Detection of Covid‐19 disease by using privacy‐aware artificial intelligence system

Abstract

Abstract Covid‐19 is a highly infectious respiratory disease that spreads quickly between individuals and has been recognized as a pandemic by the World Health Organization (WHO). Chest x‐ray images, lung computed tomography images, and polymerase chain reaction tests are generally used to diagnose this disease by the doctors. Nevertheless, manual diagnosis of Covid‐19 disease is laborious and requires highly experienced professionals. Therefore, automated systems are always needed to assist doctors in their diagnostic decisions. In the field of medicine and healthcare, artificial intelligence and deep learning currently offer incredibly effective and rapid automatic decision‐support systems. Since sensitive data is used to diagnose Covid‐19, privacy has become a major concern in research that uses artificial intelligence and deep learning. In order to eliminate these issues, this paper proposes a novel deep learning model that privately detects Covid‐19 disease. The proposed model utilizes differential privacy technique to provide data privacy and convolutional neural network to diagnose Covid‐19 disease. The performance of the proposed model was evaluated through experiments conducted on five different datasets, resulting a maximum accuracy rate of 97%.

Ähnliche Arbeiten