Computation of a high‐resolution MRI 3D stereotaxic atlas of the sheep brain

Abstract

The sheep model was first used in the fields of animal reproduction and veterinary sciences and then was utilized in fundamental and preclinical studies. For more than a decade, magnetic resonance (MR) studies performed on this model have been increasingly reported, especially in the field of neuroscience. To contribute to MR translational neuroscience research, a brain template and an atlas are necessary. We have recently generated the first complete T1-weighted (T1W) and T2W MR population average images (or templates) of in vivo sheep brains. In this study, we 1) defined a 3D stereotaxic coordinate system for previously established in vivo population average templates; 2) used deformation fields obtained during optimized nonlinear registrations to compute nonlinear tissues or prior probability maps (nlTPMs) of cerebrospinal fluid (CSF), gray matter (GM), and white matter (WM) tissues; 3) delineated 25 external and 28 internal sheep brain structures by segmenting both templates and nlTPMs; and 4) annotated and labeled these structures using an existing histological atlas. We built a quality high-resolution 3D atlas of average in vivo sheep brains linked to a reference stereotaxic space. The atlas and nlTPMs, associated with previously computed T1W and T2W in vivo sheep brain templates and nlTPMs, provide a complete set of imaging space that are able to be imported into other imaging software programs and could be used as standardized tools for neuroimaging studies or other neuroscience methods, such as image registration, image segmentation, identification of brain structures, implementation of recording devices, or neuronavigation. J. Comp. Neurol. 525:676-692, 2017. © 2016 Wiley Periodicals, Inc.

Computed tomography provides enhanced techniques for longitudinal monitoring of progressive intracranial volume loss associated with regional neurodegeneration in ovine neuronal ceroid lipofuscinoses

“…Overall, ventricular enlargement was more pronounced in CLN5 − / − animals than in CLN6 − / − animals, which may indicate a differing pattern of brain atrophy in the different types of ovine NCL. In recent past, a high‐resolution sheep brain MRI atlas has been established based on the brains of 18 sheep and it showed that the mean volume of the lateral and third ventricles combined was approximately 4.9 ml (Ella & Keller, 2015; Ella, Delgadillo, Chemineau, & Keller, 2017). The discrepancy between these measures and the measures recorded in the present study is likely consequent to differences in the measurement techniques and issues with the spatial resolution of the CT scanner used, and the present measures reflect the ventricular volumes sufficiently to compare affected and normal sheep within the same trial.…”

Ex Vivo MRI Analytical Methods and Brain Pathology in Preterm Lambs Treated with Postnatal Dexamethasone †

“…Manual segmentation of the cortical structures was based upon published MRI sheep brain atlases [24,25]. The hippocampus was defined based upon the Michigan State sheep brain atlas [26] and was determined by a detailed examination in each of the orthogonal imaging planes.…”

“…Briefly, we defined the frontal cortex by the areas defined as 1, 4, 13, 14, 15, 19, and 22 in Ella et al 2017 [24]. This region was chosen because the tissue boundaries were readily discernible and it was not significantly impacted by scanning artefacts due to undissolved gas (in contrast to the lateral ventricles).…”

“…This region was chosen because the tissue boundaries were readily discernible and it was not significantly impacted by scanning artefacts due to undissolved gas (in contrast to the lateral ventricles). Using the 3D rendered image of the masked brain volume, we placed fiducial points on the brain volume surface based upon landmarks in the Ella et al atlas [24] with the "Surface Cut" ( The hippocampus was segmented manually using "Segment Editor" in 3D Slicer, based upon closely examining slice planes and the 3D volume render. The segmentation was performed using a combination of the "Surface Cut" tool in the coronal plane, followed by editing with the "Paint" and "Smoothing" tool.…”

Sulci of the canine brain: a review of terminology

“…The diagnostic and research interpretation of such techniques requires the knowledge and unambiguous identification of areas and regions; one of the most complex CNS structures being the brain. Brain atlases are therefore fundamental tools for several clinical and research purposes [1–5]. One essential aspect of such research activities is the need to compare, and thus identify, brain structures and areas across different modalities (from macroscopic modalities to histological studies).…”