16+
DOI: 10.18413/2518-1092-2025-10-1-0-8

RECOGNITION AND CLASSIFICATION OF MRI IMAGES OF THE BRAIN USING THE NEURAL NETWORKS

The article presents a study devoted to the development of neural network tools for classifying magnetic resonance imaging (MRI) images. The study is devoted to solving a pressing scientific and technical problem aimed at improving the accuracy of diagnosing oncological diseases of the human brain. A model based on the use of a feedforward neural network is proposed. The model has three hidden layers of neurons using the Relu activation function. The output layer of neurons uses the Softmax activation function. The network was created using the Keras library and the OpenCV software library. The sizes of images used as training data are substantiated. The study showed that 38 training cycles are sufficient to configure such a neural network. The performance of the proposed neural network was tested, which showed high accuracy of image classification results. The use of this model allows to increase the accuracy of diagnosing oncological diseases of the human brain by 9.6% compared to traditional methods.

Number of views: 20 (view statistics)
Количество скачиваний: 54
Full text (PDF)To articles list
  • User comments
  • Reference lists

While nobody left any comments to this publication.
You can be first.

Leave comment: