Medical Image Segmentation Based on Improved Hunter Prey Optimization

Effective segmentation of medical images is the key to diagnosing diseases and evaluating treatment outcomes. At the same time, efficient medical image segmentation can not only reduce the time and cost of disease diagnosis, but also help doctors and patients implement diagnosis and treatment measures faster and better. However, due to the irregular shape of medical image targets, differences in imaging methods, and inherent differences in patient organ structures, medical images face significant challenges in segmentation accuracy and efficiency. This paper applies the Hunter Prey Optimization (HPO) to the field of medical image segmentation for the first time, and uses chaotic logistic mapping to improve and optimize population initialization. Taking fuzzy entropy as the objective function, combined with fuzzy median aggregation, a fuzzy improved Hunter Prey Optimization (FIHPO) is formed. Medical images are segmented under different threshold values by FIHPO, and compared with the original HPO, the improved Coyote Optimization Algorithm (FICOA) and the fuzzy Artificial Bee Colony and aggregation algorithm (FABCA). Finally, using Peak Signal to Noise Ratio (PSNR), Feature Similarity Index Measure (FSIM), and Structural Similarity Index Measure (SSIM) as evaluation criteria, a detailed data analysis was conducted on FIHPO, HPO, FICOA and FABCA. Through experimental analysis and comparison, it is found that the FIHPO can achieve better segmentation results in medical image thresholding.