Abstract P770: An Mri-Based Method for Assessing the Extent and Location of Bleeding in a Rat Model of Subarachnoid Hemorrhage

Abstract

Listen

Translate article

Objective: Studies in subarachnoid hemorrhage (SAH) animal models are limited to early timepoints, after which the brain is extracted to confirm hemorrhage. For long-term studies, blood is resorbed by the time of euthanasia and thus alternative methods of confirming SAH are necessary. Here, we used in vivo MRI to determine the presence, extent, and location of blood in the endovascular perforation rat model of SAH. Methods: Animals were divided into two groups - SAH and sham controls (no perforation). Rats were imaged at 24 hours in a 9.4T MRI scanner. T2-weighted imaging (T2WI) was used to quantify and localize the volume and extension of blood. T2WI was performed with a fast spin echo sequence with TR/TE = 5000/60 ms, FOV = 40x40 mm 2 , 0.75 mm slice thickness, and no interslice gap. SAH was identified as hypointense signal at the skull base, a normally hyperintense region due to cerebrospinal fluid in the basal cisterns. Regions of interest were identified and segmented manually. Pixels were converted to areas, and the total hypointense volume (V hypo ) was computed by summing all areas. V hypo was normalized to total brain volume (TBV) for each animal. Ability to detect blood via MRI was compared to the gold standard of direct visualization of blood after brain extraction using receiver operating characteristic (ROC) curves. The location of blood was plotted along a standardized antero-posterior axis. 3D reconstructions were developed in 3D Slicer to visualize the spatial distribution of hemorrhage. Results: V hypo values were 1.6±1.4 mm 3 sham (n = 8) and 7.5±3.9 mm 3 for SAH (n = 10) [p=0.0003]. This difference persisted when V hypo was normalized to TBV (p=0.0005). There was no difference in the TBV between groups (p=0.1493). V hypo distinguished between SAH and sham animals (p=0.001, area under curve = 0.963). A cutoff of V hypo = 4.63 mm 3 corresponded to 100% specificity and 80% sensitivity for confirming SAH. The spatial distribution of bleeding varied along the antero-posterior axis within the subarachnoid space. Conclusions: We developed an in vivo MRI method for the quantification of SAH blood volume and location in rats. This will be used for future long-term studies to confirm and grade severity of SAH.