The effective disease management depends on the timely and precise diagnosis of COVID-19 patients. As a useful diagnostic technique for COVID-19, chest CT scans provide light on the abnormalities of the lungs linked to the illness. In this paper, we provide a novel machine learning approach for identifying COVID-19 in chest CT scans using a framework combining an RNN, which is a recurrent neural network, and a network of convolution neurons. For this model training, we use COV19-CT-DB dataset, which is made up of chest CT scans. After preliminary processing, the dataset is split into training and testing groups. TensorFlow is used to build and construct the basic CNNRNN model architecture, that includes pre-trained ResNet50 for sequential analysis and feature extraction. This proposed model outperforms previous approaches, as evidenced by the experimental findings. Techniques for domain adaptation are also investigated to improve model generalization and resilience. Over all, the results highlight the possibility of using learning algorithms for the dataset as well as their contribution to bettering patient care and diagnostic accuracy in clinical settings that result in 98.8% of classification accuracy.