BrainAgeNeXt: Advancing Brain Age Modeling for Individuals with Multiple Sclerosis.

Area of Science:

• Neuroimaging
• Artificial Intelligence
• Biomarker Development

Background:

• Aging affects brain structure and cognitive function, with brain age serving as a key imaging biomarker.
• Deviations in brain age from chronological age can indicate neurodegenerative processes.
• Accurate brain age prediction is crucial for understanding healthy aging and disease progression.

Purpose of the Study:

• To introduce BrainAgeNeXt, a novel convolutional neural network for predicting brain age from T1-weighted MRI scans.
• To evaluate BrainAgeNeXt's performance against existing state-of-the-art methods.
• To assess BrainAgeNeXt's utility as a prognostic biomarker in multiple sclerosis (MS).

Main Methods:

• Developed BrainAgeNeXt, a deep learning model based on the MedNeXt framework.
• Trained and validated the model on 11,574 MRI scans from diverse datasets (healthy volunteers, ages 5-95, 3T/7T MRI).
• Compared BrainAgeNeXt's performance using Mean Absolute Error (MAE) against three other methods and tested robustness across varying image quality.

Main Results:

• BrainAgeNeXt achieved a lower MAE (2.78 ± 3.64 years) compared to existing methods (3.55, 3.59, 4.16 years).
• The model demonstrated robust performance even with motion artifacts and 7T MRI data.
• In MS cohorts, brain age exceeded chronological age (4.21 ± 6.51 years), with a faster increase in those experiencing worsening disability.

Conclusions:

• BrainAgeNeXt is a highly accurate and robust tool for brain age prediction from MRI.
• Brain age is a potential prognostic biomarker for MS progression.
• Brain age may serve as a valuable endpoint in clinical trials for neurodegenerative diseases like MS.