Application of Neural Networks for Image Analysis of Magnetic Resonance Imaging (MRI) in Multiple Sclerosis Patients

Abstract

In this article is explored the potential of the image analysis in the process of multiple sclerosis diagnosis, through Machine Learning algorithms to build applications that support the specialist during the identification stage. To accomplish this, a neuronal network is trained with a database containing combined with both: multiple sclerosis images as well as disease-freeimages; then a backpropagation algorithm is applied to determine the optimal parameters to classify the new input data, according to its characteristics..