Privacy Protection Technique for Medical Database

With the development of computer science, databases have huge benefits in data collection, accumulation, aggregate, statistics, analysis and administration, academic research and other applications by analyzing information provided by the implicit message. Databases can assist the relevant project to develop or policy-making. When the data can be collected more easily, the user may infer additional information and some particular individual's privacy has been damaged. Especially, the medical information database contains most of the content is extremely patient privacy and unwilling to be leaked insurance and medical treatment records. However, that indiscriminate hide may result in leakage of privacy information will be difficult to carry out relevant research. Therefore, we have a need for the medical database to explore the privacy protection technology. Without prejudice to the existing data processing procedures, we design an anonymizing classification system for medical database takes into account both quality and patient privacy. This research is not only a case for the hospital medical database, analysis of field characteristics, patient privacy and the relevance of medical and health research but the results by analogy to other hospital medical database. In this two year project, “Privacy protection technique for medical database”, it contains the following two parts: The first year, we will explore the field characteristics of the medical database and data classification, including evaluate the field characteristics and data classification of a medical database on quality and security level; evaluate indicators based on different dataset, the classification of each and design the fast algorithm. The second year, we will design anonymizing classification system which combines the patient privacy and the quality of information for medical database, including explore the demand for anonymizing classification; discuss different fields can affect the level of security, information quality and system efficiency in the specific bin size; implement the different needs of the anonymizing method of classification and to compare with the other anonymizing ones by using different subsets of database.

Project ID:PF10001-0902
External Project ID:NSC99-2410-H182-017-MY2