Authors: Chathurika Ranasinghe, Dineth Weerakoon, Malsha Bandara, Thivanka Gunawardana
Abstract: Biomedical imaging systems generate large volumes of sensitive data that must be securely stored, reliably retrieved, and retained for long durations to meet regulatory, clinical, and research demands. ZFS, a high-integrity, copy-on-write file system with integrated volume management, has emerged as a viable solution for long-term imaging storage in healthcare and biomedical research institutions. This review explores the suitability of ZFS for managing medical imaging archives highlighting its built-in features such as end-to-end checksumming, atomic snapshots, native encryption, and tiered storage capabilities. The paper examines ZFS's alignment with regulatory requirements like HIPAA, GDPR, and FDA 21 CFR Part 11, and discusses how its auditability, snapshot lifecycle management, and disaster recovery features help ensure compliance and data integrity. We delve into ZFS performance tuning for imaging workloads, including optimizations using ARC, L2ARC, SLOG, and record size configuration, which are critical for high-throughput radiology and pathology systems. Integration with PACS, RIS, and AI processing pipelines is reviewed, along with real-world deployments in clinical and research environments. Operational challenges such as resource overhead, secure deletion limitations, and administrative complexity are addressed, alongside emerging trends like object storage extensions, support for storage-class memory, and container-native workflows. Through this comprehensive review, ZFS is positioned not only as a technically robust and scalable imaging storage platform, but also as a strategic foundation for future-proof, compliant biomedical data management.
DOI: https://doi.org/10.5281/zenodo.15847617
