A Deep Learning Framework for Multi-Class Lung Disease Classification Using Chest X-Ray Images

Abstract

Lung diseases are among the leading causes of mortality worldwide, and early detection plays a crucial role in effective treatment and diagnosis. Chest X-ray (CXR) imaging is one of the most widely used and cost-effective diagnostic tools for detecting lung abnormalities. However, manual analysis of large volumes of medical images can be time-consuming and prone to human error. To address this challenge, this study proposes a deep learning-based architecture for multi-class lung disease classification using chest X-ray images. The dataset used in this study is collected from the Kaggle repository and includes labeled images belonging to different lung disease categories such as pneumonia, tuberculosis, COVID-19, and normal cases. The proposed framework begins with image preprocessing techniques including grayscale conversion, resizing, normalization, and noise filtering to improve image quality and standardize the dataset. Subsequently, Region of Interest (ROI) extraction is performed to isolate the lung region and eliminate irrelevant background information, thereby enhancing feature learning. The processed dataset is divided into training and testing subsets using a 70:30 ratio. Several deep learning models including VGG16, VGG19, InceptionV3, ResNet50, and DenseNet201 are implemented for feature extraction and classification of lung diseases. These convolutional neural network architectures learn hierarchical features from the chest X-ray images through convolutional layers, pooling layers, and fully connected layers. The performance of the proposed system is evaluated using various statistical metrics such as Accuracy, Precision, Sensitivity (Recall), F1-score, Kappa statistic, and Confusion Matrix. The experimental results demonstrate that deep learning models can effectively classify multiple lung diseases from chest X-ray images and assist in automated medical diagnosis. The proposed framework provides an efficient and reliable computer-aided diagnostic system that can support healthcare professionals in early detection and decision-making for lung disease diagnosis.