Fig 1.
Comparison of a canine brain with normal lateral cerebral ventricles (A) and enlarged ventricles (B).
Fig 2.
Volume determination based on MRI-datasets.
Image segmentation of white matter and grey matter using manual segmentation on a slice-by-slice basis from transverse images. Each tissue of interest is labelled red and thereby assigned to a group (mask). All masks are then assembled and the tissues can be depicted in volume form.
Fig 3.
Volume rendering of brain tissues of interest.
3D viewer mode of the graphical software AMIRA. The voxels of the tissue of interest (white matter/grey matter) of each slice have been assembled and are now displayed as a 3D model. Each tissue can be displayed solid or transparent. The localizer lines support the segmentation process. As they are displayed in both the 2D images and the 3D model, the thalamus, medulla and cerebellum can be accurately separated from the volume of interest.
Fig 4.
Volume models of grey matter and white matter.
Frontal (A) and dorsal view of a 3D model of the brain of a Jack Russell Terrier. The white matter is labelled in red and the grey matter is transparent grey.
Table 1.
Breed, body weight and diagnosis of the dogs included in groups one and two.
Fig 5.
Linear regression model analysis of the changes in relative grey matter (GM), white matter (WM)- and CSF volume and the WM/GMratio with increasing bodyweight in dogs.
The relative grey matter (GM), white matter (WM), and CSF volume of the lateral venmtricles as well as the WM/GM-ratio is plotted against the bodyweight in two groups of dogs. The open circles represent dogs with subjectively normal ventricles, the filled circles represent dogs with enlarged ventricles. The adjusted means in the middle of the regression lines of the groups differs significantly.
Table 2.
Results of the one-way analysis of covariance.