PREDICTION OF MULTILUNG DISEASE USING DEEPLEARNING

The prediction of multi-lung diseases using deep learning techniques from chest radiographs has emerged as a crucial task for aiding early diagnosis and improving healthcare outcomes. This study presents an innovative approach for the detection of six distinct lung conditions, namely, Normal, Pneumonia (Bacterial and Viral), COVID-19, Tuberculosis, and Emphysema, leveraging the ChestX6 dataset, a high-quality collection of labeled chest X-ray images. We explore the use of convolutional neural networks (CNN), MobileNet, and DenseNet to classify these diseases effectively. Each algorithm is evaluated on their ability to accurately classify chest radiographs into their respective categories, providing insights into their strengths and weaknesses. CNN, being a traditional and widely used deep learning model, serves as the baseline for comparison. MobileNet and DenseNet are utilized for their efficiency in handling complex image data while maintaining computational feasibility. The model's performance is assessed using various metrics, including accuracy, precision, recall, and F1-score, ensuring robust evaluation across multiple disease classes. The results show that deep learning-based models can significantly improve the accuracy of diagnosing multiple lung diseases, thus enabling faster and more reliable detection, which is critical in managing pandemics like COVID-19. This research underscores the potential of AI in transforming medical diagnostics and offers a pathway to automate and streamline the process of disease detection in chest radiographs.

Keywords: Multi-lung disease detection, Chest radiographs, Deep learning, Convolutional neural networks, MobileNet, DenseNet, Pneumonia, COVID-19, Tuberculosis, Emphysema.