Main Article Content
The Rheumatoid Arthritis (RA) affects many people targeting their joints, especially small joints, and it targets all ages which it is more common in women. This disease has many symptoms similar to other diseases. Therefore, it is very hard to detect. Also, the diagnostic tools are complex and uneconomical. In this paper, artificial intelligence network used for diagnosis and early detection of RA in accordance with criteria developed by the American College of Rheumatology. The best performance occurs with the minimum number of neurons required when the number of neurons is 6. So that, the performance is equal to 3.8968x1010-. When reducing the number of neurons to 5 or increasing to 8, the result is 0.0041 and 1.0611×10-10, respectively. However, all results can be consider acceptable and indicate that the best choice from this structure will be 6 neurons in the form of complexity and accuracy.