer-reviewed, and starting to show up in forward-thinking cardiology practices. And if you care about your heart health in any serious way, you need to know about it.

This article breaks down what the technology is, what the research actually shows, who it's right for, what it won't tell you, and how to actually act on the information once you have it.

What Is Coronary Artery Inflammation, Really?

Most people think of heart disease as a plumbing problem: arteries clog up with cholesterol, blood can't get through, heart attack happens. That model isn't wrong exactly — but it's incomplete. The real story starts with inflammation.

When inflammation takes hold inside the wall of a coronary artery, it destabilizes cholesterol plaques. A stable plaque might sit quietly for decades. An inflamed, unstable plaque can rupture — and a ruptured plaque is what actually causes most heart attacks. Your LDL level tells you how much fuel is in the room. Coronary artery inflammation tells you whether there's a lit match.

The problem is that traditional imaging — even coronary CT angiography (CCTA), which can show the actual anatomy of your arteries — doesn't directly measure that inflammatory signal. You can see a plaque, but you can't easily see whether it's on fire.

How the AI Tool for Coronary Inflammation Works

Here's where it gets clever. Surrounding your coronary arteries is a layer of fat called pericoronary adipose tissue, or PCAT. This fat isn't just padding. It's metabolically active, and when the artery beneath it becomes inflamed, that inflammation changes the composition of the surrounding fat in measurable ways — specifically, by reducing fat cell size and altering the fat's density in a way detectable on a CT scan.

Think of it like a smoke detector. The coronary artery is the room. The PCAT is the air around the room. When there's a fire (inflammation) inside the artery, the smoke (changes in fat density) drifts out into the surrounding space — and the right tool can detect it.

The AI tool most studied here is CaRi-Heart, developed by Caristo Diagnostics at the University of Oxford. It uses a machine-learning algorithm trained on thousands of cardiac CT scans to quantify PCAT density around the major coronary arteries and generate a Cardiac Risk Index (CRI) score. That score reflects the degree of pericoronary inflammation and has been independently validated against actual cardiovascular outcomes.

It doesn't require a new scan. If you've already had a coronary CT angiogram, the AI can re-analyze those existing images. That's the elegant part: the information was always there in the scan data. The AI is just now trained to read it.

What Does the Research Actually Show?

This is where the story gets genuinely interesting — and where you should apply some appropriate skepticism too.

• The CRISP-CT study (2018, Nature Medicine): The Oxford group published a landmark paper showing that the fat attenuation index (FAI) around the right coronary artery independently predicted cardiac mortality over a 5-year follow-up period. People in the highest FAI quartile had a 9-fold higher risk of cardiac death compared to those in the lowest quartile — after adjusting for traditional risk factors, plaque burden, and other confounders. That's not a small signal.
• The ORFAN study (2023, The Lancet): This prospective cohort study followed over 3,900 people and found that the AI-derived Cardiac Risk Index score significantly improved prediction of fatal and non-fatal cardiac events beyond standard risk scores like SCORE2 and traditional CT measures. Crucially, it reclassified a substantial portion of patients previously labeled "low risk" into higher-risk categories — and those patients did, in fact, go on to have more cardiac events.
• FDA clearance (2023): The CaRi-Heart tool received FDA 510(k) clearance, which means it met regulatory standards for safety and performance in the US market. This matters because a lot of longevity diagnostics are sold with research polish but no regulatory teeth. This one cleared the bar.

Here's the honest caveat though: most of this research is observational. The tool is very good at identifying who is at higher risk. What hasn't been proven yet in a randomized trial is whether using this information to guide treatment actually changes outcomes. The assumption is logical — if you know you're inflamed, you treat the inflammation, you reduce the risk — but that chain hasn't been closed in a prospective RCT. Promising, but that gap is real.

The Reality Check: What AI Can and Can't See

Let's be clear about what this tool is and isn't.

It is a risk stratification tool. It tells you something about the biology of your arteries that traditional imaging misses. If your score is high, that's a meaningful signal that deserves attention and clinical follow-up. If it's low, that's genuinely reassuring — though not a hall pass to ignore everything else.

It is not a diagnosis. A high CRI score doesn't mean you're having a heart attack. It doesn't tell you which specific plaques are at risk. It doesn't replace a conversation with a cardiologist who knows your full picture. And it doesn't override the fundamentals: blood pressure, blood sugar, smoking status, LDL, inflammation markers like hs-CRP — all of that still matters enormously.

The internet, predictably, is getting ahead of the science on this one. Some wellness influencers are framing it as the definitive heart disease test. It isn't. It's an important new layer of information in a multi-variable picture. More signal is better. One signal is not the whole answer.

Who Is This Actually Right For?

This technology is most likely to change your clinical decisions if you fall into one of these profiles:

• You've already had a CCTA or coronary calcium scan and want to extract more information from that imaging without additional radiation exposure.
• You're in the "intermediate risk" zone on standard cardiovascular risk calculators (ASCVD score 7.5%-20%) and your clinician is trying to decide whether to start or intensify therapy.
• You have a family history of early cardiovascular disease but your LDL and blood pressure look fine. You're worried there's something under the surface. There might be.
• You have metabolic risk factors — insulin resistance, elevated fasting glucose, visceral adiposity, elevated triglycerides — that are known drivers of vascular inflammation but don't show up dramatically in standard lipid panels.
• You're proactively managing longevity and want a more granular cardiovascular baseline than a 10-year ASCVD score can give you.

If you're 25 with no risk factors, this probably isn't where to start. If you're 45 and you've been meaning to get a real handle on your cardiovascular health, this is now a tool worth knowing about.

Risks and Side Effects: What to Know

The AI analysis itself carries no direct risk — it's a software analysis of existing scan data. The risks to consider are those associated with the CT scan itself:

• Radiation exposure: A CCTA delivers a meaningful radiation dose, typically 3-10 mSv depending on the protocol. Not trivial for annual screening, but appropriate for clinical decision-making in intermediate-to-high risk individuals.
• Contrast dye: CCTA requires iodinated contrast, which carries a small risk of allergic reaction and kidney stress in people with impaired renal function.
• Anxiety from findings: Knowing your pericoronary inflammation score is only useful if you have a plan. Getting the information without a clinician to help you interpret and act on it is stressful and counterproductive.
• Over-treatment risk: A high score might prompt aggressive statin or aspirin therapy in someone where the risk-benefit calculation isn't clearly in favor. This is why clinical context is essential.

What You Actually Do With This Information

This is where a lot of people get stuck. You find out your coronary artery inflammation is elevated. Now what?

The evidence-based levers for reducing vascular inflammation are well-established, even if they're not glamorous:

• Lipid management: Statins reduce both LDL and vascular inflammation. If your score is high, this conversation gets urgent.
• Blood sugar and insulin control: Hyperglycemia and insulin resistance are major drivers of vascular inflammation. Continuous glucose monitoring, dietary changes, and medications like Metformin or SGLT2 inhibitors can make a significant dent here.
• Weight management: Visceral fat is pro-inflammatory. Reducing it reduces the inflammatory burden on your arteries. GLP-1 receptor agonists have shown meaningful anti-inflammatory cardiovascular effects beyond their weight-loss properties.
• Exercise and sleep: Not optional. Both are among the most potent anti-inflammatory interventions available, and neither requires a prescription.
• Targeted longevity interventions: Some longevity-oriented compounds, including rapamycin, metformin, and certain SGLT2 inhibitors, have mechanistic and clinical evidence for reducing systemic inflammation and improving cardiovascular risk markers.

How Healthspan Can Help You Act on This

Knowing your coronary artery inflammation score is step one. Step two is actually doing something evidence-based about it — and that's where the gap between information and action tends to open up.

At Healthspan, the approach starts with your biology, not a protocol. The Longevity Optimization program builds a full metabolic and cardiovascular baseline: comprehensive labs including hs-CRP, fasting insulin, lipid particle analysis, HbA1c, and more. From there, a Healthspan clinician helps you map the findings to a personalized intervention plan.

If your labs and imaging suggest metabolic-driven vascular inflammation, the SGLT2 Protocol may be clinically relevant: SGLT2 inhibitors like canagliflozin have demonstrated cardiovascular outcome benefits in large trials, and they work partly through anti-inflammatory mechanisms. If blood sugar dysregulation is in the picture, Metformin has decades of safety data and real evidence for reducing cardiovascular risk markers. If weight is a factor, GLP-1 Longevity Care addresses both adiposity and direct vascular inflammation through mechanisms that go well beyond the scale.

Every Healthspan protocol includes medical supervision, labs, follow-up consultations, and dose adjustments over time. It's not a supplement stack you order online. It's a clinical relationship built around your actual numbers.

If cardiovascular inflammation is something you're taking seriously, the right first step is a conversation with a clinician who can look at your full picture. Schedule a consultation with Healthspan and start there.

Frequently Asked Questions About Coronary Artery Inflammation AI

What is the AI test for coronary artery inflammation?

The most studied tool is CaRi-Heart, developed at the University of Oxford. It uses artificial intelligence to analyze the density of fat surrounding your coronary arteries on a CT scan — a measure called the fat attenuation index (FAI). Changes in that fat signal inflammation in the arterial wall. The tool generates a Cardiac Risk Index score that predicts cardiovascular risk beyond what standard imaging or blood tests can show. It received FDA clearance in 2023.

Do I need a new scan to use coronary inflammation AI?

Not necessarily. If you've already had a coronary CT angiogram (CCTA), the AI software can analyze those existing images without additional imaging or radiation exposure. The scan data has always contained this information — the AI is trained to extract it. If you haven't had a CCTA, you'll need one performed at a center that runs the AI analysis.

How accurate is AI at detecting coronary artery inflammation?

In the ORFAN study published in The Lancet in 2023, the AI-derived Cardiac Risk Index independently predicted fatal and non-fatal cardiac events in a prospective cohort of over 3,900 people, beyond standard risk scores. In earlier research, high pericoronary fat attenuation was associated with a 9-fold higher risk of cardiac death over 5 years. That's a meaningful signal, though it's a risk stratification tool, not a definitive diagnostic.

What causes high coronary artery inflammation?

The main drivers are metabolic: insulin resistance and elevated blood sugar, visceral adiposity, dyslipidemia (especially high small dense LDL and triglycerides), hypertension, smoking, chronic systemic inflammation, and in some cases, autoimmune conditions. Lifestyle factors including poor sleep, sedentary behavior, and an ultra-processed food diet amplify all of the above. Genetic predisposition plays a role too, which is why some people with "normal" cholesterol still have significant coronary inflammation.

Can coronary artery inflammation be reversed?

The evidence suggests yes, at least partially. Interventions that reduce systemic inflammation — including statins, SGLT2 inhibitors, metformin, GLP-1 receptor agonists, aggressive blood pressure control, weight loss, and sustained aerobic exercise — have all shown measurable improvements in vascular inflammatory markers. Whether the pericoronary fat attenuation index itself normalizes with treatment is an active area of research, but the biological mechanisms strongly suggest it should.

Is coronary artery inflammation the same as atherosclerosis?

They're related but not the same thing. Atherosclerosis is the buildup of plaque inside the arterial wall. Coronary artery inflammation is the inflammatory process that drives plaque formation and, more dangerously, plaque instability. You can have atherosclerosis with low inflammation (more stable) or high inflammation (more dangerous). The AI tool specifically measures the inflammatory component, which is why it adds information beyond what a coronary calcium score or standard CCTA can show.

How do I get access to the coronary inflammation AI test?

Access currently requires a CCTA at a center that either partners with Caristo Diagnostics (makers of CaRi-Heart) or uses equivalent AI software. Availability is expanding as the technology gains adoption. Your best starting point is a cardiologist or longevity clinician who can order the appropriate imaging and help you interpret the results in the context of your full cardiovascular risk profile — not just the score in isolation.