Interpretation for Individual Brain Age Prediction Based on Gray Matter Volume.

Area of Science:

• Neuroscience
• Medical Imaging
• Machine Learning

Background:

• The relationship between aging and the central nervous system (CNS) is a critical area of research.
• Existing methods primarily focus on predicting age or disease from MR images, with less emphasis on underlying mechanisms.
• Understanding individual aging mechanisms is crucial for diagnosing degenerative brain diseases.

Purpose of the Study:

• To explore the correlation between gray matter volume (GMV) and individual age.
• To investigate the dynamic changes in GMV and their interaction networks with aging.
• To develop an interpretable machine learning approach for studying brain aging mechanisms.

Main Methods:

• Utilized interpretable machine learning models to analyze the relationship between GMV and age.
• Examined both individual gray matter regions and their interaction networks.
• Investigated the dynamic properties of these brain structures in relation to individual age.

Main Results:

• Achieved a mean absolute error (MAE) of 7.95 years in age prediction.
• Identified specific gray matter locations and their interactions that play distinct roles in aging.
• Demonstrated that these roles change dynamically across the lifespan.

Conclusions:

• The data-driven approach provides novel insights into brain aging mechanisms.
• The findings offer a new avenue for the diagnosis of age-related degenerative brain diseases.
• Highlights the dynamic nature of brain structure changes throughout the aging process.