Comparison of MRI IVIM and MR Perfusion Imaging in Acute Ischemic Stroke Due to Large Vessel Occlusion
Guangming Zhu,Christian Federau,Max Wintermark,Hui Chen,David G. Marcellus,Blake W. Martin,Jeremy J. Heit
DOI: https://doi.org/10.1177/1747493019873515
2020-01-01
International Journal of Stroke
Abstract:Purpose Intravoxel incoherent motion is a diffusion-weighted imaging magnetic resonance imaging technique that measures microvascular perfusion from a multi-b value sequence. Intravoxel incoherent motion microvascular perfusion has not been directly compared to conventional dynamic susceptibility contrast perfusion-weighted imaging in the context of acute ischemic stroke. We determined the degree of correlation between perfusion-weighted imaging and intravoxel incoherent motion parameter maps in patients with acute ischemic stroke. Methods We performed a retrospective cohort study of acute ischemic stroke patients undergoing thrombectomy treatment triage by magnetic resonance imaging. Intravoxel incoherent motion perfusion fraction maps were derived using two-step voxel-by-voxel post-processing. Ischemic core, penumbra, non-ischemia, and contralateral hemisphere were delineated based upon diffusion-weighted imaging and perfusion-weighted imaging using a Tmax >6 s threshold. Signal intensity within different brain compartments were measured on intravoxel incoherent motion (IVIM f, IVIM D*, IVIM fD*) parametric maps and compared the differences using one-way ANOVA. Ischemic volumes were measured on perfusion-weighted imaging and intravoxel incoherent motion parametric maps. Bland–Altman analysis and voxel-based volumetric comparison were used to determine the agreements among ischemic volumes of perfusion-weighted imaging and intravoxel incoherent motion perfusion parameters. Inter-rater reliability on intravoxel incoherent motion maps was also assessed. Significance level was set at α < 0.05. Results Twenty patients (11 males, 55%; mean age 67.1 ± 13.8 years) were included. Vessel occlusions involved the internal carotid artery (6 patients, 30%) and M1 segment of the middle cerebral artery (14, 70%). Mean pre-treatment core infarct volume was 19.07 ± 23.56 ml. Mean pre-treatment ischemic volumes on perfusion-weighted imaging were 10.90 ± 13.33 ml (CBV), 24.83 ± 23.08 ml (CBF), 58.87 ± 37.85 ml (MTT), and 47.53 ± 26.78 ml (Tmax). Mean pre-treatment ischemic volumes on corresponding IVIM parameters were 23.20 ± 25.63 ml (IVIM f), 14.01 ± 16.81 ml (IVIM D*), and 27.41 ± 40.01 ml (IVIM fD*). IVIM f, D, and fD* demonstrated significant differences ( P < 0.001). The best agreement in term of ischemic volumes and voxel-based overlap was between IVIM fD* and CBF with mean volume difference of 0.5 ml and mean dice similarity coefficient (DSC) of 0.630 ± 0.136. Conclusion There are moderate differences in brain perfusion assessment between intravoxel incoherent motion and perfusion-weighted imaging parametric maps, and IVIM fD* and perfusion-weighted imaging CBF show excellent agreement. Intravoxel incoherent motion is promising for cerebral perfusion assessment in acute ischemic stroke patients.