Edge-based partial volume averaging estimation for FLAIR MRI with white matter lesions

Area of Science:

• Medical Imaging
• Quantitative MRI
• Neuroimaging

Background:

• Partial volume averaging (PVA) poses challenges in accurately segmenting tissues in Magnetic Resonance Imaging (MRI).
• White matter lesions (WML) in FLAIR MRI often exhibit partial volume effects, complicating precise quantification.

Purpose of the Study:

• To develop and validate a new edge-based quantification technique for partial volume averaging (PVA) in FLAIR MRI, specifically for white matter lesions (WML).
• To enable accurate segmentation of PVA and pure tissue classes using intensity and fuzzy edge strength measures.

Main Methods:

• A novel PVA quantification technique combining intensity and fuzzy edge strength measures was developed.
• An edge-based approach was employed, probing for PVA voxels using a global estimate of tissue proportion change (α").
• The estimate was refined via a probabilistic threshold to determine the tissue fraction (α) in mixture voxels.

Main Results:

• The technique successfully segmented partial volume averaging (PVA) and pure tissue classes in FLAIR MRI.
• Demonstrated accurate quantification of tissue proportions within mixture voxels.
• Illustrated the utility of the method across several example images.

Conclusions:

• The developed edge-based method provides a non-model based approach for the detection and quantification of PVA in FLAIR MRI WML.
• This technique offers improved accuracy in segmenting and quantifying tissue mixtures, crucial for lesion analysis.