Age Correction in Dementia: Matching to a Healthy Brain
Author Information
Author(s): Juergen Dukart, Matthias L. Schroeter, Karsten Mueller
Primary Institution: Max Planck Institute for Human Cognitive and Brain Sciences
Hypothesis
Applying linear age detrending prior to statistical evaluation should increase the diagnostic accuracy for differentiation of dementia patients and control subjects using SVM.
Conclusion
The proposed age correction method improves the classification accuracy in differentiating Alzheimer's disease patients from healthy controls.
Supporting Evidence
- SVM classification accuracy improved from 83.0% to 85.0% after age correction.
- Age correction reduced misclassification rates between AD patients and controls.
- VBM analyses showed significant differences in GM atrophy patterns when age was included as a covariate.
Takeaway
This study found a way to make sure that age doesn't mess up the results when looking at brain scans of people with dementia and healthy people.
Methodology
The study used support vector machine classification and voxel-based morphometry to analyze MRI data from Alzheimer's disease patients and healthy controls, comparing results with and without age correction.
Potential Biases
Potential mutual correlation between age and other covariates used in subsequent analyses.
Limitations
The method may not be applicable in studies where matching is possible, and the group size for controls must be sufficiently large to provide robust estimates.
Participant Demographics
80 Alzheimer's disease patients and 79 healthy control subjects, matched for gender, age, and education.
Statistical Information
P-Value
0.004
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website