Evaluation of methods for detecting perfusion abnormalities after stroke in dysfunctional brain regions

Area of Science:

• Neuroimaging
• Neuroscience
• Clinical Neurology

Background:

• Structural changes are commonly studied in lesion-behaviour research.
• Structurally intact brain areas can be non-functional due to malperfusion.
• Perfusion-weighted MRI (PWI) detects these functional deficits.

Purpose of the Study:

• To introduce a novel, bias-reduced method for analyzing perfusion deficits using PWI.
• To compare the new method with traditional region of interest (ROI) comparisons.
• To validate the method for whole-brain analysis in neuroscience and clinical settings.

Main Methods:

• Developed a voxelwise inter-hemispheric comparison technique for PWI analysis.
• Compared homologous voxels to account for regional perfusion differences (e.g., grey vs. white matter).
• Utilized spatial normalization of perfusion maps and evaluated time-to-peak (TTP) delay maps.

Main Results:

• The traditional mean contra-region of interest (ROI) comparison is prone to biases, especially in periventricular regions.
• Voxelwise inter-hemispheric comparison effectively reduces these biases by comparing like tissues.
• TTP delay maps with a 3-second threshold showed strong agreement with manual delineation of perfusion deficits.

Conclusions:

• The automated voxelwise method offers a more accurate and reliable approach to analyzing perfusion deficits than traditional methods.
• This technique eliminates observer-dependent reference region selection and is suitable for comprehensive brain analysis.
• The findings enhance the utility of PWI in both neuroscience research and clinical diagnostics for identifying abnormal brain perfusion.