Purpose: The purpose of this study was to evaluate the optimal end time of scanning and the influence of varying the number of source images adopted for calculation of the parameter values in computerized tomography (CT) perfusion.
Materials and methods: Nineteen CT perfusion studies in 14 patients with cerebrovascular disease were retrospectively analyzed. CT perfusion scanning was performed using continuous scans of 1 sec/rotation x60 sec with 5-mm-thick sections. To determine the appropriate end time of scanning, the time-density curves (TDCs) of the anterior cerebral artery (ACA), superior sagittal sinus (SSS), and basal ganglia were analyzed. The functional maps of cerebral blood flow (CBF), cerebral blood volume (CBV), and mean transit time (MTT) were retrospectively generated from various numbers (30, 35, 40, 45, 50, 55, and 60) of source images. Defining the values calculated from the number of source images equal to the end time of the first pass as standard values, the percentage difference between the standard values and the values generated from various numbers of sources were evaluated.
Results: The TDCs of SSS showed the latest end time of the first pass (mean, 38.6+/-5.2 sec; range, 32 to 48 sec). Therefore, the values calculated from the number of source images equal to the end time of the first pass of SSS were defined as standard values. Increase and decrease of the number of source images for calculation resulted in increase of the percentage difference in every parameter value. The percentage differences ranged up to 49.6% for CBF, 48.0% for CBV, and 20.0% for MTT.
Conclusions: Scanning until the end time of the first pass of SSS is necessary and sufficient for reliable measurement. Variable scan time based on the TDC of the SSS may be of better utility than use of fixed scan time. Further, the radiation dose could be minimized in many cases by reducing the scan time.