Spotting Silent Heart Threats: Is AI the Future of Plaque Detection?

Artificial intelligence is rapidly moving into cardiology, offering doctors an unprecedented look at the hidden dangers within our blood vessels. By analyzing advanced imaging, these AI tools can map out the precise volume and characteristics of arterial blockages. However, a critical question remains: will this high-tech data actually translate into fewer heart attacks?

The Hidden Danger: Hard vs. Soft Plaque

Atherosclerosis, frequently referred to as "hardening of the arteries," occurs when fatty deposits build up along arterial walls. But the nickname is somewhat misleading. While some of this buildup calcifies and hardens over time, a significant portion remains soft and unstable.

From a medical standpoint, soft plaque poses a much greater threat. Because it is unstable, it is far more prone to rupturing. When a rupture occurs, it can trigger a sudden blood clot, completely blocking blood flow and causing a heart attack

To visualize these different deposits, doctors use a specialized imaging test called coronary computed tomography angiography (CCTA). A CCTA scan is highly effective at identifying both hard and soft plaque, as well as tracking blockages in the vessels that supply the heart.

Building on this technology, several medical tech firms have introduced AI algorithms designed to deep-dive into CCTA scans. These programs automatically measure and quantify the amount of each type of plaque. This advanced insight comes with a premium: major private insurers and Medicare recently approved coverage for these AI-driven reports, which add an extra $850 to $1,000 to the baseline cost of a standard CCTA.

Will AI Insights Actually Change Patient Care? The availability of this data raises an important debate. Does knowing the exact ratio of your soft and hard plaque genuinely change how a cardiologist treats you? More importantly, will altering a treatment plan based on AI data successfully lower your risk of a cardiac event?

"Right now, we really don’t know. But several studies to find out are currently under way," notes Dr. Michael Lu, director of artificial intelligence for the Cardiovascular Imaging Research Center at Harvard-affiliated Massachusetts General Hospital.

Launching the PREEMPT Study

To find definitive answers, a major nationwide effort is underway. In July 2026, 25 research institutions across the United States, including Massachusetts General Hospital, began enrolling participants for a new clinical trial funded by the National Institutes of Health (NIH).

The study focuses on discovering more effective preventative measures for cardiac events. According to Dr. Lu, who serves as the trial's co-principal investigator and is an associate professor at Harvard Medical School, the objective is to locate younger and middle-aged individuals showing early signs of atherosclerosis. Researchers will evaluate various medical interventions, including AI-enhanced plaque analysis, to see which methods work best to decrease coronary plaque. The comprehensive trial is projected to wrap up around 2032 (for more details, visit https://preemptstudy.org).

Evolution of Diagnostics: From Stress Tests to Advanced CT Scans

Cardiology has come a long way from traditional diagnostic protocols. To understand how much of a leap AI-assisted CCTA represents, it helps to look at how doctors traditionally evaluated heart health:

Stress Tests: Historically, the first line of defense was a stress test, which tracks blood flow, heart rate, and electrical activity while a patient exercises or takes medication. The flaw? These tests generally only flag abnormalities when an artery is severely obstructed (typically a blockage of more than 70%).

Traditional Angiography: If a stress test raised red flags, patients were sent for an invasive angiogram. This requires threading a catheter through the wrist or groin up to the heart to inject a contrast dye. While effective, it only provides a two-dimensional snapshot of blood flow.

Coronary CT Scans (CCTA): Compared to traditional angiography, a CCTA is rapid and non-invasive. Contrast dye is administered through a simple arm IV, and high-speed X-rays generate a highly detailed, three-dimensional view of the heart's blood vessels. Today, it is the preferred first-line diagnostic tool for patients experiencing chest pain or other concerning cardiac symptoms.

Coronary Artery Calcium (CAC) CT: Another common tool is the CAC scan. While useful, it features a major limitation: it only detects hard, calcified plaque, leaving soft plaque completely invisible. Because of this, CAC scans are strictly recommended to estimate future risk in asymptomatic patients who fall into borderline or moderate risk categories.

Weighing the Pros and Cons of AI Cardiology

Using CCTA to monitor how arterial blockages respond to therapies is not entirely new. Researchers have used these scans for years in clinical environments to track changes over one- to two-year periods.

However, the manual process is incredibly tedious; it typically takes a specialist over an hour to review a single scan and calculate plaque volumes.

In theory, artificial intelligence can complete this painstaking data aggregation in a fraction of the time. Yet, the technology isn't flawless. Dr. Lu emphasizes that no AI application is 100% accurate, meaning a qualified physician must still audit and verify every report.

While early observational data indicates that AI plaque tracking provides valuable insights that could optimize patient care, the medical community remains cautiously optimistic. Many cardiologists are holding out for concrete evidence proving that AI-guided treatment plans actively prevent heart attacks before entirely embracing the added financial cost.

Photo by www.kaboompics.com