A Hybrid Method for Image Segmentation Based on Artificial Fish Swarm Algorithm and Fuzzy c-Means Clustering

Area of Science:

• Medical image processing
• Artificial intelligence in healthcare
• Computational imaging

Background:

• Fuzzy c-means (FCM) is a popular medical image segmentation technique.
• FCM suffers from sensitivity to initial centers, local optima, and noise.
• Improved segmentation is crucial for accurate medical diagnosis.

Purpose of the Study:

• To address FCM limitations in medical image segmentation.
• To introduce a hybrid algorithm for enhanced segmentation performance.
• To improve noise resistance and precision in medical imaging.

Main Methods:

• Developed a Hybrid Artificial Fish Swarm Algorithm (HAFSA).
• Integrated Artificial Fish Swarm Algorithm (AFSA) with FCM.
• Incorporated Metropolis criterion and noise reduction into AFSA.
• Tested on artificial grid graphs and Magnetic Resonance Imaging (MRI) data.

Main Results:

• HAFSA demonstrated superior global optimization searching and parallel computing.
• The algorithm exhibited enhanced convergence rate and anti-noise capabilities.
• Experimental results showed HAFSA outperformed FCM and suppressed FCM (SFCM) in precision.
• Evaluation indicators confirmed HAFSA's excellent performance.

Conclusions:

• HAFSA effectively overcomes FCM's limitations in medical image segmentation.
• The proposed hybrid approach offers improved accuracy and robustness against noise.
• HAFSA represents a significant advancement for medical image analysis and diagnostic applications.