Pre and Post-Treatment Temporal Parametric Analysis of Neurovascular Disease Using Gamma Variate Fitting of Time Density Curves From DSA Sequences
C Ionita1*, D Bednarek1, T Dumont1, A Siddiqui1, E Levy1, K Snyder1, S Rudin1, (1) SUNY Buffalo, Buffalo, NY,TU-A-116-6 Tuesday 8:00AM - 9:55AM Room: 116
Purpose: To calculate temporal parameters of contrast flow, using digital subtraction angiography (DSA), and provide parametric imaging (PI) of pre- and post-treated neurovascular lesions to aid neuro-interventionalists decision.
Methods: We developed a program to evaluate changes in contrast flow before and after treatment of neurovascular lesions. The program records each pixel value in a DSA sequence and fits the data with a gamma-variate function. Using the function we calculate parameters such as mean-transit-time (MTT), bolus-arrival-time (BAT) and time-to-peak (TTP) and generate PIs for each quantity. The program was used for pre and post-treatment parametric calculations for four representative neurovascular lesions: ischemic stroke, intracranial stenosis, arteriovenous malformation (AVM) and intracranial aneurysm treated with a flow diverter. Finally we compared the pre and post-treatment PIs for changes in temporal parameters of the contrast flow.
Results:Processing times for a typical DSA acquisition at 10 frames per second were between 15 and 20 minutes. In the ischemic stroke and intracranial stenosis cases we observed increased flow after clot removal and balloon angioplasty respectively, BAT decreased by 2x while MTT and TTP decreased 1.5x as compared with the pre-treatment PI. For the AVM the venous arrival time increased three times after glue embolization, indicating decreased flow through the AVM. For the aneurysm treatment the MTT in the aneurysm dome was 5x larger compared with the pretreatment run indicating a reduced flow in the aneurysm.
Conclusion: Pre and post-treatment PIs showed significant differences, indicating reestablished flow for ischemic stroke and intracranial stenosis and decreased flow for the AVM and intracranial aneurysm. Execution time can be decreased significantly if software and hardware is optimized. This kind of information could be a very useful tool for treatment assistance during intracranial minimally invasive procedures in the angiographic suites where CT-perfusion is not immediately available. Support NIH-2R01EB002873
Funding Support, Disclosures, and Conflict of Interest: Support from NIH Grant 2R01EB002873