<![CDATA[Chest X-ray have an important function in the three areas of healthcare, namely diagnosis, treatment and re-examination. Studies from China suggest Chest Radiographs (X-Ray) and Chest Computed Comography (CT) scans can help diagnose COVID-19. Therefore, chest x-rays (x-rays) and chest computed tomography (CT) are appropriate methods for lung infections due to COVID-19. Based on this, the authors tried to make a model for the classification of digital images of Chest X-Ray results with the labels of Normal, Pneumonia, Tuberculosis (TBC), and COVID-19. Through the resulting model, the best model to use will be compared. The method used to create this model is through training and testing the dataset using the Convolutional Neural Network (CNN) architectural model, namely VGG19, ResNet50, and InceptionV3. The number of images used is 1,000 Chest X-Ray images. The dataset is divided into training and validation datasets in several ratios of 20% : 80%, 50%: 50%, and 80%: 20%. While testing uses 10% from train datset chest x-ray images as a confusion matrix dataset and 4 chest x-ray images as a prediction dataset. From the results of the research that has been done, the best model is VGG19 at 41 of 61 epoch and a ratio of 20% : 80%. Where the VGG19 model produces 94.44% for accuracy and 0,1084 loss value for training. Whereas at the testing stage with a configuration matrix, 95% accuracy value was obtained. Then for testing the new data prediction produces the best accuracy with 98.97% accuracy for the Normal label, 99.16% for COVID-19, 99.56% for Pneumonia, and 99.79% for Tuberculosis (TBC).]]>
