On the Concordance Between Cardiac Magnetic Resonance and Computed Tomography for Left Heart Function Assessment

Area of Science:

• Cardiovascular Imaging
• Medical Physics
• Radiology

Background:

• Cardiac magnetic resonance (CMR) is the gold standard for assessing cardiac function.
• Cardiac computed tomography (CCT) offers potential advantages in speed and availability.
• Evaluating agreement between CCT and CMR is crucial for clinical adoption.

Purpose of the Study:

• To compare CCT and CMR measurements of left ventricular (LV) and left atrial (LA) function.
• To assess the impact of segmentation (papillary inclusion/exclusion) and volume calculation methods on CCT accuracy.
• To determine the influence of slice spacing on CCT-based measurements.

Main Methods:

• Retrospective analysis of 24 patients who underwent both CCT and CMR within 14 days.
• LV and LA volumes, ejection fraction, and LV mass measured by CCT using voxel-based volumetry with papillary inclusion/exclusion.
• Simulations of Simpson's Area-Length and Disk-Summation techniques on CCT images with varying slice spacings.

Main Results:

• CCT showed excellent correlation with CMR for LV volumes, ejection fraction, and mass (r=0.82-0.98, ICC=0.72-0.94).
• Papillary exclusion segmentation (PES) provided unbiased LV ejection fraction, while papillary inclusion (PIS) overestimated it.
• LA volume agreement was moderate (ICC=0.21-0.55), but improved with simulated slice spacing ≤8 mm.

Conclusions:

• CCT can accurately assess LV function compared to CMR when using voxel-based methods and consistent papillary segmentation.
• Papillary segmentation choice and slice spacing significantly affect measurement accuracy and agreement.
• Standardization of CCT acquisition and analysis is necessary for clinical interchangeability with CMR.