Deep learning based intelligent diagnostic model detecting for brain hemorrhage using computed tomography images

Traumatic brain injuries can cause brain hemorrhage. If these bleedings are not treated in a short time and with the right treatment methods, they may result in disability or death. At the present time, the protocol applied in health institutions to diagnose and determine the location of cerebral hemorrhage requires the examination and interpretation of computed tomography scans by expert radiologists. This may cause misinterpretation from experts. These misinterpretations made by experts can cause irreversible damage to patients. Intelligent diagnostic models designed for this purpose are currently used in a wide variety of application areas to help today's healthcare experts to diagnose more effectively and quickly.In this study, it is aimed to develop an intelligent diagnosis model in order to quickly detect brain hemorrhage and hemorrhage area from computed tomography images by investigating deep learning methods, which are a sub-branch of machine learning. Deep learning algorithms have gained use in many different applications, especially in the field of health. For this purpose, U-Net Architecture, which provides successful results in medical image processing and segmentation applications in deep learning algorithms, will be used. It is planned to make the performance evaluations of the U-net architecture to be used. It is aimed to present performance evaluation comparatively over existing deep learning models.