Evaluation of ResNet Architecture’s Performance for Early Brain Infarction Detection

The main objective of this research, is to compare the performance of ResNet models when deployed on the images of Brain CT scans. In this research three ResNet models are selected with the goal of identifying the most effective one for detecting and classifying cerebral infarcts. These models demonstrated exceptional results, highlighting the promising capabilities of deep learning in the field of medical imaging. The remarkable accuracy, specificity, recall, and sensitivity exhibited by these models generate enthusiasm for the potential application of deep learning in real-world clinical settings. Beyond merely identifying infarcts, these findings suggest that leveraging deep learning could contribute to refining model performance in practical, clinical scenarios.