Perfusion-based segmentation of the human brain using similarity mapping.

Area of Science:

• Neuroimaging
• Medical Physics
• Radiology

Background:

• Accurate segmentation of brain tissues is crucial for understanding neurological conditions.
• Dynamic susceptibility contrast-enhanced MRI (DSC-MRI) provides insights into brain perfusion.
• Current methods may have limitations in precisely segmenting specific tissue types.

Purpose of the Study:

• To introduce and validate a novel method for segmenting human brain MR scans based on perfusion characteristics.
• To assess the reliability of the proposed segmentation technique for white and gray matter.
• To explore the potential of this method for identifying pathological brain regions.

Main Methods:

• The method utilizes a similarity measure between time-intensity curves from DSC-MRI and a reference model.
• Segmentation isolates specific tissues, such as white and gray matter.
• Validation involved segmenting white and gray matter in six healthy controls.

Main Results:

• The segmentation method demonstrated reliable differentiation between white and gray matter.
• The gray-to-white matter ratio of relative regional blood volume was 2.1 +/- 0.2, aligning with existing literature.
• The approach showed potential for segmenting abnormal perfusion patterns.

Conclusions:

• The developed method offers reliable segmentation of white and gray matter in human brain MR scans.
• This technique holds promise for characterizing brain pathologies by segmenting abnormal perfusion regions.
• Further applications may involve combining this method with indicator-dilution analysis for enhanced diagnostic capabilities.