pH‐sensitive amide proton transfer effect dominates the magnetization transfer asymmetry contrast during acute ischemia—quantification of multipool contribution to in vivo CEST MRI
Because magnetization transfer (MT) asymmetry (MTRasym) analysis provides an efficient means to correct the direct water saturation, MTRasym has been used commonly in quantifying the in vivo APT effect. However, it has been recognized that MTRasym is susceptible to concomitant radiofrequency (RF) saturation effects, particularly the slightly asymmetric semisolid MT and nuclear Overhauser effect (NOE), which may somewhat limit its specificity to changes in pH and amide proton concentration 8. To address this, alternative quantitative measures have been investigated. For example, it has been shown that the regression of MTRasym against MT and longitudinal relaxation rate (R1) helps mitigate non‐pH heterogeneity within the brain, therefore improving pH specificity for automated ischemic lesion segmentation 5. In addition, the APT effect can be derived from the difference between a Z‐spectrum and Lorentzian line fitting that avoids the need of using upfield reference scans 18. In addition, the spillover and MT can be estimated with a three‐way subtraction, allowing less contaminated APT quantification 20. It is also possible to take the signal difference between the peak of interest and the mean of the two neighboring offsets to determine APT and NOE effects independently 10. However, this three‐point approach assumes a simplistic linear baseline, which is limited to high‐field applications 22. Moreover, an inverse Z‐spectrum analysis was introduced to help quantitate direct‐saturation and MT effects, revealing good contrast between the ischemic lesion and normal tissue 23.
Our study aimed to determine the origins of in vivo MTRasym contrast during acute ischemic stroke, particularly in the diffusion lesion, perfusion lesion, and their mismatch using a well‐established middle cerebral artery occlusion rat model of acute stroke. Five potential sources to MTRasym contrast, including NOE, MT, direct water saturation, amine and amide CEST effects, were numerically solved from Z‐spectrum using a multipool Lorentzian function 10. Our study showed that, although MTRasym analysis has concomitant contributions from multiple origins, its change at 3.5 ppm is dominated by pH‐sensitive APT contrast. This finding is helpful to elucidate the pH specificity of the commonly used MTRasym, support simplified and expedient pH‐sensitive image postprocessing, and ultimately, facilitate translating pH MRI to the acute stroke setting.