Beschreibung
In this study, functional connectivity of the canine brain was investigated using functional magnetic resonance imaging (fMRI) in the German shepherd dog under isoflurane anesthesia. Due to the evolutionary process the same or similar functionally connected brain regions can be found from mice to dogs to humans. The resting-state networks are brain areas in which spontaneous low-frequency BOLD fluctuations can be measured without performing a specific task, reflecting the underlying neuronal activity.
For this study, 14 purebred German shepherd dogs without macromorphologic changes and using isoflurane anesthesia, were examined in a 3 Tesla magnetic resonance imaging scanner. This veterinary study is also the first to successfully use MP2RAGE as an anatomical MRI sequence when performing a resting-state examination. In addition, to minimize signal dropout in the frontal sinus region, a special sectioning of the epi-BOLD sequence was applied. Furthermore, the pre-processing was modified for the specific study. For this purpose, the software FSL (FMRIB Software Library), ANTs (Advanced Normalization Tool), SPM12 (Statistical Parametric Mapping), FreeSurfer and GIFT Toolbox (Group ICA of fMRI Toolbox) were combined and the results of the detected functional activity areas in their localization and extension in the German shepherd dog in the resting-state were described in detail.
The published results of the group-ICA (group-independent component analysis), using 20 components, indicate a total of 14 possible resting networks in the German shepherd dog. A total of 106 functionally linked activity areas were detected in the gray matter of the neocortex. For the first time, a connected default-mode network (component 8), a frontal (component 13) and an orbitofrontal network (component 4) as well as a dissociated right (component 17) and left auditory network (component 14) could be shown in the dog. In addition, a "second" dissociated default-mode network (component 0 and 11) as well as a visual network (component 19), two lateralized striatal networks (component 6 and 7), a cerebellar network (component 10), an olfactory network (component 15) and possibly a subcortical network (component 18) could be detected. In addition, there is evidence for the presence of a thalamocortical network (component 3, component 5, component 12) as well as a formatio reticularis network (component 1, component 9).
Another finding is that there is activation of the olfactory bulb, amygdala, and hippocampus in multiple components. This overlapping of functional activity areas in different components indicates important brain regions that are responsible for cognitive performance. In addition, the activity areas detected here raise the question of how our dogs perceive their environment and whether this may be similar to the phenomenon of synesthesia in humans. Thus, based on the data collected here, it is conceivable that dogs literally "see" odors (see components 15 and 19 with activation of visual cortex and bulbus olfactorius as well as lobus piriformis on both sides). However, at the present time this is only a hypothesis that needs to be further investigated in the future.
The functional data collected here in the resting-state will help to better understand the behavior of our dogs in the future and to detect individual character traits of dogs that are particularly suitable for specific tasks (e.g. assistance dogs). Furthermore, this work can contribute to diagnose neurodegenerative diseases such as canine cognitive dysfunction intra vitam using fMRI in the future and in the broadest sense, provide the basis for establishing new therapeutic options for comparable human medical diseases such as Alzheimer's disease, Parkinson's disease or attention deficit hyperactivity disorder (ADHD).
