Reproducibility of 3D pH-weighted chemical exchange saturation transfer contrast in the healthy cervical spinal cord

Abstract

Spinal cord ischemia and hypoxia can be caused by compression, injury, and vascular alterations. Measuring ischemia and hypoxia directly in the spinal cord noninvasively remains challenging. Ischemia and hypoxia alter tissue pH, providing a physiologic parameter that may be more directly related to tissue viability. Chemical exchange saturation transfer (CEST) is an MRI contrast mechanism that can be made sensitive to pH. More specifically, amine/amide concentration independent detection (AACID) is a recently developed endogenous CEST contrast that has demonstrated sensitivity to intracellular pH at 9.4 T. The goal of this study was to evaluate the reproducibility of AACID CEST measurements at different levels of the healthy cervical spinal cord at 3.0 T incorporating B₁ correction. Using a 3.0 T MRI scanner, two 3D CEST scans (saturation pulse train followed by a 3D snapshot gradient-echo readout) were performed on 12 healthy subjects approximately 10 days apart, with the CEST volume centered at the C4 level for all subjects. Scan-rescan reproducibility was evaluated by examining between and within-subject coefficients of variation (CVs) and absolute AACID value differences. The C4 level of the spinal cord demonstrated the lowest within-subject CVs (4.1%-4.3%), between-subject CVs (5.6%-6.3%), and absolute AACID percent difference (5.8-6.1%). The B₁ correction scheme significantly improved reproducibility (adjusted p-value = 0.002) compared with the noncorrected data, suggesting that implementing B₁ corrections in the spinal cord is beneficial. It was concluded that pH-weighted AACID measurements, incorporating B₁-inhomogeneity correction, were reproducible within subjects along the healthy cervical spinal cord and that optimal image quality was achieved at the center of the 3D CEST volume.

Keywords: CEST MRI; pH‐weighted contrast; postprocessing corrections; reproducibility.