Principal Component Analysis of Diffusion Tensor Images to Determine White Matter Injury Patterns Underlying...

Area of Science:

• Neuroimaging
• Neuroscience
• Medical Data Analysis

Background:

• Mild traumatic brain injury (mTBI) can lead to persistent posttraumatic headache (PTH).
• Identifying specific white matter injury patterns associated with PTH is crucial for understanding mTBI outcomes.
• Diffusion tensor imaging (DTI) provides insights into white matter integrity.

Purpose of the Study:

• To utilize principal component analysis (PCA) on fractional anisotropy (FA) maps to detect white matter injury patterns correlating with PTH after mTBI.
• To establish a data-driven approach for identifying neuroimaging biomarkers for PTH.

Main Methods:

• DTI and neurocognitive testing were performed on mTBI patients and controls.
• PCA was applied to coregistered FA maps to identify major components.
• Regression analysis correlated principal components with PTH, followed by screening against other symptoms and demographics.

Main Results:

• Principal component 4 (PC4) significantly correlated with PTH in mTBI patients (OR=2.32, P=.01).
• Decreased FA in the corpus callosum's midsplenium and increased FA in the genu were associated with PC4.
• PC4 demonstrated good discriminatory power for PTH (AUC=0.73) and was specific to this symptom.

Conclusions:

• PCA is an effective data-mining tool for linking DTI-based white matter injury patterns to clinical symptoms in mTBI.
• A specific pattern of FA changes in the corpus callosum (splenium and genu) identified by PCA can help predict PTH risk post-mTBI.