Abstarct: Early diagnosis of MS is of great importance, as this disease can significantly affect the quality of life of patients. MS is a chronic inflammatory disease that affects the central nervous system, and its correct diagnosis can be challenging. Common methods such as using MRI images require detailed analysis and advanced models to identify the affected areas. In this regard, the use of artificial intelligence and deep learning methods can help improve the accuracy of diagnosis. In this thesis, advanced machine learning and deep learning techniques have been used on spinal MRI images to diagnose MS. First, using the data augmentation technique, the data was preprocessed and their samples were increased. Then, using the NASNetMobile deep neural network, key features were extracted from the images and the feature dimensions were reduced using the t-SNE algorithm. In the final stage, classification was performed using the KNN algorithm and samples with MS, Myelitis, and normal were identified. baxsed on the simulation results, the average accuracy of the proposed method for diagnosing MS is 98.14%, which is an improvement over the compared methods.