Average saturation efficiency filter (ASEF) for CEST imaging

Abstract

Purpose: Endogenous CEST signal usually has low specificity due to contamination from the magnetization transfer effect and from fast exchanging labile protons with close Larmor frequencies. We propose to improve CEST signal specificity with an average saturation efficiency filter (ASEF).Methods: ASEF measures the difference between CEST signals acquired with similar average irradiation power but largely different duty cycles (DC), for example, a continuous wave or a high DC pulse train versus a low DC one. Simulation and Cr phantom studies were performed to evaluate the characteristics of ASEF for CEST. Results: Theoretical and simulation studies show that ASEF can suppress fast exchanging processes, with only a small loss of chemical exchange contrast for slow-to-intermediate exchange rates if the difference in DC is large. In the RF offset range of 2 to 5 ppm with an averaged saturation power of 0.8 and 1.6 microteslas, there is a mismatch of ∼0.1% to 2% in the magnetization transfer signal between saturation by continuous wave and a pulse train with DC = 15% and pulse duration of 24 ms, respectively. This mismatch can be minimized by careful selection of saturation power, pulse duration, and DC differences or by applying a small fudge factor between the 2 irradiation powers. Phantom studies of Cr confirmed that ASEF can minimize the magnetization transfer effect and reduce sensitivity to fast exchange processes. Conclusion:ASEF can improve the specificity of slow-to-intermediate exchanging CEST signal with a relatively small loss of sensitivity.

Average saturation efficiency filter ASEF‐CEST MRI of stroke rodents

“…This is different with our recent study in which the ASEFR MT signal is effectively minimized in agar phantoms. 41 Thus, we postulate that the proteins were not completely immobilized in the heated-BSA samples, and there may exist a very small amount of BSA with limited mobility and detectable CEST or aromatic nuclear Overhauser enhancement effect. This small amount of BSA may have contributed to the majority of the residue ASEFR signal at 3.4 ppm.…”

“…Saturation preparation schemes consisted of either a single CW block pulse or a train of binomial pairs of Gaussian pulses with a kurtosis of 4. 41 Whereas magnetization transfer ratio asymmetry (MTR asym ) is commonly used as a convenient measure of the CEST signal, it has recently been established that asymmetry analysis introduces contaminations from the MTC asymmetry of the semisolid macromolecules and the nuclear Overhauser enhancement effect, which could be larger than the CEST signal of interest. 13,19,30,36 Thus, only positive RF frequencies were acquired in this study.…”

“…Although the effectiveness of the exchange rate filtering effect has been confirmed in a pilot phantom study of ASEF signal, 41 a small mismatch of MT effect exists when saturation pulses with the same average B 1 power but different duty cycles are used 33,42 . This intrinsic disparity of the MT signal grows more apparent with a higher difference between the DC of the 2 saturation pulse trains and with a higher average B 1 power (B 1,avg ) used for saturation, as demonstrated in literature 33 and reported in our proof‐of‐principle study 41 . In addition to this intrinsic mismatch (Figure 1A), there is also the possibility of imperfections in MR hardware in RF power linearity and instability of delivering a sustained long RF irradiation; that is, the actual B 1,avg may differ slightly even though the normal B 1,avg is set to be equal for the 2 irradiation schemes, which can contribute to an imbalance of the background MT signals.…”

“…Although the effectiveness of the exchange rate filtering effect has been confirmed in a pilot phantom study of ASEF signal, 41 a small mismatch of MT effect exists when saturation pulses with the same average B 1 power but different duty cycles are used 33,42 . This intrinsic disparity of the MT signal grows more apparent with a higher difference between the DC of the 2 saturation pulse trains and with a higher average B 1 power (B 1,avg ) used for saturation, as demonstrated in literature 33 and reported in our proof‐of‐principle study 41 .…”

“…33,42 This intrinsic disparity of the MT signal grows more apparent with a higher difference between the DC of the 2 saturation pulse trains and with a higher average B 1 power (B 1,avg ) used for saturation, as demonstrated in literature 33 and reported in our proof-of-principle study. 41 In addition to this intrinsic mismatch (Figure 1A), there is also the possibility of imperfections in MR hardware in RF power linearity and instability of delivering a sustained long RF irradiation; that is, the actual B 1,avg may differ slightly even though the normal B 1,avg is set to be equal for the 2 irradiation schemes, which can contribute to an imbalance of the background MT signals. To improve the specificity of the ASEF CEST signal, the background ASEF MT signal caused by these confounding effects needs to be minimized.…”