Medical image segmentation method based on FCM algorithm in fusion with improved GSA algorithm

Abstract

The invention discloses a medical image segmentation method based on an FCM algorithm in fusion with a GSA algorithm. According to the approximate process of the method, an optimal clustering center determination method of the FCM algorithm and the improved GSA algorithm are fused, a spatial position, namely an optimal solution, of the heaviest particle is solved step by step by property adjusting active gravitational mass, passive gravitational mass and inertial mass of particles, and the optimal solution is used as an optimal clustering center of pixels during clustering. According to the method, a spatial fuzzy c-means clustering algorithm is utilized to divide the image pixels into uniform regions, the FCM algorithm and the GSA algorithm are fused, the improved GSA algorithm is included into the fuzzy c-means clustering algorithm to find the optimal clustering center, the fitness function value of fuzzy c-means clustering is minimum, and therefore the segmentation effect is improved. Experiment results show that compared with traditional clustering algorithms, the method is more effective in terms of medical images complicated in segmentation.

Description

technical field [0001] The invention relates to a medical image segmentation method based on FCM algorithm fusion and improved GSA algorithm. Background technique [0002] Image segmentation technology has always been a research hotspot in the field of medical image processing, but there is no general segmentation algorithm suitable for all image segmentation. At present, there is no effective general solution and unified standard for medical image segmentation. One of the reasons is that medical images have extremely complex diversity, the complexity of human anatomy, the irregular shape of tissues and organs, and the anatomy between people. There are considerable differences in organizational structure, which increases the difficulty of medical image segmentation, and needs to solve problems such as blurring, unevenness, individual differences, and complexity. Its development is a gradual process from manual segmentation to semi-automatic segmentation and automatic segment...

AI Technical Summary

Problems solved by technology

Aiming at the characteristics of the complexity of medical image processing, this paper uses the classic FCM algorithm to conduct image segmentation research in the early stage. By segmenting images of different complexity, it is found that because the traditional FCM algorithm only considers the grayscale characteristics of pixel values, it does not consider the image. Spatial structure information, which isolates the pixels between images, and lacks the thinking about the relationship between these pixels, makes the FCM algorithm more sensitive to the noise in medical images, resulting in the distinction between noise and real feature points in a certain class Not open, or even the same, resulting in poor anti-noise performance of the algorithm and unsatisfactory image segmentation quality

Images