Stroke detection in the brain using MRI and deep learning models

When it comes to finding solutions to issues, deep learning models are pretty much everywhere. Medical image data is best analysed using models based on Convolutional Neural Networks (CNNs). Better methods for early detection are crucial due to the concerning increase in the number of people suffering from brain stroke. Among the several medical imaging modalities used for brain imaging, magnetic resonance imaging (MRI) stands out. When it comes to analysing medical photos, the deep learning models currently utilised with MRI have showed good outcomes. To improve the efficacy of brain stroke diagnosis, we suggested several upgrades to deep learning models in this work, including DenseNet121, ResNet50, and VGG16. Since these models are not purpose-built to solve any particular issue, they are modified according to the present situation involving the detection of brain strokes. To make use of all of these cutting-edge deep learning models in a pipeline, we proposed a strategy based on supervised learning. Results from the experiments showed that optimised models outperformed baseline models.