The study deals with the development of a new computer system for the diagnosis of uterine cancer from CT images. Traditional approaches have difficulty effectively distinguishing between normal, benign, and malignant uterine tissue due to the complexity of the anatomical structures. The research team created a hybrid model that combines a DenseNet121 convolutional neural network with a transformer to improve image analysis. The proposed system was tested on a set of KAUH uterine cancer CT images with three categories: normal, benign, and malignant tissue. The results show a high efficiency of the model with an accuracy of 87.44%, a sensitivity of 87.13%, a specificity of 95.20% and an F1 score of 87.17%. The model outperforms benchmark systems such as VGG16, VGG19, MobileNetV2 and ResNet50. The authors conclude that this approach has the potential to serve as a computer-aided tool to assist radiologists in the detection of uterine cancer.