Comparison Results of Medical Image Segmentation with Genetic Algorithm and Particle Swarm Optimization

Digital-based medical images play an important role in modern health services. Image processing in the form of segmentation imposed on medical images is carried out in order to obtain clear boundaries on images based on certain characteristics. This is expected to reduce errors in further image analysis. In terms of the number of threshold values, the multilevel thresholding method will be applied in this research instead of the bi-level one. To find out the threshold value, the genetic algorithm (GA) and the particle swarm optimization (PSO) algorithm are implemented. These two metaheuristic methods maximize efficiently the form of Shannon, Renyi, and Masi entropies. Furthermore, the threshold value will be applied in the segmentation process of medical images taken from a hospital in Central Java, known as RSUD Kraton Pekalongan. The results of this segmentation are compared by using some image quality indices, including PSNR (Peak Signal to Noise Ratio) and SSIM (Structural Similarity Index Measure) to figure out the most effective method-entropy combination.