A groundbreaking tool utilizing artificial intelligence is enhancing how doctors at Methodist Healthcare assess artery plaque during standard coronary CT angiograms. This innovative approach allows for a more nuanced analysis, providing critical information about heart health.
Soft, non-calcified plaque lurking in arteries poses a greater risk since it is more likely to rupture, potentially causing heart attacks or strokes. In contrast, hardened, calcified plaque is easier to identify and generally less hazardous.
Dr. Michael Lane, the Medical Director of Advanced Cardiovascular Imaging at Methodist Healthcare, played a key role in validating this new AI technology, which is now operational in 11 hospitals and four outpatient clinics within the Methodist network.
“In a year, this will likely become one of the most common tests in our emergency department,” Lane stated. “It’s invaluable to understand the extent of plaque in a patient’s arteries.”
Understanding the composition of plaque is essential for cardiologists, as it serves as a more accurate predictor of heart attack risk compared to mere artery narrowing. The newly developed AI tool has been trained on extensive heart disease datasets, enabling rapid differentiation between soft and hardened plaque.
Cardiovascular disease remains the leading cause of death globally and in the United States. Alarmingly, the heart disease mortality rate in Bexar County surpasses both national and state averages, according to 2019–2023 reports from the National Institutes of Health.
Heart Disease and Plaque
In his office, Lane examines a 360-degree view of a patient’s arteries after they presented with chest pain. The calcified plaques are easily identifiable, appearing as bright spots in greyscale images.
In contrast, soft plaque remains nearly invisible without advanced technology.
By using the AI tool with a simple click, the soft plaque is highlighted, illuminating various sections of the arteries in vibrant yellow. While Lane does not completely rely on these results, he finds the predictions strikingly accurate.
“This patient is not going home; they’re going to the cath lab for a stent placement to save heart tissue,” Lane remarked. “It’s crucial for their treatment.”
Over time, plaque accumulates within artery walls, composed of fats, cholesterol, and other substances, impeding the flow of oxygen-rich blood from the heart throughout the body.
This accumulation, known as atherosclerosis, often goes undetected for many years until it culminates in a plaque rupture, leading to a complete or partial blockage and resulting in heart attacks or ischemic strokes.
“When I speak with my patients, I urge them to understand that a heart attack at age 60 likely began at age 40,” Lane explained.
Lipid-rich soft plaque, or low-attenuation plaque, carries a greater risk of rupture compared to hardened, calcified plaque. “The soft plaque is what truly endangers lives,” Lane said.
Many common heart disease screening methods fail to account for plaque composition, often overlooking patients with high-risk soft plaque despite little or no arterial narrowing.
Typical screenings, like calcium score tests, focus on identifying hardened plaque to assess short-term cardiovascular risk and artery narrowing.
Several studies show that a significant number of patients with a calcium score of zero still harbor non-calcified plaque undetected by standard calcium scans.
Using advanced AI technology to gain clearer insight into plaque composition could enable physicians to identify patients who might require treatment that would have otherwise gone undetected.
How the Technology Works
This innovative technology is part of a growing suite of AI tools developed by companies like California’s HeartFlow. These tools analyze coronary CT scans to detect and quantify plaque in the arteries, with healthcare systems across the U.S. increasingly integrating AI to bolster various clinical and administrative functions.
This plaque analysis technology leverages AI to evaluate and compare extensive training data from coronary CT scans alongside intravascular ultrasounds (IVUS). This minimally invasive procedure uses a specialized catheter equipped with ultrasound technology to generate 3D images of plaque accumulation from within the arteries.
“Results showed that these AI models not only excelled at detecting blockages in arteries causing chest pain but also identified many other issues,” Lane noted. “We began to recognize soft plaque on the same CT scans where we were searching for blockages.”
Located in the South Texas Medical Center, Lane and his team conducted early testing on this technology. Currently, Methodist Healthcare operates 16 machines across 11 hospitals and four outpatient clinics, each capable of integrating the AI component into its scans.
Patients may request these advanced scans if their hospital is equipped with the technology, especially if they experience atypical chest pain or have a known history of coronary artery disease, according to Lane.
This new capability may prove pivotal in diagnosing cardiovascular disease among women, who often present with less total plaque than men but are more likely to possess undetected non-calcified plaque, which traditional imaging struggles to identify.
“This will be particularly beneficial for women aged 40 to 60, as we can visualize plaque that previously went unnoticed,” Lane concluded.
