CSF Pulsation Artifacts on ADC Maps Obtained with Readout-segmented EPI

Abstract

Purpose: Diffusion-weighted imaging (DWI) using readout-segmented EPI (rs-EPI) can minimize distortion and blurring artifacts; however, we sometimes encounter cerebrospinal fluid (CSF) pulsation artifacts on apparent diffusion coefficient (ADC) maps, especially when the number of readout segments (NRS) is increased. The purpose of this study was to evaluate the effect of the NRS on the CSF pulsation artifacts in the ADC maps of healthy volunteers.

Methods: In 10 healthy volunteers, we obtained DWI from rs-EPI with a b-value of 0 and 1000 s/mm². The NRS was set to 5, 7, or 9. An ADC map was generated from the trace image and the b = 0 image. Each scan was performed twice. A circular region of interest (ROI) was drawn in the pons and bilateral thalami. The standard deviation (SD) of the ROI was measured to assess the artifacts. Bilateral SD values were averaged for the ROIs in the thalami. The SD values from two successive scans of each NRS were averaged for the pons and thalami, respectively. For the qualitative analysis, the CSF pulsation artifacts on each ADC map were graded by two observers independently as 0, no artifact; 1, mild artifact; 2, moderate artifact; or 3, severe artifact.

Results: In the quantitative analysis, the SD values tend to increase with the increasing of NRS in both thalami and pons; however, the difference in the SD values from each NRS did not reach a statistically significant level. In the qualitative analysis, there was a statistically significant difference in the scores between 5 and 9 segments and between 7 and 9 segments with both the observers, respectively (P < 0.05).

Conclusion: The CSF pulsation artifacts on ADC maps obtained with rs-EPI are affected by the NRS.