Evaluating Reproducibility of the ADC and Distortion in Diffusion-weighted Imaging (DWI) with Reverse Encoding Distortion Correction (RDC)

Abstract

Purpose: To compare image distortion and reproducibility of quantitative values between reverse encoding distortion correction (RDC) diffusion-weighted imaging (DWI) and conventional DWI techniques in a phantom study and in healthy volunteers. Methods: This prospective study was conducted with the approval of our institutional review board. Written informed consent was obtained from each participant. RDC-DWIs were created from images obtained at 3T in three orthogonal directions in a phantom and in 10 participants (mean age, 70.9 years; age range, 63-83 years). Images without distortion correction (noDC-DWI) and those corrected with B0 (B0c-DWI) were also created. To evaluate distortion, coefficients of variation were calculated for each voxel and ROIs were placed at four levels of the brain. To evaluate the reproducibility of apparent diffusion coefficient (ADC) measurements, intra- and inter-scan variability (%CV[ADC]) were calculated from repeated scans of the phantom. Analysis was performed using Wilcoxon signed-rank test with Bonferroni correction, and P < 0.05 was considered statistically significant. Results: In the phantom, distortion was less in RDC-DWI than in B0c-DWI (P < 0.006), and was less in B0c-DWI than in noDC-DWI (P < 0.006). Intra-scan %CV[ADC] was within 1.30%, and inter-scan %CV[ADC] was within 2.99%. In the volunteers, distortion was less in RDC-DWI than in B0c-DWI in three of four locations (P < 0.006), and was less in B0c-DWI than in noDC-DWI (P < 0.006). At the middle cerebellar peduncle, distortion was less in RDC-DWI than in noDC-DWI (P < 0.006), and was less in noDC-DWI than in B0c-DWI (P < 0.0177). Conclusion: In both the phantom and in volunteers, distortion was the least in RDC-DWI than in B0c-DWI and noDC-DWI.