Effects of heart rate, filling and slice thickness on the accuracy of left ventricular volume measurements in a dynamic cardiac phantom using ECG-gated MDCT.

The objective of the study was to assess the effect of heart rate, filling condition and slice thickness on the accuracy of volumetric analysis based on multidetector-row computed tomography (MDCT) of a cardiac phantom. Retrospective electrocardiogram-gated MDCT of a pulsating phantom was performed under different conditions. End-diastolic volume (EDV) and end-systolic volume (ESV) for different heart rates (60-75 beats per minute), filling volumes and reconstructed slice thicknesses (2 mm and 5 mm) were obtained by three observers. Results were analysed by a linear mixed-effects model. Significant effects on the accuracy were found for heart rate (F-value, 7.3-39.2; p < 0.004) and filling condition (F-value, 7.4-55.6; p < 0.004), but not for slice thickness. Small relative differences in the assessment of EDV were found (range, -3% to 3%), but there was a trend for overestimation of the ESV (range, -1% to 18%). Underestimation of stroke volume and ejection fraction (range, -1% to -11%) became smaller under conditions of improved temporal resolution and larger EDV. Good interobserver agreement was found (SD < 1.8 ml and < 0.5%). In conclusion, MDCT allows sufficient and reliable measurements of ventricular volumes and calculation of left ventricle function for clinical applications. Heart rate and filling conditions significantly affect the accuracy of volumetrics, as demonstrated in this cardiac phantom. Thicker slices provide similar accuracy to thin slices.