Enhancing Lung Diseases Prediction Using CNN ResNet-101 Model

This venture is targeted on growing an automatic gadget that predicts lung sicknesses from X-ray images using a deep mastering version. The version classifies pix into 4 classes: COVID-19, everyday, Pneumonia, and Tuberculosis. The system core is a Convolutional Neural community (CNN) the use of the ResNet-one hundred and one structure, pretrained at the ImageNet dataset and changed for lung disorder classification. X-ray photos are preprocessed and divided into training and validation units. The education procedure includes picture to enhance model effectiveness. The gadget become skilled using TensorFlow and Keras, using switch studying to shorten education time and decorate accuracy. The dataset is split into an 80-20 ratio for training and validation. version overall performance is assessed the use of accuracy, loss metrics, and a confusion matrix. Early stopping based totally on validation accuracy is implemented to beautify prediction reliability. The gadget generates a visual display of prediction probabilities for all ailment categories, along side a consumer-pleasant heatmap for studying confusion matrices. The very last model achieves aggressive accuracy and can be utilized as a diagnostic tool for healthcare specialists. A graphical interface has been created for users to add X-ray photos and consider prediction effects, showing the likelihood of each disorder. The mission showcases the effectiveness of deep getting to know in scientific imaging of lung sicknesses.