Increased MRI-based Brain Age in chronic migraine patients

Area of Science:

• Neuroimaging
• Machine Learning
• Brain Aging

Background:

• Migraine is associated with brain structure and function alterations.
• The impact of aging on these brain changes in migraine is not well understood.
• The Brain Age framework offers a novel approach to investigate brain aging in neurological conditions.

Purpose of the Study:

• To investigate brain aging patterns in migraine patients using a machine learning-based Brain Age model.
• To determine if migraine patients exhibit an increased Brain Age Gap (predicted age minus chronological age) compared to healthy controls.
• To explore the relationship between brain age and migraine characteristics.

Main Methods:

• A machine learning model was trained to predict Brain Age using 2,771 T1-weighted MRI scans from healthy subjects.
• The model utilized 1,479 imaging features (morphological and intensity-based) within a 10-fold cross-validation.
• The trained model was applied to a cohort of 247 subjects, including healthy controls, episodic migraine (EM), and chronic migraine (CM) patients.

Main Results:

• Chronic migraine patients demonstrated a significantly increased Brain Age Gap (4.16 years) compared to healthy controls (-0.56 years).
• Episodic migraine patients showed a trend towards an increased Brain Age Gap (1.21 years), but it was not statistically significant.
• No correlation was found between the Brain Age Gap and migraine frequency, duration, or headache frequency. Key imaging features driving age differences were previously linked to migraine.

Conclusions:

• Brain-predicted age serves as a sensitive biomarker for chronic migraine patients.
• This approach reveals distinct brain aging patterns in individuals with chronic migraine.
• Neuroimaging-based brain age estimation can offer insights into the pathophysiology of migraine.