Blockchain-Based COVID-19 Detection Framework Using Federated Deep Learning

ABSTRACT

With the increase of COVID-19 instances worldwide, a reliable method for diagnosing COVID-19 cases is needed. The major issue in detecting COVID-19 clients is a lack of diagnostic techniques that are both reliable and affordable. Due to the virus’s rapid dissemination, medical professionals are having difficulties finding positive cases. The second real-life issue is sharing data across clinics worldwide but keeping in mind the organizations’ privacy concerns. Developing a collaborative approach and protecting personal information are two important issues while creating a global classifier. This article offers a system that uses Ethereum - based federated learning to gather a modest quantity of data from many sources and train a global deep learning model. The data is authenticated using blockchain technology, and federated learning trained the system worldwide while maintaining the institution’s anonymity. The suggested structure may make use of current data to enhance diseases recognition. Our findings show that our method is more effective in detecting COVID-19 participants. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.