Application of a gene diagnosis model in the identification of Kashin-Beck disease and similar joint diseases

Kashin-Beck disease (KBD), a local osteoarthrosis with high morbidity and disability rate, seriously affects the labor and health quality of the population in the disease area. In traditional way, it's unreliable to distinguish KBD from other common joint diseases just by analyzing the early clinical manifestations and physical signs. Thus, this paper proposes a gene diagnosis model using two sets of gene expression data from patients' cartilaginous tissues to diagnose illness exactly at the genetic level. Firstly, wavelet transform is used to filter noises caused in the process of genetic sequencing. Then, we select both up-regulated genes and down-regulated genes as a feature gene set and further decrease the number of the feature genes through Student't test. Next, Pearson correlation coefficient are applied to remove redundant genes and find out the final feature genes. Finally, support vector machine (SVM) as the classifier can help to distinguish between common joint diseases. According to the experiment results, the gene diagnosis model performs well and also provide sets of effective genes for further clinical translational research.