Privacy-preserving Data Mining, Sharing and Publishing

In the constantly developing world, e-health systems hold great promise for improving global access to healthcare services. Current significant technological visions of innovation in healthcare systems identify an approach to join different technological sectors and the need for technological platforms as well. These are: standardized electronic health records (eHR), aggregated public health data, genomic medicine, remote healthcare and diagnostics (telemedicine). Advancements enable medical consultation, remote imaging services, specialized medical diagnostics, and etc. There is an increasing demand for good health data management. According to [16] 75% of Americans would like to communicate via e-mail with their physicians and 60% would like to track their medical records electronically. A nationwide system of electronic medical records promises to facilitate the exchange of medical knowledge and patient data among physicians and other health providers. The question is how can healthcare institutions share patient information with a third party without compromising the privacy of individual patients? At the beginning we recall terms privacy, confidentiality, and information security [25,26]: 1. Privacy is the right of an individual to control disclosure of his or her medical information. 2. Confidentiality is the understanding that medical information will only be disclosed to authorised users at specific times of need. It entails holding sensitive data in a secure environment limited to an appropriate set of authorized individuals or organizations. 3. Information security includes the processes and mechanisms used to control the disclosure of information. It is the protection of computer-based information from unauthorized destruction, modification, or disclosure. The privacy and security aspects have an effect on the electronic storage and transmitting of patient health information, see Fig. 1. Vast quantities of data are generated through the health care process in medical institutions. We can distinguish different types of patient data: registration data (e.g. contact info), demographics (e.g., DOB, gender, race), billing information (e.g. diagnosis codes), genomic information (e.g. SNPs), medication and allergies, immunization status, laboratory test results, radiology images and so on. All kind of medical data connected with patient interacts in EMR System which consolidates particular systems, such as Registration System data (date and time of visit), Lab System, Pharmacy System, Radiology System (reports, images), Billing System (diagnosis codes), Order Entry System (prescriptions, orders), Decision Support System (clinical knowledge, guidelines). Physicians are the point of the transition/movement/usage of data. When they have access to all types of medical data related to the patients, can better diagnose and treat diseases with the help of Decision Support System. Medical information systems involve subsystems containing among others patient information, reporting tools, decision support systems and clinical scheduling.