Corona COVID 19 hastalarının aksesuar röntgeni görüntülerinden sınıflandırılması

For medical professionals to successfully focus their attention and give therapy to patients, it is essential for them to have an accurate diagnosis and categorization of COVID 19. Because COVID 19 has developed into a leading cause of mortality, the medical community must give it significant scrutiny. As a result, one of our objectives is to create a structure, with the purpose of conducting research on the collection of photographs collected from the Kaggle platform. In order to accomplish this goal, we used image merging strategies into our research. At first, we made use of a Convolutional Neural Network, abbreviated as CNN. further on, a strategy was created by merging CNN with Gated Recurrent Units (GRU). This method, which we describe to as our recommended model, was further developed. Through a preprocessing phase that entailed combining pictures, we were able to increase the overall quality of the CT scans. In addition, the color format of the photographs was changed from BGR to RGB by us. In addition, the dataset containing the micro-CT images was broken up into sets for the purposes of testing, training, and validation. According to the confusion matrix analysis, the findings collected by the network (CNN) demonstrated a remarkable accuracy rate of 98%. In addition to that, the value of 97% was displayed in the correlation matrix of our CNN-GRU model.