An enhanced security framework for secured transfer of DICOM images by ensuring the CIA triad

An enhanced security framework for secured transfer of DICOM images by ensuring the CIA triad

연구 분야: Verification

학회: Multimedia Tools and Applications

With the advent of modern medicine, the speed of treatment is of paramount importance. Interconnection of medical equipment allows faster testing and reporting and allows patients and doctors to view reports from anywhere in the world. To facilitate this interconnection, the DICOM standard was proposed in 1985, to handle medical images in an efficient and effective way. However, there is now a need for these machines to be connected to the internet, and this has bought about many new threats to the security of medical data. In the last five years, a large number of cybersecurity incidents have occurred in the medical field targeting DICOM images. Since the DICOM standard does not enforce security, a framework which improves the confidentiality, authentication, and integrity for the header and pixel data during transmission, through cryptographic, watermarking and anonymization techniques is proposed in this literature. The cryptographic techniques use AES in Galois/Counter Mode (GCM), and the hashes for watermarking are generated using SHA. To protect the patients’ privacy and reliability in diagnosis and treatment, the security triad CIA has been ensured in the proposed work through anonymization, hashing, encryption, timely access, and backup. It is observed that the time difference between SHA3 and SHA2 hashing are almost negligible (< 2% in worst case), when smaller workloads are performed with the modern architectures. And it is always better to prefer SHA3 when hardware acceleration is available (< 0.4% in worst case) for processing. The anonymization techniques allow DICOM image metadata to be stripped from the files in case they are to be published publicly. Overall, the proposed framework offers a robust solution to mitigate cybersecurity risks associated with DICOM image transmission, ensuring the confidentiality, authentication, and integrity of medical data in transit. By comparing with the other existing methods, it is noted that the proposed method achieves a structured similarity index of 0.9999 and the peak-signal-to-noise ratio of 79.58.