On-the-fly DICOM-RTV metadata pseudonymization during a real-time streaming

With the current advancement of the information technologies and the growing demand for data sharing, the risk of private data leakages increases. Moreover, the aggregation of multiple data, required for improving quality, may favor the prediction of unrevealed private and sensitive information, in particular in the medical data. Thus, with the current laws and regulations (GDPR, CNIL, HIPAA, etc.), it becomes required to protect the privacy of the patients when dealing with their sensitive data, and in particular when sending them outside the clinical site. Inline with this objective, the data anonymization and pseudonymization emerged among the solutions for safely sharing private data with remote peers. In this work, a particular interest is given to the pseudonymization of the DICOM Real-Time Video (DICOM-RTV) associated metadata that can be generated inside the operating room (OR), and shared via a streaming technology. The pseudonymization, by definition, includes both the notion of de-identification and re-identification, necessary for ensuring patient safety. Thus, the main objective of this paper is to define and conceptualize an architecture for de-identifying DICOM-RTV metadata on-the-fly before streaming them together with the DICOM-RTV video of the ongoing surgery outside the medical facility in real-time.