Progressive privacy-preserving batch retrieval of lung CT image sequences based on edge-cloud collaborative...

Area of Science:

• Medical Imaging
• Computer Science

Background:

• Computer tomography image (CI) sequences are large time-series datasets.
• High computational costs for similarity retrieval in large CI sequences strain cloud resources.

Purpose of the Study:

• To propose a privacy-preserving batch retrieval scheme for lung CI sequences.
• To address computational and I/O challenges in cloud-based retrieval.

Main Methods:

• Developed a progressive privacy-preserving Batch Retrieval (BRS) scheme.
• Utilized edge-cloud collaborative computation with batch similarity measure, CIB-based privacy, uniform indexing, and edge buffering.

Main Results:

• The BRS method significantly reduces response time and network communication costs.
• Demonstrated enhanced retrieval safety and accuracy compared to state-of-the-art methods.
• Showcased improved efficiency and scalability.

Conclusions:

• The BRS method offers an efficient and scalable solution for privacy-preserving retrieval of lung CI sequences.
• Edge-cloud collaboration effectively mitigates computational burdens and enhances performance.