Multilevel threshold image segmentation with diffusion association slime mould algorithm and Renyi's entropy for...

Area of Science:

• Computer Vision
• Medical Image Analysis
• Artificial Intelligence

Background:

• Image segmentation is crucial for image analysis and medical diagnostics.
• Existing segmentation methods may struggle with complex medical images and avoiding local optima.
• Accurate segmentation of chronic obstructive pulmonary disease (COPD) CT scans is vital for diagnosis and treatment planning.

Purpose of the Study:

• To propose an improved Slime Mould Algorithm (DASMA) for Renyi's entropy multi-threshold image segmentation.
• To enhance the diversity and convergence speed of the Slime Mould Algorithm.
• To apply the proposed method to segment COPD CT images and evaluate its performance.

Main Methods:

• An improved Slime Mould Algorithm (DASMA) incorporating a diffusion mechanism (DM) and an association strategy (AS).
• Application of DASMA to Renyi's entropy multilevel threshold image segmentation using a non-local means 2D histogram.
• Validation on the Berkeley Segmentation Dataset and Benchmark (BSD) and real-world COPD CT scans.

Main Results:

• The DASMA algorithm demonstrated superior performance in image segmentation compared to existing methods.
• Effective segmentation of COPD CT images was achieved, enabling qualitative and quantitative analysis of lesion tissue.
• Experimental results evaluated by image quality metrics confirmed the algorithm's extraordinary performance.

Conclusions:

• The DASMA-based image segmentation technique offers a robust and effective solution for medical image analysis, particularly for COPD.
• The improved algorithm enhances diagnostic accuracy and aids in developing appropriate treatment plans.
• The method shows significant potential for clinical applications in medical image interpretation.