[Improved magnetic resonance myelography using image fusion].

PURPOSE: To demonstrate that the disadvantage of missing anatomical information in heavily T2-weighted MR myelography images can be eliminated by image fusion and phase encoding in the coronal direction of the source images, resulting in MR myelography images comparable to the gold standard, i. e., post-myelography CT.

MATERIALS AND

METHODS

This study included 110 patients suffering from extradural pathologies of the cervical and lumbar spine. All patients were investigated using 3D MR myelography and post-myelography CT. The MRI data were post-processed using image fusion and reconstruction algorithms and were compared to the corresponding images of post-myelography CT.

RESULTS

Our approach for visualization (3D MR myelography) was able to depict intradural structures in high spatial resolution and without artifacts. The results of our visualization approach were comparable to the gold standard - post-myelography CT. Anatomical correlation was reached by image fusion of different MR data sets. The required post-processing steps were performed quickly and were available on a commercial workstation.

CONCLUSION

Image fusion of different MR data sets allows for visualization of 3D data sets with enhanced quality. The results for the visualization of MR myelography in particular are comparable to conventional myelography and post-myelography CT. The missing anatomical information in heavily T2-weighted MR myelography images can be compensated by image fusion with conventional MRI.