Hybrid Approaches for Intracerebral Hemorrhage Stroke Classification Using Deep Learning Techniques

Intracerebral Hemorrhage (ICH) is a type of stroke and it is a second leading fatal condition. Due to ICH, most of the people are losing their individuality in life. The mortality rate of ICH is also increasing every year. This issue needs an emergency treatment and timely classification. Radiologists are now using the Computer Tomography (CT) images for the analysis and identification of ICH but it is a time-consuming process. So, this research study concentrates on the analysis of deep learning models to develop a better solution for solving the existing problem. This research study utilizes the hybrid approaches of VGGNet and ResNet based deep learning approach. Here, small size filters of 3×3 are used to perform the better feature map. Definitely, this deep learning approaches will produce best accuracy and develop an effective model for the classification of ICH when compared to other tools. Here, the dataset is collected from Kalinga Institute of Medical Science, Bhubaneswar. Totally, 1164 CT images are collected from the 62 patients.