Automatic quantitative analysis and localisation of protein expression with GDF

When detecting the difference of protein expression between normal and cancerous tissues, the shape measurement of protein mostly depends on semi-automatic analysis of image software, which makes the results vulnerable to subjective factors. In this paper, GDF (generalised data field) is proposed to discover protein expression region and further locate it by taking cell nucleus as a reference. Based on the potential distribution, pixels of the image are firstly divided into different clusters. Each cluster represents protein expression in a different degree to precisely describe the details. Then, the clusters are merged into two groups under the requirements of experts or users. Meanwhile, the shapes of cell nuclei are measured, which favours the localisation of the protein expression. Compared with KM and EM, experimental results demonstrate that by using GDF, the protein can be extracted from an image easily and objectively, and the noises of background are further eliminated. Automatic quantitative analysis and localisation of protein expression.