<![CDATA[Artificial Intelligence can be used as a medical image diagnosis model by training a deep learning Convolutional Neural Networks (CNNs) image classification model to classify X-Ray images of the lungs. In this study, X-Ray images of the lungs classified as pneumonia or normal were used by the CNNs model that had been built. In the process of diagnosis there are difficulties, one of which is the lack of predictability of a follow up examination procedure only through X-Ray image. We need a method that can help diagnose X-Ray image objectively, quantitatively, and has high accuracy. There are used the CNNs model is then applied to Streamlit web-based application systems. The aim of this research is to develop a pneumonia disease classification application to simplify the diagnosis process. In this study, 5,855 X-Ray images of the lungs were used. The model is given epoch values of 10, 15 and 20 with 326 steps per epoch and the highest accuracy value is obtained at epoch 20. The result of accuracy value is 91.504%. The accuracy value can be affected from the epoch value, but the addition of the epoch value cannot fully increase the resulting accuracy value. The CNNs program model that has been created is then deployed to a web application using Streamlit and can be used for X-Ray image classification. The results of using this web application can be used for the initial detection process of X-Ray images with a diagnosis of pneumonia.]]>
Copyrights © 2023
