Comparison of Echo Planar and Turbo Spin Echo Diffusion-Weighted Imaging in Intraoperative MRI.

Abstract

Background: Diffusion-weighted imaging (DWI) is routinely used in brain tumor surgery guided by intraoperative MRI (IoMRI). However, conventional echo planar imaging DWI (EPI-DWI) is susceptible to distortion and artifacts that affect image quality. Turbo spin echo DWI (TSE-DWI) is an alternative technique with minimal spatial distortions that has the potential to be the radiologically preferred sequence.

Purpose: To compare via single- and multisequence assessment EPI-DWI and TSE-DWI in the IoMRI setting to determine whether there is a radiological preference for either sequence.

Study type: Retrospective.

Population: Thirty-four patients (22 female) aged 2-61 years (24 under 18 years) undergoing IoMRI during surgical resection of intracranial tumors.

Field strength/sequence: 3-T, EPI-DWI, and TSE-DWI.

Assessment: Patients were scanned with EPI- and TSE-DWI as part of the standard IoMRI scanning protocol. A single-sequence assessment of spatial distortion and image artifact was performed by three neuroradiologists blinded to the sequence type. Images were scored regarding distortion and artifacts, around and remote to the resection cavity. A multisequence radiological assessment was performed by three neuroradiologists in full radiological context including all other IoMRI sequences from each case. The DWI images were directly compared with scorings of the radiologists on which they preferred with respect to anatomy, abnormality, artifact, and overall preference.

Statistical tests: Wilcoxon signed-rank tests for single-sequence assessment, weighted kappa for single and multisequence assessment. A P-value <0.001 was considered statistically significant.

Results: For the blinded single-sequence assessment, the TSE-DWI sequence was scored equal to or superior to the EPI-DWI sequence for distortion and artifacts, around and remote to the resection cavity for every case. In the multisequence assessment, all radiologists independently expressed a preference for TSE-DWI over EPI-DWI sequences on viewing brain anatomy, abnormalities, and artifacts.

Data conclusion: The TSE-DWI sequences may be favored over EPI-DWI for IoMRI in patients with intracranial tumors.

Level of evidence: 2 TECHNICAL EFFICACY: Stage 5.