Biological Brain Age Prediction Using Cortical Thickness Data: A Large Scale Cohort Study.

Area of Science:

• Neuroimaging
• Radiology
• Biomedical Engineering

Background:

• Brain age estimation is crucial for health informatics and early neurocognitive disorder prediction.
• Normal brain aging exhibits a predictable pattern detectable through anatomical degeneration.
• Predicting biological age from brain structure is an active area of research.

Purpose of the Study:

• To develop and validate a method for brain age estimation using cortical thickness data.
• To apply advanced machine learning techniques for accurate age prediction from brain MRI.
• To assess the utility of the proposed method in identifying patterns associated with normal aging.

Main Methods:

• Utilized a dataset of 2,911 cognitively normal subjects' brain MRI scans (ages 45-91).
• Applied Sparse Group Lasso (SGL) for feature selection based on anatomical grouping.
• Employed Gaussian Process Regression (GPR) for non-parametric, non-linear age prediction modeling.

Main Results:

• Achieved a mean absolute error of 4.05 years in brain age estimation.
• The proposed method demonstrates performance comparable to or exceeding existing techniques.
• Successfully extracted cortical thinning patterns linked to normal brain aging.

Conclusions:

• The developed brain age estimation model is accurate and reliable.
• This method can serve as a valuable tool for clinicians in diagnosing neurodegenerative diseases.
• Cortical thickness analysis provides a robust biomarker for tracking brain aging.