Brain blood vessel segmentation using line-shaped profiles

Area of Science:

• Medical Imaging
• Neurosurgery
• Computational Anatomy

Background:

• Accurate segmentation of cerebral blood vessels is critical for diagnosing and treating neurological conditions like aneurysms and arteriovenous malformations (AVMs).
• Current segmentation methods often require complex fusions of techniques and significant user input, hindering efficiency in clinical workflows.

Purpose of the Study:

• To introduce a novel, efficient line-shaped profiling technique for segmenting brain blood vessels and AVMs.
• To reduce user interaction and computational time in the segmentation process.

Main Methods:

• Developed a line-shaped profiling approach considering local and neighborhood pixel information.
• Applied the method to segment brain vasculature and AVM structures from 3D imaging data.

Main Results:

• The line-shaped profiling method demonstrated efficiency in resolving fine details and large vessel structures.
• Achieved accurate segmentation of complex AVMs with significantly reduced user interaction.
• The technique requires selection of only one parameter for segmentation.

Conclusions:

• Line-shaped profiling offers an efficient and user-friendly alternative for cerebral blood vessel and AVM segmentation.
• This method has the potential to improve pre-surgical planning for AVM embolization and other neurovascular interventions.