Segmentation of cDNA Microarray Image Using Fuzzy C-means Algorithm Optimized by Particle Swarm
|Course||Precision instruments and machinery|
|Keywords||Microarray Image Segmentation Particle Swarm Optimization Fuzzy c-means clustering algorithm Segmentation algorithm evaluation|
Gene contains a large number of genetic information , information on these research has far-reaching significance. However, previous research methods in dealing with high-throughput genetic information is inefficient , then the mid-1980s a highly accurate gene chip technology should be shipped raw . Gene chips have a wide range of applications, is a hot research direction . Image processing is an essential gene -chip applications is an important step , through effective image processing chip can be efficiently obtain accurate information contained in the high throughput . Therefore, microarray image processing has a very important significance . This paper focuses microarray image processing research. Image processing including key steps: image preprocessing, locator , image segmentation and segmentation evaluation , signal extraction were also introduced. Image preprocessing and locator is to be able to better and more accurate segmentation carried out . Image processing, segmentation is difficult , split a direct impact on the final signal extraction results. So this will be microarray image segmentation technique as a key . This gene chips for the whole process of image segmentation launched a comprehensive study into the algorithm from the evaluation of segmentation algorithms are carried out in detail. And in the summary of previous segmentation algorithm is proposed based on an approach based on Fuzzy c-means clustering gene chip adaptive image segmentation method , and this algorithm is based on a further improvement was proposed based on particle swarm optimization the Fuzzy c-means clustering of microarray image segmentation algorithm clustering method than the original noise immunity stronger, not easy to fall into local optimum . For a more objective evaluation of the segmentation algorithm , this paper presents multiple segmentation algorithm evaluation criteria , and presents a composite image using the ratio of gene expression measurement accuracy of the final evaluation criterion. For a variety of end-use evaluation criteria used microarray image segmentation methods and the proposed segmentation algorithm was evaluated and compared.