A review of research on medical image confidentiality related technology coherent taxonomy, motivations, open challenges and recommendations

Abstract

With the modernization of biomedical image equipment and image processing technology, the amount of medical image data increases rapidly. These medical images need to be transmitted over a public network between doctors or hospitals to provide telemedicine services and sometimes need to be uploaded to a telemedicine data center or health cloud due to limited local computing and storage resources. Medical images involve confidential and sensitive information of patients, and the confidentiality technology to protect patient privacy is becoming more and more important. These problems have been studied by many scholars, but these studies are largely limited and scattered. To the authors’ knowledge, there is no systematic literature review related to medical image confidentiality technology. To provide valuable insights into the technical environment and to support researchers, the available options and gaps in this research area must be understood. Therefore, in this study, the recent 5 years’ research results in the field of medical image security are reviewed to map the field of study into a coherent taxonomy. We use five major academic databases which are (1) Medline, (2) IEEE Xplore, (3) WoS, (4) Scopus, and (5) ScienceDirect, and every literature related to (1) medical images, and (2) encryption were searched with a focus in these databases. These databases contain literature about medical image confidentiality technology. According to the classification scheme, the final data set consists of 123 pieces of literature, and are divided into five categories. The first category, which also occupies the most proportion, focuses on medical image confidentiality algorithms, and most of them adopt chaotic mapping correlative technology The second category considers multiple security requirements while studying medical image confidentiality. The third category combines medical image encryption and image compression technologies. The fourth category focuses on the hardware related medical image confidentiality design. The final category is the medical image security systems design. We then identify the basic characteristics of this emerging field from the following aspects; the motivations of using medical image confidentiality technology, the challenges of medical image confidentiality technology, and the recommendations for further research and use of medical image confidentiality technologies. Finally, the quality evaluation criteria of the medical image security algorithm are summarized, including but not limited to the evaluation of security and the evaluation of speed and time complexity. The results of the reference evaluation show that most studies on medical image security algorithms lack sufficient safety evaluation. This was evident from less than 30% of the total articles that used all kinds of security evaluation methods.