Alzheimer's Imaging Consortium

Area of Science:

• Neuroimaging
• Biomarker Discovery
• Alzheimer's Disease Research

Background:

• Arterial Spin Labelling (ASL) is a non-invasive MRI technique measuring cerebral blood flow (CBF).
• CBF changes are indicative of Alzheimer's disease (AD) progression.
• ASL-derived CBF may serve as an imaging biomarker for dementia.

Purpose of the Study:

• Introduce a novel AD meta-region of interest (meta-ROI) using ASL data.
• Assess the correlation between the ASL meta-ROI and cognitive scores.
• Evaluate ASL as a potential non-invasive alternative to PET and other biomarkers for tracking disease progression across cognitively normal (CN), mild cognitive impairment (MCI), and AD stages.

Main Methods:

• Analyzed ASL data from the Alzheimer's Disease Neuroimaging Initiative (ADNI3).
• Processed ASL data to obtain calibrated CBF maps using the qasl pipeline.
• Constructed a meta-ROI from 17 ROIs showing linear correlation with ADAS-Cog scores and trained a regression model to predict ADAS-Cog.

Main Results:

• Observed a significant correlation (R=0.72) between the ASL meta-ROI scores and cognitive performance.
• This correlation is comparable to established biomarkers like PET (R=0.75) and CSF/plasma biomarkers.
• The findings suggest ASL's efficacy in reflecting disease progression.

Conclusions:

• The strong correlation supports the use of ASL-derived biomarkers for predicting and monitoring Alzheimer's disease progression.
• ASL offers a promising non-invasive imaging approach for clinical applications in dementia.
• Further validation of ASL meta-ROI as a reliable biomarker is warranted.