Fast Implementation Method of Protein Spots Detection Based on CUDA / 生物医学工程学杂志

In order to improve the efficiency of protein spots detection, a fast detection method based on CUDA was proposed. Firstly, the parallel algorithms of the three most time-consuming parts in the protein spots detection algorithm: image preprocessing, coarse protein point detection and overlapping point segmentation were studied. Then, according to single instruction multiple threads executive model of CUDA to adopted data space strategy of separating two-dimensional (2D) images into blocks, various optimizing measures such as shared memory and 2D texture memory are adopted in this study. The results show that the operative efficiency of this method is obviously improved compared to CPU calculation. As the image size increased, this method makes more improvement in efficiency, such as for the image with the size of 2,048 x 2,048, the method of CPU needs 52,641 ms, but the GPU needs only 4,384 ms.

Analysis of Protein Spot Characteristics Based on Similarity Map / 中国医学影像学杂志

There are no intuitive and unified standards for selecting features in existing protein gel image point matching method, for which an analysis method based on similarity map is proposed. Firstly, the definition and generation methods of similarity map were presented. Secondly, trait and merits of features such as coordinate similarity, shape-context similarity and morphology similarity were analyzed using similarity map method. Finally, comprehensive utilization of multi-features named product-method which has a better effect than mean-method was proposed based on the results. Many experiments using different 2-DE gel images were carried out to prove the validity of similarity map and product-method. The results showed that similarity map could be used for intuitional, effective analysis of matching performance and to guide the selection and comprehensive utilization of multi-features.