Using artificial intelligence (AI) tools to assess CT scans can make a significant impact on patient care, according to a new analysis published in the Journal of the Society for Cardiovascular Angiography and Interventions.[1] In fact, the study’s authors found that they influenced cardiologists to modify their own evaluation of the patient in a majority of cases.
Plaque Analysis, the AI offering at the heart of this analysis, was developed by HeartFlow to evaluate signs of coronary plaque identified in coronary CT angiography (CCTA) scans. It received U.S. Food and Drug Administration approval back in October 2022.
“To date, plaques have been quantified manually by CCTA readers which is time-consuming and less reproducible than using an AI-based process,” wrote first author Sarah Rinehart, MD, medical director of cardiovascular imaging, CT and nuclear at Charleston Area Medical Center, and colleagues.
For the DECODE study, Rinehart et al. asked a team of trained cardiologists to interpret CCTA images and review other key details for 100 patients. The cardiologists then worked together to produce an initial management plan for each patient.
At this point, the group was given access to the AI-powered assessment provided by HeartFlow’s Plaque Analysis offering. They then looked back at that initial management plan to see if they wanted to make any modifications now that they had additional details.
The patient population had a median age of 64 years old, and 59% were men. While 84% patients presented with dyslipidemia, another 71% presented with hypertension and 23% presented with diabetes. A family history of premature coronary artery disease (CAD) was seen in 48% of patients, and 42% were already on statin therapy. The median coronary artery calcium (CAC) score was 99.5.
Overall, patient management plans were modified for 66% of patients. When the patient’s CAC score was 0, management plans were changed 47.4% of the time. When the CAC score was 400 or more, on the other hand, management plans were changed 96% of the time.
“The majority of management changes consisted of up-classification and intensification of medical therapy,” the authors added. “We found a high rate of medical therapy reclassification across the spectrum of CCTA-defined CAD.”
Click here to read more in the JSCAI, a journal from the Society for Cardiovascular Angiography and Interventions.
Michael has more than 16 years of experience as a professional writer and editor. He has written at length about cardiology, radiology, artificial intelligence and other key healthcare topics.