Implementing ResNetv2_50 with SVM to enhance accuracy in the classification of multiple sclerosis outperforms VGG19

The intention of this research is to improve accuracy for the classification of multiple sclerosis systems from MRI images by using ResNetv2_50 compared with VGG19.Improved the accuracy of multiple sclerosis MRI images by comparing ResNetv2_50 with VGG19. The analysis has been conducted using the SPSS software to enhance the overall results. There were two groups, groups 1 and 2, and each of them had 20 samples. The G power and alpha values obtained were 80% and 0.001. The accuracy performance is increased by ResNetv2_50 and VGG19 for multiple sclerosis by 85.56% and 78.60%, respectively. In this case, the t-test for the independent sample value is obtained as 0.001 (p<0.05). This suggests there are significant statistical differences in their results between each of the two methods. This concludes that ResNetv2_50 has better accuracy performance in comparison to VGG19 in multiple sclerosis. ResNetv2_50 demonstrates high accuracy in estimating multiple sclerosis.