Diagnosis of kashin-beck disease and other common joint diseases via a gene model

Kashin-Beck disease and common joint disease display similar early clinical features, result in the difficulty of identification of diagnosis and treatment by observing the clinical manifestations and physical signs. In this paper, we use two sets of gene expression data from cartilaginous tissues and a health tissues, collected from 16 samples. We propose a gene model to achieve the diagnosis of Kashin-Beck disease and other common joint diseases using the expression. In the model, we try to filter various noises caused by improper operation, aging of equipment and so on by wavelet transform. Bhattacharyya distance and Pearson correlation coefficient are used to remove irrelevant genes and redundant genes respectively and acquire the rest as feature genes. In order to deal with the comprehensive data, we choose support vector machine (SVM) to achieve identification. The experiment results demonstrate that the gene model proposed in this paper offers a set of feature genes for further biological interpretation and play an important role in the diagnosis.