Lesional and perilesional tissue characterization by automated image processing in a novel gyrencephalic animal model of peracute intracerebral hemorrhage

Abstract

Intracerebral hemorrhage (ICH) is an important stroke subtype, but preclinical research is limited by a lack of translational animal models. Large animal models are useful to comparatively investigate key pathophysiological parameters in human ICH. To (i) establish an acute model of moderate ICH in adult sheep and (ii) an advanced neuroimage processing pipeline for automatic brain tissue and hemorrhagic lesion determination; 14 adult sheep were assigned for stereotactically induced ICH into cerebral white matter under physiological monitoring. Six hours after ICH neuroimaging using 1.5T MRI including structural as well as perfusion and diffusion, weighted imaging was performed before scarification and subsequent neuropathological investigation including immunohistological staining. Controlled, stereotactic application of autologous blood caused a space-occupying intracerebral hematoma of moderate severity, predominantly affecting white matter at 5 h post-injection. Neuroimage post-processing including lesion probability maps enabled automatic quantification of structural alterations including perilesional diffusion and perfusion restrictions. Neuropathological and immunohistological investigation confirmed perilesional vacuolation, axonal damage, and perivascular blood as seen after human ICH. The model and imaging platform reflects key aspects of human ICH and enables future translational research on hematoma expansion/evacuation, white matter changes, hematoma evacuation, and other aspects.

Keywords: Brain; hemorrhage; large animal; perfusion; segmentation.

Figure 1.

Exemplary structural MRI of intracerebral hemorrhage (ICH) model in sheep and key features for the structural image processing. (a) Exemplary coronal and transversal sections T1w, T2w TSE, and T2w* MRI sequences from ICH in a sheep brain registered to the ovine standard space. Crosshairs represent the coordinate [x;y;z] of the ovine space at [−19;0;8]. The 3D-rendered surface provides the spatial localization of the lesion and depicts gray (GM) and white matter (WM), as well as the ventricle. (b) A T1w image attenuated for T2-signals (LogS₀S₁) served as input for the unified segmentation procedure of SPM8 to adequately reflect the heterogeneous lesion. The tissue probability based on the ovine tissue probability maps for GM and WM were modified by the lesion probability, which served as a separate tissue class.

Figure 2.

Volumetry and morphology of gray (GM) and white (WM) matter as well as ventricles with and without intracerebral hemorrhage (ICH). (a) GM and WM volume of both hemispheres at baseline and ICH: The gray, dotted line represents the median of the unaltered tissue volumes. (b) shows the tissue ablation (I^a) and lateralization index (I^lat) for GM, WM, and ventricle of both hemispheres at baseline and after ICH. (c) Ventricle-to-brain index (I^v:b) of contra- and ipsilateral hemisphere are given at baseline and after ICH. (d) The voxel-based-morphometry revealed a compression of the ipsilateral lateral ventricle (green, left), while the olfactory ventricles were dilated bilaterally after ICH (green, right). ICH location is indicated as red ellipsis. boxplot values: median, 25–75% percentile, 95%CI, + outlier, white – without ICH, black – with ICH; V_with – ventricle within hemorrhagic altered brain; V_without – ventricle without ICH; R – right hemisphere; L – left hemisphere; T – t-value statistics; */**/*** indicate p-value below 0.05/0.01/0.001.

Figure 3.

Parametric perfusion (mean-transit-time, MTT) and diffusion (apparent diffusion coefficient, ADC) maps and the lesion/deficit volumes. (a) T2w TSE image (left), related map of apparent diffusion-coefficient (ADC, middle), and perfusion-weighted-imaging via map of mean-transit-time (PWI, right, including T2 TSE overlay) were aligned to ovine stereotaxic space. Crosshairs located at [−17, 0, 8]. (b) Lesion and deficit mismatches: Mismatch analysis of the ratio of L^DWI:L^PWI and L^logS0S1:L^PWI deficit volumes revealed a significant perfusion excess (p < 0.01 each). The ratio between L^ADC and L^DIFF was also significantly reduced (p < 0.05). There was a significant correlation between L^PWI and the diffusion-perfusion mismatch L^DWI:L^PWI (p < 0.05). Interestingly, this was also the case for the correlation between L^PWI and L^DIFF:L^PWI (p < 0.05), but not for L^PWI and L^DIFF:L^PWI (p = 0.346). (c) Overlay of individual structural lesion (L^logS0S1,blue), PWI (L^PWI, green), diffusion deficit (L^DWI as sum of apparent (L^ADC), and free diffusion (L^DIFF), red) superimposed with structural brain mask (pale blue). Boxplot values: median, 25–75% percentile, 95%CI, + outlier, R = correlation coefficient; */*/* indicate p-value below 0.05/0.01/0.001

Figure 4.

Histopathology of brain tissue around experimental ICH in sheep brain. (A) Low magnification view of hematoma (Hta) within subcortical white matter: Parenchymal tissue close to the hematoma exhibited vacuolation (V) suggesting ischemic damage. (B) Higher magnification view of region shown in (A) shows the margin of vacuolated white matter (V). (C) Blood (asterisk, *) occurred in the perivascular space around a small arterial vessel, close to the hematoma. (D–G) Sections showing tissue adjacent to the hematoma immunolabelled for microglial cells (Iba-1, panel D) or astrocytes (GFAP, E) myelin (MBP, panel F) or axonal neurofilaments (panel G). (H, I) Higher magnification of neurofilament labelled tissue shows swelling and bulbing of axons (arrows). (I) Neurofilament labelled healthy white matter from the same animal, for comparison. (J) Shrunken, eosinophilic pyramidal neurons (arrows) were seen in overlying cortical grey matter (arrows). Many non-shrunken neurons were also evident (arrowheads). (K, L) sections labelled for the plasma marker fibrinogen. Generalised and cellular fibrinogen was detected close to the hematoma (K). Fibrinogen positive cells were also seen relatively distant from the hematoma (arrows in panel L). Fibrinogen and blood cells (asterisk) were evident in the perivascular space around a small blood vessel. (M, N) ICH altered sheep brain specimen. Coronal gross sections (upper = anterior, bottom = posterior) reveal sharply demarcated intracerebral hematoma (white arrow) in radiate corona. Panels A, B, C, J: standard haematoxylin-eosin stain. D–I, K: immuno-histochemical labelling with DAB chromogen (brown) and hematoxylin nuclear counterstain (blue). Scale bars (A–l) 500 µm (a), 20 µm (H, I), 100 µm (b–G, K–L) and 20 mm (M,N).