Correction of T1 Effects in Calculation of Relative Recirculation in Ischemic Stroke Patients

Abstract

Relative recirculation (rR) is a promising surrogate permeability biomarker derived from dynamic susceptibility contrast (DSC) magnetic resonance imaging (MRI), and can be applied to predict hemorrhagic transformation (HT). However, the influence of T1 effects on the calculation of rR has not been properly investigated. Furthermore, a correction method for T1 effects in rR calculation has not been reported. In this study, a two-compartment model is used to simulate the influence of T1 effects on rR under various imaging conditions. The results of the simulation demonstrate that the rR values calculated with T1 effects are smaller than, or even have the opposite sign, compared to those calculated without T1 effects. Data of 17 DSC–MRI scans were obtained from 14 ischemic stroke patients known to have blood–brain barrier (BBB) disruption on post-contrast T1-weighted images. A method is proposed to correct the T1 effects in the calculation of rR, and applied to analyze the ischemic stroke patient dataset. The statistics from an in vivo study show that the corrected rR in BBB disruption regions is significantly higher than the uncorrected rR (p < 0.05). Specifically, the corrected rR is positive, whereas the uncorrected rR is negative. rR values calculated without removing T1 effects can be severe underestimations. Using the proposed T1 correction method, the obtained rR can be adjusted to a level in accordance with known vascular physiology. As a semi-quantitative biomarker, rR with a high accuracy can potentially improve clinical assessment, such as the prediction of HT.