Coronary calcium scoring on contrast-enhanced spectral coronary computed tomography angiography using a...

Area of Science:

• Medical imaging
• Radiology
• Cardiovascular CT

Background:

• Coronary artery calcium (CAC) scoring typically requires a separate non-contrast scan, increasing radiation dose and scan time.
• Existing virtual non-contrast (VNC) methods often underestimate CAC scores and have limited detection of low-density calcifications.
• Improved VNC methods are needed to accurately quantify CAC from contrast-enhanced CT angiography (CCTA).

Purpose of the Study:

• To evaluate a novel self-developed CAC-preserving VNC (TrueCa) algorithm.
• To assess the TrueCa method's ability to improve the detection and quantification of coronary artery calcium (CAC) on CCTA.
• To compare TrueCa performance against conventional CAC CT scoring and standard VNC methods.

Main Methods:

• A phantom model with coronary artery calcifications of varying densities (75-800 mg/cc) was used.
• Scans were performed using dual-layer spectral CT with both CAC scoring and CCTA protocols.
• Agatston scores (AS) and volume scores (VS) were compared between CAC CT, standard VNC, and the TrueCa algorithm reconstructed at different slice thicknesses.

Main Results:

• Standard VNC CCTA significantly underestimated AS (30%-65%) and failed to detect CAC below 400 mg/cc.
• TrueCa CCTA successfully detected all calcifications, showing no significant difference compared to CAC CT across all densities.
• TrueCa demonstrated improved VS accuracy, especially for medium to low-density calcifications, outperforming standard VNC.

Conclusions:

• The TrueCa algorithm shows superior performance in detecting and quantifying coronary artery calcium across a wide range of densities.
• TrueCa CCTA holds significant potential as a reliable alternative to traditional CAC CT scans, enhancing diagnostic accuracy.
• This phantom study suggests TrueCa can improve CAC assessment on CCTA, potentially reducing radiation dose and scan time.