Room: Track 1
Purpose: To develop and validate an optimal timing protocol for a two-volume dynamic CT pulmonary perfusion technique.
Methods: A total of 24 swine (48.5 ± 14.3 kg) underwent contrast-enhanced dynamic CT acquisition over 20-30 seconds to capture the pulmonary and aortic enhancement curves. Multiple contrast injections (volume: 0.5 or 1 ml/kg; rate: 5-10 ml/s; duration: 2–15 seconds) were made under different cardiac outputs (1.4 - 5.1 L/min) for each animal, resulting in a total of 154 pulmonary and aortic arterial input functions (AIF). Using the AIFs, an optimal timing protocol for prospective acquisition of two volume scans for perfusion measurement was developed. Specifically, the first volume scan was obtained at the base of the AIF simulating bolus tracking. The second volume scan was obtained at the peak of the AIF using a time-to-peak relation that was empirically derived using the contrast injection duration and regression analysis. After development of the optimal timing protocol, four swine with 29 total dynamic CT perfusion acquisitions were then used for validation. Specifically, simulated prospective two-volume perfusion measurements were quantitatively compared to a previously validated retrospective technique using t-tests, regression, and Bland-Altman analysis. The CT dose index (CTDI_vol³²) and size-specific dose estimate (SSDE) of two-volume perfusion measurement were also determined.
Results: Using the optimal timing protocol, the pulmonary artery time-to-peak (T_PA) was related to one-half the contrast injection duration (T_Inj/2) by T_PA=1.06T_Inj/2+0.90 (r=0.97). Simulated prospective two-volume perfusion measurements (P_PRO) were then related to the reference standard retrospective measurements (P_RETRO) by P_PRO= 0.93P_RETRO+0.89 (r=0.94). The CTDI_vol³² and SSDE of the two-volume dynamic CT perfusion measurement were estimated to be 28.4 and 47.0 mGy, respectively.
Conclusion: optimal timing protocol will enable an accurate and prospective two-volume dynamic CT perfusion technique.