Medical Image Segmentation Using Fruit Fly Optimization and Density Peaks Clustering.

Area of Science:

• Medical Imaging
• Computer Vision
• Artificial Intelligence

Background:

• Medical image segmentation is crucial for diagnosis and treatment planning.
• Traditional methods like Density Peaks Clustering (DPC) often require manual parameter tuning.
• Optimization algorithms can improve segmentation but may suffer from local optima.

Purpose of the Study:

• To develop a novel, automated medical image segmentation algorithm.
• To overcome limitations of existing Density Peaks Clustering and fruit fly optimization algorithms.
• To enhance segmentation accuracy and robustness in medical imaging.

Main Methods:

• A hybrid algorithm combining Density Peaks Clustering (DPC) with a modified Fruit Fly Optimization Algorithm (FOA).
• Automated parameter selection for DPC, eliminating manual input.
• Introduction of random step size in FOA to prevent local optima.
• Utilization of image entropy for selecting cut-off distance and cluster centers.

Main Results:

• The proposed algorithm automatically determines parameters, unlike standard DPC.
• Random step size in FOA improves convergence and avoids local optima.
• Image entropy-based selection enhances structural feature capture.
• Experiments demonstrate faster convergence and superior robustness on benchmark and proprietary datasets.

Conclusions:

• The novel algorithm offers an adaptive and robust solution for medical image segmentation.
• It overcomes key limitations of existing clustering and optimization techniques.
• The method shows significant potential for improving clinical image analysis workflows.