Abstract 18160: Prospectively ECG Triggered Rapid Kilovolt (kV) Switching Dual Energy CT for Beam Hardening Reduction In Quantitative Myocardial CT Perfusion

Abstract

PURPOSE We investigated the effectiveness of beam hardening (BH) reduction in quantitative myocardial CT perfusion with a prospectively ECG triggered rapid kV switching dual energy (spectrum) CT (DECT) protocol.

METHODS On four normal farm pigs (55±20 kg), axial scans were triggered at every 1-2 mid-diastole(s) for 35 s after bolus injection of iodinated contrast (Isovue 370, 0.7 mgI[[Unable to Display Character: ⋅]]kg-1) at 4 ml[[Unable to Display Character: ⋅]]s-1 using a GE Healthcare Discovery CT 750HD scanner with a DECT protocol: 140/80 kilovolts (kV) alternated at 0.2 ms intervals, 640 mA and 0.35 s gantry period. Dynamic contrast enhanced (DCE) images of each pig were reconstructed at 140 kV as conventional single energy (spectrum) CT (SECT) acquisition and monochromatic 70 keV using the two kV projection sets. The 70 keV images were further corrected for BH using an image-based correction algorithm (iBHC). Each image set was analyzed using CT Perfusion (GE Healthcare) from which myocardial perfusion (MP) was calculated using a deconvolution analysis of the arterial and myocardial time-density curves (TDCs). TDCs and mean MP in the lateral, apical and septal wall over 8 × 5 mm slices of the LV were compared between the three image sets.

RESULTS Myocardial TDCs in the three walls measured from 140 kV DCE images showed larger differences between them. In contrast, those measured from 70 keV and 70 keV with iBHC DCE images exhibited more overlapping throughout the first-pass phase. The grand mean of MP measured from 140 kV (139.3±32.8 ml[[Unable to Display Character: ⋅]]min-1[[Unable to Display Character: ⋅]]100g-1) was significantly higher than those from 70 keV (108.1±22.7, p<0.05) and 70 keV with iBHC (86.9±22.5, p<0.05). Furthermore, 140 kV showed a large difference in mean MP (33.1%) between the septal and lateral wall (164.2±35.8 vs 123.4±16.0, p<0.05). Such difference was reduced to 15.8 % with 70 keV (115.5±27.0 vs 99.7±13.4) and further minimized to 6.1 % after iBHC was applied (84.3±21.6 vs 79.5±12.8).

CONCLUSION In SECT myocardial CT perfusion, BH induced large changes in CT number in different myocardial segments leading to inconsistent overestimation of MP. BH errors were significantly reduced when a 70 keV DECT protocol was employed, as evident by the more homogenous MP measurement within the normal myocardium. The residual BH error in the septum could be further reduced with iBHC.