A cross domain access control model for medical consortium based on DBSCAN and penalty function

Area of Science:

• Health Informatics
• Cybersecurity
• Blockchain Technology

Background:

• Cross-domain access to patient medical information is crucial for healthcare but poses privacy risks.
• Existing access control methods often overlook security auditing of authorized user behavior.
• Sensitive patient data requires robust control mechanisms to prevent breaches.

Purpose of the Study:

• To design a blockchain-based system for intelligent recording and analysis of doctors' cross-domain access logs.
• To develop a method for detecting abnormal cross-domain access behavior using DBSCAN clustering.
• To implement a dynamic control mechanism using a penalty function to mitigate data breach risks.

Main Methods:

• Development of a blockchain-based system for secure logging of cross-domain access.
• Application of DBSCAN clustering algorithm for identifying anomalous access patterns in logs.
• Integration of a penalty function to dynamically regulate doctor access based on behavior analysis.

Main Results:

• The proposed system effectively records, queries, and analyzes cross-domain access logs.
• DBSCAN clustering successfully identified abnormal cross-domain access phenomena.
• The penalty function demonstrated a good control effect on doctor access behavior, reducing breach risks.

Conclusions:

• The blockchain-based access log auditing system offers a feasible solution for controlling cross-domain access in medical consortia.
• The integration of DBSCAN and a penalty function provides effective dynamic control over doctor access.
• This approach enhances the security of sensitive patient data in inter-institutional healthcare settings.