Purpose: To create a group-specific vein-atlas based on healthy control subjects to visualize the average venous system under normal conditions and to compare the venous volume portion in multiple sclerosis (MS) lesions with that atlas.
Materials and methods: Susceptibility-weighted imaging (SWI), as well as T1-weighted imaging, was performed at 7 Tesla on nine healthy controls and nine age-matched MS patients. Automatic vein segmentation was performed on SWI data. The vessel segmentation results of the healthy controls were nonlinearly transformed into a model space, and subsequently averaged to create the vein-atlas. Thirteen normal-appearing white matter (NAWM) regions and 18 MS lesions were manually segmented in the patient data, and were used to calculate the venous volume portion in individual patient data and in corresponding regions within the vein-atlas.
Results: The vein-atlas illustrates the average venous network of the control group. The venous volume portion in MS lesions was significantly higher (P < 0.05) compared with the corresponding regions in the vein-atlas. NAWM regions did not differ significantly (P > 0.05) from corresponding atlas regions.
Conclusion: The developed vein-atlas shows the average venous system of a specific population and allows, therefore, the evaluation of the venous system of individual subjects.
Keywords: 7 Tesla; blood vessel; multiple sclerosis; susceptibility-weighted imaging; vein-atlas.
© 2013 Wiley Periodicals, Inc.