WE-E-18A-12: Sampling Correction of Parametric Imaging Maps Derived From Digital Subtraction Angiography in Vascular Phantoms

Abstract

Purpose: To implement corrections for system sampling distortions in parametric imaging maps derived from digital subtraction angiography and test this approach in a controlled flow experiment using a neurovascular phantom. Methods: Depending on the exposure time and frame rate, imaging system sampling can cause distortion of the time density curves (TDC) used in parametric imaging. To solve this problem we implemented a distortion correction algorithm. A flow circuit set to physiologic conditions was created using an internal carotid artery phantom. Parametric imaging maps (PIM) were derived by measuring TDC's at each pixel location in the map. Each TDC was corrected by adding its numerical derivative multiplied by the sampling interval. PIM's were derived for two different contrast bolus injection volumes (5 and 10 ml), and four injection rates (5, 10, 15 and 20 ml/s). We derived PIM's for mean-transit-time (MTT), time-to-peak (TTP) and bolus-arrival-time (BAT). Colored maps were derived for each parameter; values were recorded at precise locations and compared between corrected and uncorrected cases for all injections techniques. Results: For corrected MTT's maps, the correlation with the injection techniques was still strong but the correction resulted in 10 to 30% smaller MTT's. At the inlet location, the corrected MTT's were between 2-0.25 seconds. For corrected BAT and TTP maps, the inlet values changed 20%. Conclusion: The experiment showed that the distorted PIM's for MTT are strongly correlated with the injection conditions and they agree with the expected theoretical values. Sampling correction reduces the BAT and TTP dependence on the injection technique which is expected since these particular parameters should be mostly related to the flow conditions rather than the injection technique. Supported by NIH Grant: 2R01EB002873 and equipment grant from Toshiba Medical Systems Corporation