AI used to develop tools to improve risk stratification of atherosclerosis
Click Here to Manage Email Alerts
Key takeaways:
- Researchers used AI to develop improved risk stratification tools for atherosclerotic CVD.
- The nomograms provide plaque volume estimates based on age, sex and clinical risk factors.
With the use of AI-enabled quantitative coronary plaque analysis, researchers developed nomograms designed to improve risk stratification of patients who may be at risk for atherosclerosis.
Their findings were presented at the Scientific Meeting of the Society of Cardiovascular Computed Tomography and simultaneously published in the Journal of Cardiovascular Computed Tomography.
“The risk factors of diabetes, hypertension, dyslipidemia and smoking are known drivers of coronary plaque leading to end outcomes. However, the exact impact of those risk factors on quantified total plaque volume on coronary CT angiography needs to be better understood,” Alexander Haenel, MD, clinical fellow in advanced cardiac imaging in the department of radiology at St. Paul’s Hospital & University of British Columbia in Vancouver, British Columbia, Canada, told Healio.
“With the current coronary CT angiography reporting standard of CAD-RADS 2.0, plaque burden in clinical practice is assessed qualitatively and, therefore, partially subjective. Plaque quantification further optimizes the use of diagnostic information from coronary CT angiography,” he said. “Traditionally, plaque quantification has required time-consuming and labor-intensive manual segmentation, limiting its clinical application. AI-enabled quantitative coronary plaque analysis allows time-efficient, accurate and reproducible implementation of plaque quantification into the clinical workflow.”
To study the impact of risk factors on plaque volume, researchers had 4,430 patients from the ADVANCE registry undergo AI-enabled quantitative coronary plaque analysis.
The researchers developed nomograms by plotting percentiles of total, calcified, noncalcified and low-attenuation plaque volume for each risk factor in different patient age groups.
Haenel and colleagues reported higher median total, calcified, noncalcified and low-attenuation plaque volume for patients with at least one risk factor: diabetes, hypertension, hyperlipidemia or smoking.
Total plaque volume observed with AI-enabled quantitative coronary plaque analysis for patients with at least one risk factor also increased with age.
The researchers reported that, excluding sex (OR for women = 0.513; 95% CI, 0.449-0.587; P < .0001), the strongest predictor of moderate total plaque volume — defined as volume more than 250 mm3 — was hypertension (OR = 1.579; 95% CI, 1.384-1.802), followed by diabetes (OR = 1.46; 95% CI, 1.241-1.717), smoking (OR = 1.406; 95% CI, 1.235-1.602) and hyperlipidemia (OR = 1.333; 95% CI, 1.169-1.519; P for all < .0001).
“This study is among the first to explore the relationship between traditional risk factors and quantified plaque volume and components,” Haenel told Healio. “By establishing nomograms for atherosclerotic plaque measures based on risk factors, age and sex, our research aims to support the use of AI-enabled quantitative coronary plaque analysis. This tool could enhance individual risk assessment, guide treatment decisions and improve therapy monitoring.”
For more information:
Alexander Haenel, MD, can be reached at 2775 Laurel St., 9th Floor, Vancouver, BC V5Z 1M9, Canada.
References:
- Haenel A, et al. J Cardiovasc Comput Tomogr. 2024;doi:10.1016/j.jcct.2024.05.182.
- SCCT 2024: HeartFlow to present new data on coronary artery disease management with coronary computed tomography angiography. https://www.heartflow.com/newsroom/scct-2024-heartflow/. Published July 16, 2024. Accessed July 30, 2024.
Collapse
Haenel A, et al. Poster session 16 – Plaque imaging. Presented at: Scientific Meeting of the Society of Cardiovascular Computed Tomography; July 18-21, 2024; Washington, D.C.
Read more about
Click Here to Manage Email Alerts