An AI Enabled Framework for MRI-based Data Analytics for Efficient Brain Stroke Detection

These days, it seems that medical image processing is a good fit for deep learning-based models. There is a potential basis for exploiting healthcare services thanks to advancements in deep learning and pre-trained models in computer vision applications. According to the World Health Organization (WHO), a clinical decision support system (CDSS) based on deep learning has the potential to advance the state of the art in medical image analysis. Timely management is essential due to the high death and disability rates associated with stroke. Timely study of brain imaging allows for rapid medical action. Existing research on MRI-based brain stroke analysis requires a refined and enhanced strategy to reap the potential advantages of deep learning, and MRI is discovered to provide additional possibilities for medical picture analysis. As may be seen in the material, a lot of focus has been placed here. In this study, we present the Deep Automated Brain Stroke Detection Framework (DABSDF), a deep learning-based system for detecting strokes in brain MRI. The Deep Convolutional Neural Network-based Pipeline for Brain Stroke Detection is a method we proposed (DCNNP-BSD). To test the effectiveness of the proposed framework and its algorithms, a prototype application has been developed in the Python data science environment. We evaluate our model against current deep learning models. The effectiveness of the various models on the MRI dataset varies widely. In terms of performance, VGG16 fares the worst while the suggested model, DCNNP-BSD, fares the best. With a dice similarity coefficient of 0.8822979, 0.8554022 sensitivity, 0.99595785 specificity, and 0.97774774 accuracy, the suggested CNN-based deep learning model beat the state-of-the-art.