Högskolan i Skövde

Alzheimer’s disease is characterized by early neuroinflammatory changes, particularly in the hippocampus, where myo-Inositol (mI), a glial metabolite detectable via magnetic resonance spectroscopy (MRS), may serve as a promising biomarker. This study investigated the detectability and test-retest reliability of hippocampal myo-Inositol quantification using single-voxel MRS at 3T in healthy adults. Eleven participants underwent four scans across two days, with acquisition parameters varied to assess session- and day-specific effects. Data was processed using Osprey, and reliability was evaluated using intra-class effect decomposition (ICED). Myo-Inositol was successfully detected in all participants (SNR = 62.01 ±13.22; FWHM = 7.94 ±1.23), and coefficients of variance (CoVs) were below 10% in most cases. However, the ICED model indicated a high proportion of residual error variance (53.8%, p = .001) alongside modest but significant trait variance (40.7%, p = .043). Session- and day-specific sources of variance were non-significant. These results suggest that while hippocampal myo-Inositol can be reliably detected, its current test-retest reliability limits clinical applicability. Technical challenges, including motion-related artifacts and interindividual differences in hippocampal size relative to the fixed voxel, may have contributed to measurement noise. Future studies should explore improved acquisition strategies to enhance stability and biomarker viability.