Multimodality Neuroimaging in Mild Cognitive Impairment: A Cross-sectional Comparison Study

Area of Science:

• Neuroimaging
• Neurology
• Biomarkers

Background:

• Mild cognitive impairment (MCI) is a critical research area due to its association with Alzheimer's disease (AD).
• Early and objective phenotyping of MCI is essential for timely intervention and clinical management.
• Current diagnostic methods often rely on subjective cognitive assessments and structural MRI alone.

Purpose of the Study:

• To evaluate a novel logistic regression-based classifier for early, objective phenotyping of MCI.
• To compare the diagnostic utility of a multimodal MRI approach against structural MRI alone.
• To assess the classifier's ability to differentiate MCI from healthy controls and from early AD converters.

Main Methods:

• Participants included 33 with stable amnestic MCI, 15 MCI converters to AD, and 20 healthy controls.
• Multimodal MRI data acquired: volumetric T1-weighted MRI, diffusion tensor imaging (DTI), and proton magnetic resonance spectroscopy (¹H MRS).
• A multimodal classifier integrated regional volumes, DTI metrics (fractional anisotropy, mean diffusivity), and ¹H MRS metabolite ratios (NAA/Cr, Cho/Cr, mI/Cr, NAA/mI).

Main Results:

• Diffusion tensor imaging (DTI) showed the highest sensitivity (90.9%) for MCI vs. controls, but lower specificity (50%).
• For MCI vs. AD classification, DTI was the best individual modality (72.7% sensitivity, 87.9% specificity).
• The multimodal classifier achieved high performance: AUC=0.89 (93.9% sensitivity, 70% specificity) for MCI vs. controls, and AUC=0.93 (93% sensitivity, 85.6% specificity) for MCI vs. AD.

Conclusions:

• Combining gray matter atrophy, white matter tract integrity (DTI), and metabolite variations (MRS) significantly improves MCI classification accuracy.
• The multimodal MRI classifier outperforms individual MRI biomarkers for differentiating MCI from controls and predicting conversion to AD.
• This integrated approach offers a promising tool for early, objective phenotyping of MCI in clinical settings.