DRPSO:A multi-strategy fusion particle swarm optimization algorithm with a replacement mechanisms for colon cancer...

Area of Science:

• Medical image analysis
• Computational intelligence
• Cancer research

Background:

• Colon adenocarcinoma (COAD) presents diagnostic challenges due to subtle early symptoms and severe complications.
• Accurate segmentation of COAD pathology images is crucial for early diagnosis and treatment planning.

Purpose of the Study:

• To develop and validate a multi-level image segmentation (MIS) method for COAD pathology images.
• To enhance the efficiency and accuracy of COAD detection through improved image segmentation.

Main Methods:

• Proposed a novel multi-strategy fusion particle swarm optimization algorithm (DRPSO) with a replacement mechanism, featuring non-linear inertia weight, sine-cosine learning factors, and population reorganization.
• Integrated DRPSO with non-local mean 2D histogram and 2D Renyi entropy for MIS.
• Validated DRPSO against state-of-the-art algorithms on CEC2020 and CEC2022 test sets.

Main Results:

• DRPSO demonstrated superior convergence accuracy and speed compared to existing algorithms.
• The proposed DRPSO-based MIS method achieved high-quality segmentation of COAD pathology images.
• Performance metrics included PSNR = 23.556, SSIM = 0.825, and FSIM = 0.922.

Conclusions:

• The developed MIS method based on the DRPSO algorithm shows significant potential for assisting in the early diagnosis of COAD.
• This approach offers a promising tool for accurate pathology image segmentation in cancer research.