Related Experiment Video
Updated: Oct 5, 2025
06:35
Basics of Multivariate Analysis in Neuroimaging Data
Published on: July 24, 2010
17.0K
Detecting Biomarkers of Alzheimer's Disease Based on Multi-constrained Uncertainty-Aware Adaptive Sparse Multi-view
Wenbo Wang1, Wei Kong2, Shuaiqun Wang1
1College of Information Engineering, Shanghai Maritime University, 1550 Haigang Ave., Shanghai, 201306, People's Republic of China.
Journal of Molecular Neuroscience : MN
|January 26, 2022
Summary
This study introduces a new method, MC-unAdaSMCCA, to analyze genetic and brain imaging data for Alzheimer's disease (AD) research. The method improves correlation analysis and identifies biomarkers for AD and mild cognitive impairment (MCI).
Area of Science:
- Neuroscience
- Genetics
- Medical Imaging
Background:
- Image genetics links genetic data (SNP, gene expression, methylation) with brain imaging (sMRI, fMRI, PET) to understand Alzheimer's disease (AD) pathogenesis.
- Existing methods often struggle with gradient domination and feature redundancy, limiting correlation analysis and computational efficiency.
- There is a need for advanced methods to accurately explore multi-modal imaging genetic associations.
Purpose of the Study:
- To propose a novel multi-constrained uncertainty-aware adaptive sparse multi-view canonical correlation analysis (MC-unAdaSMCCA) method.
- To enhance the analysis of associations among single nucleotide polymorphisms (SNPs), gene expression data, and structural MRI (sMRI).
- To improve correlation detection, reduce computational complexity, and speed up analysis in multi-modal imaging genetics.
Main Methods:
- Developed MC-unAdaSMCCA by imposing orthogonal constraints on feature weights via linear programming, building upon traditional unAdaSMCCA.
- Reduced feature weight redundancy to enhance data correlation and algorithm efficiency.
- Benchmarked MC-unAdaSMCCA against three other adaptive sparse multi-view canonical correlation analysis methods using real and synthetic neuroimaging data.
Main Results:
- MC-unAdaSMCCA achieved better or comparable typical correlation coefficients and weights compared to benchmark methods.
- The method demonstrated effectiveness in identifying biomarkers associated with Alzheimer's disease (AD) and mild cognitive impairment (MCI).
- MC-unAdaSMCCA exhibited robustness against noise and proficiency in processing high-dimensional data.
Conclusions:
- The proposed MC-unAdaSMCCA method offers a reliable approach for multi-modal imaging genetic research.
- It effectively addresses limitations of existing methods, improving accuracy and speed.
- This method facilitates biomarker discovery for neurodegenerative diseases like AD and MCI.
