Integrating spatial fuzzy clustering with level set methods for automated medical image segmentation

Area of Science:

• Medical Imaging
• Computer Vision
• Image Processing

Background:

• Level set segmentation methods often require significant manual effort for initialization and parameter tuning.
• Optimizing controlling parameters is crucial for the performance of level set segmentation.

Purpose of the Study:

• To develop a novel fuzzy level set algorithm for automated medical image segmentation.
• To reduce manual intervention in the segmentation process.
• To enhance the robustness and accuracy of medical image segmentation.

Main Methods:

• A new fuzzy level set algorithm is proposed, utilizing spatial fuzzy clustering for direct evolution from initial segmentation.
• Controlling parameters for level set evolution are estimated from fuzzy clustering results.
• The algorithm is enhanced with locally regularized evolution for improved manipulation and robustness.

Main Results:

• The proposed fuzzy level set algorithm demonstrates effectiveness in medical image segmentation across various modalities.
• The method successfully automates initialization and parameter estimation through fuzzy clustering.
• Enhanced locally regularized evolution contributes to more robust segmentation outcomes.

Conclusions:

• The developed fuzzy level set algorithm offers an effective and less manual approach to medical image segmentation.
• The integration of fuzzy clustering and locally regularized evolution significantly improves segmentation performance.
• The algorithm shows promise for diverse medical imaging applications requiring accurate segmentation.