Equine Stereotaxtic Population Average Brain Atlas With Neuroanatomic Correlation

Abstract

There is growing interest in the horse for behavioral, neuroanatomic and neuroscientific research due to its large and complex brain, cognitive abilities and long lifespan making it neurologically interesting and a potential large animal model for several neuropsychological diseases. Magnetic resonance imaging (MRI) is a powerful neuroscientific research tool that can be performed in vivo, with adapted equine facilities, or ex-vivo in the research setting. The brain atlas is a fundamental resource for neuroimaging research, and have been created for a multitude animal models, however, none currently exist for the equine brain. In this study, we document the creation of a high-resolution stereotaxic population average brain atlas of the equine. The atlas was generated from nine unfixed equine cadaver brains imaged within 4 h of euthanasia in a 3-tesla MRI. The atlas was generated using linear and non-linear registration methods and quality assessed using signal and contrast to noise calculations. Tissue segmentation maps (TSMs) for white matter (WM), gray matter (GM) and cerebrospinal fluid (CSF), were generated and manually segmented anatomic priors created for multiple subcortical brain structures. The resulting atlas was validated and correlated to gross anatomical specimens and is made freely available at as an online resource for researchers (https://doi.org/10.7298/cyrs-7b51.2). The mean volume metrics for the whole brain, GM and WM for the included subjects were documented and the effect of age and laterality assessed. Alterations in brain volume in relation to age were identified, though these variables were not found to be significantly correlated. All subjects had higher whole brain, GM and WM volumes on the right side, consistent with the well documented right forebrain dominance of horses. This atlas provides an important tool for automated processing in equine and translational neuroimaging research.

Diffusion Tensor Imaging Tractography of White Matter Tracts in the Equine Brain

“…Even if manual delineation of ROI is prone to generate biases, it was not possible to use automated segmentation as stereotaxic white matter atlases based on DTI are not available in the equine species. A freely available standard stereotaxic brain atlas was recently published for horses (16), but it does not contain DTI data. The creation of a tractography atlas could be a potential evolution of the present work.…”

“…The use of advanced MRI techniques in clinical practice offers the opportunity to better understand the mechanisms of neurological diseases of the horse and to improve their diagnosis. In the research setting, the implementation of such advanced MRI techniques often use postmortem samples, as exemplified by the two most recent MRI studies of the normal equine brain (15,16). Indeed, in addition to the constraints related to the adaptation of facilities for equine neuroimaging, the development of MRI tools in neurologically healthy horses poses the ethical problem of performing a general anesthesia for an MRI examination which is not medically warranted.…”

“…As DTI tractography assesses the structural integrity of white matter, it has been widely used for prognostic and/or diagnostic purposes in various brain pathologies such as stroke (3)(4)(5)(6), neurodegenerative diseases (7)(8)(9)(10)(11), and brain tumors (12). With the increasing availability of high-field-strength MRI (1.5 and 3 T) in veterinary facilities (13)(14)(15)(16)(17), the use of this technique is now gradually growing for the description of white matter anatomy and structural connectivity of domestic mammals (dog, cat, ferret, and sheep) (18)(19)(20)(21)(22)(23)(24)(25)(26)(27). These collected anatomical data are of particular interest as they might have applications in both veterinary medicine and experimental research.…”

In vivo Population Averaged Stereotaxic T2w MRI Brain Template for the Adult Yucatan Micropig

“…The binary tissue probability maps for GM, WM, and CSF were calculated using both the T2 and T1 templates ( Figure 5), and can be used as a mask for brain extraction, or as an a priori input for tissue segmentation of adult micropig brains in programs such as SPM12 or FSL. The partial volume maps for GM and WM were used to compute CSF, as well as gray matter and white matter tissue volumes in the T1 template, to increase accuracy (Table 2) (Johnson et al, 2019). Cortical surface architecture was estimated by concatenating GM and WM masks, and then creating a surface mesh 3D object from the volume.…”

“…The main limitations of our population averaged brain template are the low subject number and the inclusion of only female subjects. We included 16 animals in generating the T2 brain template based on the observation that templates stabilize at around 10 subjects (Avants et al, 2014) and the numbers used in other similar studies (Zhong et al, 2016;Villadsen et al, 2018;Johnson et al, 2019); however, we recognize that many human brain templates include one hundred or more subjects, and larger numbers would increase our confidence that our template is representative of the population. All the animals used in our study were female as we work exclusively with female micropigs.…”