Integrating cybersecurity and explainable AI (XAI) for COPD detection and management

Explainable AI (XAI) is a method used to explain the working or decision-making mechanisms of artificial intelligence models. Given the current prevalence of AI, machine learning, and IoT devices, creating secure communication infrastructures becomes extremely important. However, classic rule-based and signature-based cybersecurity algorithms are struggling to cope with the ever-increasing complexity and convergence of modern technologies. As a result, there is an increasing demand for XAI to address these security concerns, ensuring well-structured explanations, interoperability, and decision-making accuracy, especially in domains like healthcare. In the healthcare domain, XAI techniques have been explored in limited applications, despite their potential to improve the detection and diagnosis of diseases like Chronic Obstructive Pulmonary Disease (COPD) by providing better transparency and accuracy compared to traditional methods. This paper addresses these challenges by applying XAI to cybersecurity and healthcare applications. First, we propose using XAI techniques to enhance the security of communication networks by providing detailed explanations, transparency, and improved interoperability in decision-making processes. Additionally, we demonstrate how the XAI-based LIME technique can be applied to diagnose COPD, providing more accurate diagnoses and precise explanations for each decision and classification compared to existing AI/ML methods. Finally, the paper presents a comprehensive performance analysis of the proposed XAI framework, comparing its effectiveness against traditional AI/ML models in both mitigating cybersecurity threats and improving COPD diagnosis. The results, measured using precision, accuracy, recall, F1 score, and other metrics, validate the superiority of XAI in both domains.