Related Experiment Video
Updated: May 21, 2025

Identification of Disease-related Spatial Covariance Patterns using Neuroimaging Data
Published on: June 26, 2013
Spatial and signal features of white matter integrity and associations with clinical factors: A CARDIA brain MRI
Faezeh Vedaei1, Dhivya Srinivasan1, Drew Parker2
1AI(2)D, Center for AI and Data Science for Integrated Diagnostics, and Center for Biomedical Image Computing and Analytics, University of Pennsylvania, Philadelphia, PA, USA.
Abstract:
White matter hyperintensities (WMH) may be indicative of age-related cerebrovascular diseases and contribute to cognitive and functional decline. Normal appearing WM (NAWM) adjacent to WMH, termed "penumbra," is known to be vulnerable to future WMH pathology. WM integrity can be evaluated using multiple magnetic resonance imaging (MRI) modalities. We aimed to identify MRI features predictive of WMH growth and to compare the implications of these features based on spatial proximity to existing WMH versus signal features in baseline NAWM. We used baseline and 5-year follow-up MRI scans in 485 middle-aged participants form the Coronary Artery Risk Development in Young Adults (CARDIA). Multimodal MRI at baseline, including fluid attenuated inversion recovery (FLAIR), diffusion tensor imaging (DTI), and cerebral blood flow (CBF), was measured within WM ROIs including baseline WMH and regions that later developed into new WMH, within and external to the baseline penumbra. Overall, we found that 80% of new WMH appeared within the baseline penumbra. We also found lower fractional anisotropy (FA) and CBF and higher FLAIR and median diffusivity (MD) in NAWM at baseline in regions with subsequent WMH growth compared to those without WMH growth. For NAWM regions defined by signal features, subthreshold FA and suprathreshold MD and FLAIR abnormality at baseline were the most robust predictors of WMH growth. Baseline systolic blood pressure had significant associations with baseline abnormalities in NAWM and subsequently with cognitive decline, particularly for FA and MD measures. The findings support the use of DTI as the predictor of WMH growth, which is correlated with subtle, adverse WM alterations and cognitive function years before developing to WMH. The results may contribute to future clinical trials aimed at preserving WM integrity.
Insights
Diffusion tensor imaging (DTI) can predict white matter hyperintensities (WMH) growth by detecting subtle white matter (WM) alterations. These changes in normal appearing white matter (NAWM) are linked to cognitive decline years before WMH develop.
Area of Science:
- Neuroimaging
- Cerebrovascular Health
- Aging Brain
Background:
- White matter hyperintensities (WMH) are linked to cognitive decline and aging.
- Normal appearing white matter (NAWM) near WMH, known as the penumbra, is vulnerable to future pathology.
- Magnetic resonance imaging (MRI) can assess white matter integrity.
Purpose of the Study:
- To identify MRI features predicting WMH growth.
- To compare predictive features based on spatial proximity versus signal characteristics in NAWM.
- To investigate the association between WM integrity, WMH growth, and cognitive decline.
Main Methods:
- Utilized multimodal MRI (FLAIR, DTI, CBF) from baseline and 5-year follow-up scans of 485 middle-aged adults (CARDIA study).
- Analyzed WM regions including baseline WMH, new WMH, and penumbra.
- Assessed fractional anisotropy (FA), median diffusivity (MD), FLAIR, and cerebral blood flow (CBF) in NAWM.
Main Results:
- 80% of new WMH developed within the baseline penumbra.
- Lower FA, CBF, and higher FLAIR, MD in NAWM predicted WMH growth.
- Subthreshold FA and suprathreshold MD/FLAIR abnormalities in NAWM were robust predictors.
- Baseline systolic blood pressure correlated with NAWM abnormalities and cognitive decline.
Conclusions:
- Diffusion tensor imaging (DTI) effectively predicts WMH growth by detecting adverse white matter alterations.
- These subtle WM changes precede WMH development and are associated with cognitive function.
- Findings support DTI's role in identifying individuals at risk for WMH and cognitive decline.

