Researchers have found that an advanced AI model can use imaging findings to make fractional flow reserve (FFR) calculations and predict when coronary artery plaque blockages require surgery. Sharing its findings in the International Journal of Cardiology, the team noted that this algorithm could help patients avoid invasive catheterization procedures.
The study’s authors explored data from 113 patients with suspected coronary artery disease who underwent coronary CT angiography (CCTA) and invasive FFR measurements from August 2013 to May 2018. The average patient age was 59 years old, and 77% of patients were men.
Using vascuCAP, a commercially available AI software the team is developing with Elucid, they then examined each patient’s CCTA images to determine which blockages may require surgical intervention. The software can examine blood flow in multiple areas, even sliding down the vessel when necessary, without any of discomfort or potential complications associated with catheterization.
Those AI predictions were then compared to the FFR measurements the team already had on file, and the authors found that their efforts were a success. The AI-based FFR assessments achieved an area under the curve (0.94), sensitivity (0.90) and specificity (0.81) much higher than other techniques.
The team plans to continue validating its software, but they did speak to the solution’s potential for boosting patient care and limiting discomfort.
“There are high-risk features that can be seen on CT that help us to predict which patients are more prone to have a heart attack in the future, regardless of what the blood flow looks like at the time,” co-author U. Joseph Schoepf, MD, a researcher at the Medical University of South Carolina, said in a prepared statement. “So physicians can identify which patients need their help right now but also which patients need more TLC in their treatment to prevent heart attacks further down the road.”