Privacy-preserving models for comparing survival curves using the logrank test

Area of Science:

• Health Informatics
• Biostatistics
• Medical Research

Background:

• Electronic health records facilitate medical research collaboration.
• Privacy regulations and security concerns restrict data sharing and collaboration among institutions.
• Existing methods for survival analysis lack robust privacy preservation.

Purpose of the Study:

• To propose novel privacy-preserving models for survival curve comparison using the logrank test.
• To enable secure collaboration between multiple medical institutions for enhanced survival analysis.
• To address the challenges posed by privacy concerns in collaborative medical research.

Main Methods:

• Developed two distinct privacy-preserving models tailored for different collaboration scenarios.
• Adapted the logrank test for privacy-preserving survival curve comparisons.
• Utilized real-world medical data for empirical evaluation.

Main Results:

• The proposed models effectively preserve data privacy during survival curve comparisons.
• The models maintain the accuracy of the logrank test in collaborative settings.
• Experimental results demonstrate the practical effectiveness of the privacy-preserving approaches.

Conclusions:

• Privacy-preserving models can facilitate secure multi-institutional collaboration in medical research.
• The proposed logrank test-based models offer a viable solution for privacy-conscious survival analysis.
• These advancements can significantly boost collaborative medical research without compromising patient confidentiality.