Preserving confidentiality when sharing medical database with the Cellsecu system

Area of Science:

• Computer Science
• Medical Informatics
• Data Security

Background:

• Medical databases contain sensitive patient information, requiring robust privacy protection.
• Removing explicit identifiers alone is insufficient to prevent re-identification through data linkage or unique characteristics.
• Existing methods struggle to balance data utility with comprehensive confidentiality.

Purpose of the Study:

• To propose Cellsecu, a novel computer system for maintaining anonymity and confidentiality in medical databases.
• To automatically enhance data privacy protection for seamless data warehouse querying.
• To introduce and implement a formal model and new confidentiality criteria for improved data security.

Main Methods:

• Developed Cellsecu, a system employing information removal, generalization, and expansion techniques.
• Utilized a formal model based on Modal logic as the theoretical foundation.
• Defined and implemented a new confidentiality criterion termed "non-uniqueness".

Main Results:

• Cellsecu effectively maintains anonymity and confidentiality of sensitive information within medical database cells.
• The system enhances data privacy protection, enabling automated query handling in data warehouses.
• Preliminary evaluations show minimal performance degradation from confidentiality modules.

Conclusions:

• Cellsecu offers a formal and effective approach to safeguarding sensitive medical data.
• The "non-uniqueness" criterion and Modal logic model provide a clear framework for data confidentiality.
• The system demonstrates a practical solution for enhancing privacy in medical data management with acceptable performance trade-offs.