T₂ relaxation time bias in gagCEST at 3T and 7T: comparison of saturation schemes

Purpose: To characterize the effects of water T₂ relaxation time in the glycosaminoglycan chemical exchange saturation transfer method (gagCEST) and compare them between 3T and 7T as well as between various saturation schemes. Methods: Simulations and a phantom experiment were conducted at 3T and 7T in a range of water T₂ values and GAG concentrations using various saturation schemes. For both simulations and MRI measurements, unsaturated signal as well as the saturated Z-spectrum were generated, and the magnetization transfer ratio asymmetry at 1 parts per million was used as a measure of the gagCEST effect size. Results: The simulations and phantom experiment results showed a clear GAG concentration and T₂ dependence of the gagCEST effect size. Whereas the gagCEST effect size was much larger at 7T, the impact of the T₂ bias was more pronounced at 3T. The saturation train length, duty cycle, and average B₁ all had clear impact on both the gagCEST effect size and T₂ bias. Conclusion: The water T₂ relaxation is important to consider in gagCEST, especially at 3T. The T₂ differences can introduce a pronounced bias, which may obscure the gagCEST effect when using low duty cycles and long saturation trains.