Resumen

SAURABH, Sudhanshu  y  GUPTA, P. K.. Detection and Classification of Multiple Sclerosis from Brain MRIs by Using MobileNet 2D-CNN Architecture. Comp. y Sist. [online]. 2024, vol.28, n.3, pp.1229-1242.  Epub 21-Ene-2025. ISSN 2007-9737.  https://doi.org/10.13053/cys-28-3-4197.

Deep learning-based object detection and classification have been widely investigated for neuroimaging. Magnetic resonance imaging (MRI) data serve as a diagnostic tool for the detection and classification of brain disorders such as Parkinson’s, Alzheimer’s disease (AD), and Multiple Sclerosis (MS). In addition, the use of the Convolutional Neural Network (CNN) framework helps in the development of predictive models from the available MRI images. This work aims to develop a CNN-based model with a pre-trained MobileNet model to detect and classify multiple sclerosis using the MRI image dataset. In this article, we have proposed a pre-trained MobileNet-2D-CNN architecture for the accurate prediction of multiple sclerosis from various MRI images. Initially, the proposed model extracted images from MRI images of the affected patient with MS and healthy control. We used MRI images to train the MobileNet-2D-CNN model to identify the MS features map that predicts MS. The proposed architecture has been validated on standard MRI scans. We also performed a class activation map for the interpretation of the prediction provided by the proposed model, which represents the behavior of neurons in the early stages. The proposed approach achieves a classification precision of 98.15% and AUC=1.00.

Palabras llave : CNN; deep learning; feature map; mobilenet; MRI; multiple sclerosis.

        · texto en Inglés     · Inglés ( pdf )