Interplatform reproducibility of CT coronary calcium scoring software

Area of Science:

• Cardiovascular Imaging
• Medical Physics
• Radiology

Background:

• Coronary artery calcium (CAC) scoring is crucial for cardiovascular risk assessment.
• Standardization of CAC scoring across different imaging workstations is essential for clinical reliability.

Purpose of the Study:

• To evaluate the vendor-independence of CAC scoring by comparing results from three different software workstations.
• To determine if CAC scoring software generates comparable results across platforms.

Main Methods:

• Fifty-nine patients underwent 64-section multidetector computed tomography (CT) for CAC scoring.
• Agatston and volume scores were calculated using software from three distinct workstations.
• Statistical analyses included regression, Spearman rank correlation, and Kruskal-Wallis tests.

Main Results:

• Excellent correlation (rs > 0.998) was observed between workstations, particularly at the 60% R-R interval.
• No significant differences in Agatston or volume scores were detected between software platforms.
• Score variability across reconstruction intervals led to divergent cardiac risk group assignments for 12 patients.

Conclusions:

• Commercially available CAC scoring software platforms yield comparable results, indicating vendor-independence.
• Variability in CAC scores based on reconstruction interval persists across platforms, limiting distinct advantages for risk stratification.