Impact of right ventricular contrast attenuation on the accuracy of right ventricular function analysis at cardiac multi-detector-row CT

Abstract

Purpose: To investigate the impact of right ventricle (RV) contrast attenuation on the accuracy of RV function analysis at cardiac CT performed for coronary imaging.

Materials and methods: We analyzed multi-phase cardiac dual-source CT studies of 40 consecutive patients (mean age 60.9+/-12.3 years; 13 women) with normal valve function. Function analyses of both the right and left ventricle (RV and LV) were performed using dedicated post processing software and stroke volumes (SV) were computed for each ventricle. The accuracy of the RV-SV measurements was determined based upon comparison to LV-SV measurement. The level of contrast attenuation of the RV was recorded at three separate regions of interest-below the pulmonary valve, mid-ventricular, and inferior RV. The accuracy of RV function assessment was correlated with the level of attenuation using regression analyses.

Results: There was a statistically significant correlation between the accuracy of RV function assessment and the level of RV attenuation in the inferior RV (adjusted R(2)=73.1%, p<0.0001). When compared to LV-SV measurements, in studies with low RV attenuation (<175 HUs, n=19) the mean deviation of RV-SV from LV-SV was 29.0+/-10.8 ml (42.1+/-13.9%). In studies with high RV attenuation (> or =175 HUs, n=21) the mean deviation of RV-SV from LV-SV was significantly lower 6.9+/-10.5 ml (9.9+/-16.0%).

Conclusion: The accuracy of RV function analysis is dependent on the level of contrast medium attenuation achieved in the inferior RV. There should be at least intermediate attenuation (>175 HUs) to enable accurate functional analysis. At low attenuation levels RV function parameters are prone to underestimation.