Post-mortem magnetic resonance microscopy (MRM) of the murine brain at 7 Tesla results in a gain of resolution as compared to in vivo MRM

Abstract

Small-animal MRI with high field strength allows imaging of the living animal. However, spatial resolution in in vivo brain imaging is limited by the scanning time. Measurements of fixated mouse brains allow longer measurement time, but fixation procedures are time consuming, since the process of fixation may take several weeks. We here present a quick and simple post-mortem approach without fixation that allows high-resolution MRI even at 7 Tesla (T2-weighted MRI). This method was compared to in vivo scans with optimized spatial resolution for the investigation of anesthetized mice (T1-weighted MRI) as well as to ex situ scans of fixed brains (T1- and T2-weighted scans) by using standard MRI-sequences, along with anatomic descriptions of areas observable in the MRI, analysis of tissue shrinkage and post-processing procedures (intensity inhomogeneity correction, PCNN3D brain extract, SPMMouse segmentation, and volumetric measurement). Post-mortem imaging quality was sufficient to determine small brain substructures on the morphological level, provided fast possibilities for volumetric acquisition and for automatized processing without manual correction. Moreover, since no fixation was used, tissue shrinkage due to fixation does not occur as it is, e.g., the case by using ex vivo brains that have been kept in fixatives for several days. Thus, the introduced method is well suited for comparative investigations, since it allows determining small structural alterations in the murine brain at a reasonable high resolution even by MRI performed at 7 Tesla.

Keywords: MRI; animal scanner; contrast-to-noise ratio; in vivo; post-mortem; segmentation; signal-to-noise ratio; spatial resolution.

FIGURE 1

By using in vivo tissue, the brain and the surrounding tissue are clearly visible (A). However, the boundaries between the cortex and the CPu are difficult to detect (A,B). Neither the hippocampal fields CA1, CA3 nor the dentate gyrus can clearly be distinguished (C). Within the cerebellum (D) the different layers cannot be differentiated. Inserts B,C,D represent magnifications of areas shown in A. CPu, caudate putamen; OB, olfactory bulb.

FIGURE 2

By using unfixed post-mortem tissue, the brain and the blood vessels located on the brain surface are clearly visible in an intensity projection of the 3D scan (A). A representative slice through a 3D volume of un-fixed post-mortem head of a mouse is shown in (B). A higher magnification of the hippocampal formation is shown in (C). Likewise in the cerebellum (D) the white matter tracts and the gray matter can be distinguished. However, the Purkinje cell layer, which is composed of a single layer of relatively large neurons, cannot be detected (a higher magnification of this area is shown in E). CA1, CA3, hippocampal area; CPu, caudate putamen; DG, dentate gyrus; ec, external capsule; ent, entorhinal cortex; gl, granular layer of the cerebellum; Hipp, hippocampus; ic, internal capsule; icf, intercrural fibers; LV, lateral ventricle; ml, molecular layer of the cerebellum; OB, olfactory bulb.

FIGURE 3

Extracted brain tissue that was fixed for several days. The olfactory bulbs (OB), the cortex, the hippocampus (Hipp), and the cerebellum can easily be distinguished from adjacent brain regions. Due to the removal of the brain from the surrounding tissue, some damage can be noted, as e.g., in the cerebellum or the OB (indicated by arrows).

FIGURE 4

Analysis of the mean thickness of the cortex. Data are presented as mean ± min to max (*p ≤ 0.05; ANOVA; Tukey’s post hoc test). In vivo: data from the in vivo scanning procedure; post-mortem: data obtained by using post-mortem brains; fixed-hours: data obtained by using extracted brains that were fixed for some hours; fixed-days: data obtained from extracted brains that were fixed for several days.

FIGURE 5

(A) Center slice in the segmentation of the in vivo sequence for the female mouse; (B) same as in A, but for the post-mortem scan; (C) segmentation of the in vivo sequence for the male mouse; (D) the same as in C, but for the post-mortem scan; (E) SNR of raw unsegmented in vivo and post-mortem scans. The labels on the horizontal axis show which sequence and mouse has been analyzed and correspond to the upper row of the figure; (F) the same as in E, but for the CNR; (G) volume visualization of the in vivo segmentation result shown in A; (H) volume visualization of the post-mortem segmentation result shown in B.