Transfer Learning-Based Classification Comparison of Stroke

One type of brain disease that significantly harms people's lives and health is stroke. The diagnosis management strokes both heavily rely on the quantitative analysis Magnetic Resonance (MR) images. early process great importance for prevention stroke cases. Stroke prediction made possible by deep neural networks with capacity enormous data learning. Therefore, in thus study, several network models, including DenseNet121, ResNet50, Xception, MobileNet, VGG16, EfficientNetB2 are proposed transfer learning to classify MR images into two categories (stroke non-stroke) order study characteristics lesions achieve full intelligent automatic detection. dataset comprises 1901 training images, 475 validation 250 testing On sets, augmentation was used increase number improve models’ experimental results outperform all state arts were same dataset. overall accuracy best model 98.8% value precision, recall, f1-score using