Quantifying metabolic asymmetry modulo structure in Alzheimer's disease

Area of Science:

• Neuroimaging
• Medical Image Analysis
• Computational Anatomy

Background:

• Metabolic asymmetry in the brain is a potential biomarker for neurological disorders like Alzheimer's disease (AD).
• Quantifying metabolic asymmetry is challenging due to confounding structural hemispheric differences.
• Existing methods may not adequately disentangle metabolic changes from anatomical variations.

Purpose of the Study:

• To introduce a new computational framework for quantifying metabolic asymmetry in the brain, independent of structural variations.
• To establish a method for comparing metabolic differences between brain hemispheres in individuals and populations.
• To apply this methodology to neuroimaging data from Alzheimer's Disease Neuroimaging Initiative (ADNI) cohort.

Main Methods:

• Utilizing large deformation diffeomorphic metric mapping (LDDMM) for anatomical atlas construction.
• Defining a structurally symmetric coordinate frame based on LDDMM invariance properties.
• Mapping individual metabolic asymmetry to a common population-wide symmetric coordinate system.

Main Results:

• Demonstrated the feasibility of creating subject-specific structurally symmetric coordinate systems.
• Successfully quantified metabolic asymmetry modulo structural differences in AD patients.
• Established a statistical framework for population-level analysis of metabolic asymmetry using LDDMM.

Conclusions:

• The proposed LDDMM-based method effectively quantifies metabolic asymmetry while controlling for structural variations.
• This approach offers a robust tool for investigating brain alterations in Alzheimer's disease and other neurological conditions.
• The framework facilitates population-level statistical analysis of brain asymmetry in neuroimaging studies.