ResNet-CBAM in Medical Imaging: A High-Accuracy Tool for Stroke Detection from CT Scans

This research aims to analyze the ResNet-CBAM deep learning model in detecting strokes from brain CT scan images and its web-based implementation. Stroke is a serious condition caused by disruption of blood flow to the brain, leading to brain cell death and is one of the leading causes of death and disability worldwide. Early detection and accurate diagnosis are essential to reduce its impact. With the advancement of artificial intelligence (AI) and deep learning technology, especially Convolutional Neural Network (CNN), medical image analysis has experienced significant improvements. The ResNet-CBAM model combines a Residual Network architecture and a Convolutional Block Attention Module, designed to improve the detection of important features from CT scan images. Test results show that this model has 95% accuracy in classifying normal and stroke conditions, with precision, recall and F1-score reaching 96% respectively for the non-stroke class and 94 %, 93 % and 93 % for the stroke class. The ROC curve shows excellent classification capabilities with AUC values reaching 1.00 for training data and 0.99 for testing data. This application allows users to upload CT scan images, process them, and get prediction results quickly and accurately. This provides great potential to be used as an aid in the early diagnosis of stroke, helping medical personnel make faster and more appropriate decisions, and improving the quality of life of patients through early intervention.