Now neural networks are used in several domains. One important application is the diagnosis process. This paper by using the neural networks can be taken a step in the diagnosis of bipolar disorder. Of the many ways such as interviews, patient records, relevant articles, books psychiatric. Etc. Parameters of this disorder have been identified. After identifying the parameters of bipolar disorder, the implementation of neural network models are discussed. MLP and RBF have models that used in this paper. At the end, the comparison between the results of the two models is done. And conclude the MLP model with the lower error (13.33%) than RBF model, can diagnosis the bipolar disorder.

Full text: IJEIR_1173_Final.pdf