Evaluating a calcium-aware kernel for CT CAC scoring with varying surrounding materials and heart rates: a dynamic...

Area of Science:

• Medical Imaging
• Radiology
• Cardiovascular Imaging

Background:

• Coronary artery calcium (CAC) scoring is crucial for cardiovascular risk assessment.
• Standard CT reconstruction techniques can be affected by surrounding tissues and patient motion.
• Optimizing CAC scoring requires robust reconstruction methods.

Purpose of the Study:

• To evaluate a novel calcium-aware (Ca-aware) CT reconstruction technique's impact on CAC scores.
• To assess the Ca-aware technique's performance with various surrounding tissues (fat, water, soft tissue) and dynamic motion (simulating heart rates).
• To determine the feasibility of Ca-aware reconstruction for accurate CAC scoring under diverse clinical conditions.

Main Methods:

• An artificial coronary artery with varying CAC densities was placed in an anthropomorphic phantom filled with different tissue-equivalent materials.
• Dynamic CT data was acquired using standard (120 kVp) and reduced radiation protocols (100 kVp, 150Sn kVp).
• Raw data was reconstructed using both standard and the novel Ca-aware techniques, with Agatston scores compared to a reference.

Main Results:

• Ca-aware reconstruction yielded Agatston scores comparable to the reference for CAC in fatty tissues across all heart rates.
• Medium-density CAC in water and soft tissue also showed comparable scores with Ca-aware reconstruction.
• Low-density CAC demonstrated deviations up to 15% in water and 23% in soft tissue.

Conclusions:

• The novel Ca-aware CT reconstruction technique enables accurate CAC scoring across different surrounding materials and heart rates.
• This technique supports patient-specific tube voltage protocols, potentially reducing radiation dose.
• While effective for scoring, CT-derived mass scores may underestimate the actual physical mass of CAC.