International Journal of Advanced Research in Computer and Communication Engineering

Abstract: Due to the COVID-19 outbreak, greater and more dependable tools were required to predict the severity of the disease and help specialists in their decision-making. Conventional approaches might be ineffective and time-consuming in the event of handling large and complex patient data. This paper presents a computer-aided diagnostic model, which is based on machine learning and the application of deep learning to predict the correct outcome. The model uses UNET to process medical images to segment them, and CNN/ResNet50 to classify the chest X-ray or CT scan. In order to enhance accuracy, a hybrid approach is formed by using optimized features with a Random Forest classifier. The system is programmed and developed to have a simple Graphical User Interface (GUI) that allows an individual to upload medical images and get an automated prediction with visualized output. Accuracy, precision, recall, and F1-score are used to compare performance, and the findings indicate that the proposed model is better than the current methods that can be used to offer an effective and convenient means of detecting and predicting the severity of COVID-19 at an early stage.

Keywords: COVID-19 Prediction; Machine learning; Deep learning; UNET; CNN; ResNet50; Random Forest; Image Segmentation; Medical Diagnosis; Computer-Aided Detection.

Downloads: | DOI: 10.17148/IJARCCE.2025.141166