Three-dimensional chemical exchange saturation transfer imaging using compressed SENSE for full z-spectrum acquisition

Purpose: To evaluate the accuracy of three-dimensional (3D) chemical exchange saturation transfer (CEST) imaging with a compressed sensing (CS) and sensitivity encoding (SENSE) technique (CS-SENSE) for full z-spectrum acquisition. Methods: All images were acquired on 3-T magnetic resonance imaging (MRI) scanner. In the phantom study, we used the acidoCEST imaging. The phantoms were prepared in seven vials containing different concentrations of iopamidol mixed in phosphate-buffered solution with different pH values. The CEST ratios were calculated from the two CEST effects. We compared the CEST ratios obtained with three different 3D CEST imaging protocols (CS-SENSE factor 5, 7, 9) with those obtained with the 2D CEST imaging as a reference standard. In the clinical study, 21 intracranial tumor patients (mean 49.7 ± 17.2 years, 7 males and 14 females) were scanned. We compared the intratumor magnetization transfer ratio asymmetry (MTRasym) obtained with 3D CEST imaging with those obtained with 2D CEST imaging as a reference standard. Results: A smaller CS-SENSE factor resulted in higher agreement and better correlations with the 2D CEST imaging in the phantom study (CS-SENSE 5; ICC = 0.977, R² = 0.8943, P < 0.0001: CS-SENSE 7; ICC = 0.970, R² = 0.9013, P < 0.0001: CS-SENSE 9; ICC = 0.934, R² = 0.8156 P < 0.0001). In the brain tumors, the means and percentile values of MTRasym at 2.0 and 3.5 ppm showed high linear correlations (R² = 0.7325–0.8328, P < 0.0001) and high ICCs (0.859–0.907), which enabled successful multi-slice CEST imaging. Conclusions: The 3D CEST imaging with CS-SENSE provided equivalent contrast to 2D CEST imaging; moreover, a z-spectrum with a wide scan range could be obtained.