gms | German Medical Science

Objective: High-field iMRI suites provide surgeons with diagnostic-quality MR scans. However, more advanced imaging, such as DWI, is highly sensitive to susceptibility artefacts. We analyzed and defined the sources of interference, compared intraoperative and postoperative MR diffusion restrictions and developed an optimized DWI sequence to reduce susceptibility artefacts.

Methods: iMRI scans were analyzed for the quality of standard DWI sequences (ap-direction of the patient’s head) and apparent diffusion coefficient (ADC) maps. Artefact-causing factors were identified by analysis of standard T1 images. Intraoperatively identified diffusion-restrictions were compared to the postoperative MRI (within 48h). Additional b0-images with opposed phase encoding direction (pa-direction of the patient’s head) were used for offline distortion reduction.

Results: 35 patients with complete imaging data sets were included in this study. DWI raw data were not usable due to severe artefacts in 10 (AP) and 7 (PA) patients. ADC maps were analyzed in all patients. T1 analysis revealed that air, caused by an insufficiently fluid-filled resection cavity, was the main cause for severe artefacts. Air collections in the anterior-posterior direction to the resection cavity (e.g. frontal lesions in supine position) influenced the image quality much more, than air collections lateral to the resection cavity (e.g. temporal lesions in lateral position). Also, positioning of the patients head close to the limits of the MR-isocenter and blood in the resection cavity severely worsened the quality of DWI signals. Best results were seen for patients with entirely fluid-filled resection cavities, however unnecessary artefacts were caused by large compresses laying on the situs while scanning in few cases. Postoperatively identified diffusion restrictions (small lesions with direct proximity to the resection cavity or larger infarctions) were seen in 26 patients. 12 (46%) of those were already identified intraoperatively, 14 (54%) were not.In 9 of these 14 cases (64%) a diffusion restriction was seen after distortion correction, which however appeared larger in the postoperative MRI in 7 of these cases.

Conclusion: DWI is an important and reliable tool to identify diffusion restrictions in iMRI. If artefact causing conditions are known and prevented by surgeons, quality and reliability can be increased easily. Additional post-processing algorithms to reduce distortions can aid in identifying diffusion restrictions.