Whole slide images are among the largest files in healthcare, and storage is often the cost that grows fastest after a lab goes digital. This guide is for lab managers, hospital IT teams and finance staff who need to estimate whole slide image storage, choose the right tiers and set policies that keep the archive useful without letting costs run away.
Estimating whole slide image storage capacity
Start with a simple model built from your own numbers. You need four inputs:
- Slides scanned per day, including peaks and any backlog you plan to digitize.
- Average file size per slide, measured from your own pilot scans rather than vendor estimates.
- Working days per year on which scanning happens.
- Retention period for each category of image.
Multiply slides per day by average file size to get daily growth, then by working days to get annual growth. As a purely illustrative example, a lab scanning 500 slides a day at an average of 1 GB per slide would add about half a terabyte per working day. Your figures will differ, which is exactly why you should measure them.
Then add headroom. Backups, replicas and disaster recovery copies multiply the raw figure, and growth in volume or a switch to higher magnification can change the curve quickly.
What drives file size
Average file size is the most sensitive input in your model, and it varies more than many people expect. The main factors are:
- Tissue area. A large resection section produces a far bigger file than a small needle biopsy.
- Scan resolution. Scanning at 40x produces files several times larger than 20x for the same tissue.
- Compression settings. Higher compression saves space but must not degrade diagnostic quality. Settings should be agreed with pathologists and covered by validation.
- Multi-plane scanning. Z-stacks for cytology multiply file size by the number of planes captured.
- Empty glass. Good tissue detection avoids scanning blank areas, while poor detection wastes space.
- Rescans. Failed scans that are kept rather than deleted add up quietly.
If you are still comparing equipment, ask each vendor for measured file sizes on your own slides as part of choosing a whole slide scanner. For background on how pyramids and compression work, see how whole slide imaging works.
Storage tiers: hot, warm and archive
Not every image needs the same speed of access. Most labs benefit from tiered storage that matches cost to how often images are used.
Hot tier
Fast storage for active cases: images scanned recently and awaiting sign-out, plus cases likely to be reopened soon for addenda, tumor boards or consultations. Viewer performance depends heavily on this tier, so do not economize here.
Warm tier
Lower-cost storage for images from recent months that are opened occasionally, such as prior cases a pathologist wants to compare with a new specimen. Retrieval should still take seconds, not hours.
Archive tier
The cheapest storage for long-term retention. Retrieval may be slower, which is acceptable for images opened rarely. Make sure the delay is understood and that urgent retrieval is still possible when a patient's history matters.
Automated policies should move images between tiers based on age and case status, while keeping the link to the case record intact wherever the file physically lives.
Retention: what to keep and for how long
Retention policy has a bigger effect on long-term cost than any hardware choice. Decide it deliberately with pathologists, quality managers and legal advisers, and align it with local regulations and professional guidance.
Questions to settle:
- Is the glass slide or the digital image the record of retention, or both?
- Do all case types need the same retention period, or can some be shorter?
- Should every slide be kept, or only those reviewed for diagnosis?
- What happens to rescans, test scans and failed scans?
- How will images used for teaching or research be handled separately?
- Who approves deletion, and how is it recorded?
Many labs continue to retain glass slides and blocks under their existing rules, which can allow a more flexible policy for digital images. Whatever you decide, write it down and apply it consistently.
On-premises, cloud or hybrid
Each model has real trade-offs.
- On-premises storage gives direct control and predictable performance on the local network, but requires capital investment, data center space, staff and hardware refresh cycles.
- Cloud storage scales without upfront hardware and offers inexpensive archive tiers, but costs are ongoing, data retrieval may be charged, and performance depends on internet connectivity.
- Hybrid setups keep recent images close to scanners and pathologists while moving older images to the cloud or a remote archive.
For cloud or hybrid models, check where data is physically stored, how it is encrypted, who can access it, and what it costs to retrieve or move the whole archive. Patient data protection rules apply wherever the images live, and slide label images may contain identifiers.
Protecting the archive and controlling cost
Protecting the archive
Storage is not only about capacity. An archive that cannot survive a hardware failure or ransomware attack is not fit for clinical use.
- Keep at least one copy of images in a separate location from the primary storage.
- Protect backups from being altered or deleted by the same credentials that manage production systems.
- Test restores regularly, including restoring a full day of scans, not just a single file.
- Verify file integrity periodically so silent corruption is caught early.
- Document recovery time objectives that pathologists and management have agreed to.
Format matters for long-term access as well. Images stored in open or well-documented formats are easier to migrate later, as discussed in our article on WSI file formats and DICOM.
Controlling cost without hurting diagnosis
Common mistakes that inflate storage costs include keeping every rescan, scanning all slides at 40x by default, ignoring tissue detection settings, and storing everything on the fastest tier indefinitely. Each is easy to correct once spotted.
Review storage growth monthly against your model. Unexpected jumps usually point to a workflow change, such as a new case type, a scanner setting changed during maintenance, or a backlog digitization project nobody accounted for.
Key takeaways
- Build your capacity model from measured file sizes and real slide volumes, then add headroom for copies and growth.
- Resolution, tissue area, compression and multi-plane scanning drive file size more than anything else.
- Use hot, warm and archive tiers, with automated movement that keeps images linked to their cases.
- Retention policy is the biggest long-term cost lever, so decide it deliberately and document it.
- Protect the archive with separate backups, integrity checks and tested restores.
If you are weighing cloud options for your digital workflow, see plans for how PATHOLOGY+ combines whole slide image viewing and case tracking on a multi-tenant cloud platform.