Quantitative susceptibility mapping for Alzheimer's disease, mild cognitive impairment, and normal aging: evaluation of corpus callosum.

Background: Abnormal iron metabolism and accumulation in the brain have been proposed as pathological changes in Alzheimer's disease (AD). These changes can be detected using quantitative susceptibility mapping (QSM). The corpus callosum (CC), an essential white matter structure in the brain, is thought to undergo volume and microstructural changes in Alzheimer's patients, with specific regional atrophy related to cognitive impairment and dementia severity. This study aimed to measure Methods: A retrospective study was conducted with 34 patients with AD, 32 patients with MCI, and 29 cases with HC. A subset of these participants had available follow-up magnetic resonance imaging (MRI) data, including 13 AD patients, 14 MCI patients, and 14 HC cases. Structural MRI data were processed using FreeSurfer software version 6.0 to segment the CC into five parts. QSM processing was performed using STISuite 3.0, and the results were registered and analyzed for susceptibilities in each CC segment using the FSL (FMRIB Software Library, version 5.0.7). Correlations between susceptibility levels and diagnosis were evaluated using the Kruskal-Wallis test, while associations between susceptibility and cognitive function [Thai Mental State Examination (TMSE) and Clinical Dementia Rating (CDR)] were assessed using Spearman's rank correlation coefficient. Changes after follow-up were assessed using paired samples Results: Significantly increased susceptibility was observed in the mid-anterior and central parts of the CC for AD patients compared to normal controls (0.051 and 0.103 ppm in AD and -0.014 and 0.003 ppm in HC; P value =0.014 and 0.009). Susceptibility in the mid-anterior, central regions, showed weakly positive correlations with CDR-global scores (r=0.296, P=0.006 and r=0.287, P=0.005). After a 2-year follow-up, susceptibility significantly increased across groups. In the HC group, significant increases were observed in the mid-anterior region (mean difference =0.074 ppm; P value =0.021). For the MCI group, a significant increase in the mid-posterior region (mean difference =0.081 ppm; P value =0.039) was found. For the AD group, a significant increase was found in the mid-posterior and posterior regions (mean difference =0.021 and 0.086 ppm; P value =0.013 and 0.005). Conclusions: The study findings suggest that increased susceptibilities in the mid-anterior and central parts of the CC can serve as a potential biomarker for the diagnosis of MCI and AD and assess cognitive function in these diseases.