Enhancing lung abnormalities detection and classification using a Deep Convolutional Neural Network and GRU with explainable AI: A promising approach for accurate diagnosis

Abstract

Accurate and timely detection and classification of lung abnormalities are crucial for effective diagnosis and treatment planning. In recent years, Deep Learning (DL) techniques have shown remarkable performance in medical image analysis. This paper presents a novel and promising approach, namely DCNN-GRU, for improving the detection and classification of lung abnormalities. Our proposed model combines the capabilities of a Deep Convolutional Neural Network (DCNN) with a Gated Recurrent Unit (GRU) while incorporating Explainable AI techniques. Specifically, the DCNN-GRU model leverages the power of CNNs to automatically extract meaningful features from lung images, capturing both local and global patterns. The extracted features are fed into a GRU, which effectively models temporal dependencies and captures sequential information inherent in lung images. This integration allows the model to understand complex lung abnormalities accurately. Additionally, we emphasize the integration of Explainable Artificial Intelligence (XAI) techniques like LIME, SHAP, and Grad-CAM to enhance the interpretability and transparency of our model. To evaluate the proposed approach, we conducted experiments on COVID-19 and Lung cancer using two different datasets. The model achieved a promising accuracy of 99.30% and 98.97% for COVID-19, and lung cancer, respectively. Furthermore, the model significantly reduces training time compared to existing approaches. The results demonstrate that our model outperforms existing approaches, achieving a high accuracy rate in detection and classification tasks. Furthermore, the XAI provides valuable insights into the model’s decision-making process, aiding clinicians in understanding and validating the predictions.