Machine Learning Predicts Brain Age from Cerebral Blood Flow
In a study, researchers employed transcranial Doppler (TCD) to assess the velocity of cerebral blood flow and utilized machine learning techniques to estimate chronological age while identifying accelerated cerebrovascular aging in individuals with health issues. The investigation involved 168 healthy participants and 277 individuals with diseases, all of whom underwent bilateral TCD recordings of the middle cerebral artery through the MOCAIP algorithm. Data derived from MOCAIP and heart rate variability were fed into regression models developed using healthy subjects to predict age. It was found that diseased individuals exhibited signs of accelerated aging. This method seeks to characterize vascular age based on physiological performance.
Key facts
- Transcranial Doppler (TCD) measures cerebral blood flow velocity in major brain arteries.
- 168 healthy subjects and 277 diseased subjects were analyzed.
- Bilateral TCD recordings of the middle cerebral artery were used.
- Morphological Analysis and Clustering of Intracranial Pressure (MOCAIP) algorithm was applied.
- MOCAIP-generated features and heart rate variability features served as inputs for regression models.
- Models were trained on healthy subjects to predict chronological age.
- Diseased subjects were expected to show accelerated cerebrovascular aging.
- The study aims to define vascular age in terms of physiological function.
Entities
—