Unit 2 · From Glass to Gigabytes: Foundations of Digital Pathology and WSI Systems
Welcome to Unit 2 · From Glass to Gigabytes.
This study guide is about how a modern digital pathology service actually works from a glass slide on the bench to pixels on a screen, files on a server, and cases flowing through scanners, viewers, and networks.
The focus is on building clear mental models of the full digital workflow, not on coding or advanced image processing. You start from familiar territory (routine histology, glass slides, day-to-day reporting) and gradually add the digital components that make up a real service: scanners, pixels, file formats, storage, viewers, validation, and a realistic rollout plan for a lab. Each chapter is an evidence-backed “essential knowledge” pack that distills the literature and turns it into practical, clinically anchored takeaways.
- Anchor every new concept to concrete use cases such as tumour boards, frozen sections, and subspecialty sign out so that the technology conversations remain rooted in patient-facing outcomes.
- Build a shared vocabulary that lets you brief biomedical scientists, information technology, and leadership without needing to master low-level configuration details.
- Capture what matters for patient safety—image fidelity, turnaround time, remote access, and QA—and translate those into diagrams and roadmaps you can reuse in committees or accreditation visits.
- Break the “glass to gigabytes” path into scanners, file formats, storage tiers, viewers, and validation steps so you can see where data, bandwidth, or integration bottlenecks appear.
- Understand the tradeoffs between compression, resolution, and throughput, and how those choices ripple through procurement, networking, and security discussions.
- Turn stakeholder requirements into realistic rollout phases with the dependencies (barcoding, LIS/EHR hooks, monitoring, and QA) made explicit before you discuss hardware specs.
Pitfall
Many labs stop at “we bought a scanner and used it for tumor boards or a few consult cases” but never manage to digitize routine sign out. Surveys from Europe, Asia, and Jordan show that only a minority of labs with scanners actually use digital pathology for day-to-day diagnosis, and most use it mainly for consultation, education, or research .
Interesting fact
A single whole slide image is often a gigapixel image (around 1,000,000,000 pixels), which is over a hundred times more pixels than a 4K ultra-high-definition frame (about 8.3 million pixels) . This is why tiling, pyramids, and storage strategy are essential topics rather than IT trivia.
Chapters at a glance
Each chapter is a concise reading pack that focuses on a specific part of the system and keeps linking back to clinical questions and real-world workflow.
Chapter 1 distils the literature, evidence base, and clinical drivers you need before diving into the technical pieces.
- Chapter 1 – Digital pathology overview and evidence → Read it now
Chapter 1: Digital pathology overview and evidence
- Core definitions of digital pathology, WSI, and vendor-neutral archives plus why adoption is happening now.
- Clinical drivers and use cases (teleconsults, MDTs, remote reporting, education, QA) grounded in evidence.
- Safety/noninferiority data, local validation caveats, and change-management steps for rollouts.
- Where AI and computational pathology fit once a digital platform is in place.
Chapter 2: Pixels, colour, and compression
- Microns per pixel as the key specification behind any “20x/40x” label and its clinical impact.
- Colour channels, colour spaces, and calibration for stains that need faithful rendering.
- Sampling and resolution tradeoffs, including when finer sampling matters.
- Bit depth and dynamic range pragmatics plus lossy versus lossless compression effects.
Chapter 3: Workflow and policy basics
- Glass-only versus glass-plus-digital day-in-the-life for a case.
- Pre-analytic, analytic, and post-analytic framing to anchor QA discussions.
- Roles across histology, scanner operators, pathologists, and IT with RACI-style thinking.
- Scanning policies by case type and stain, plus common failure points and safety nets.
Chapter 4: Scanners and image acquisition
- Scanner anatomy and how the acquisition sequence really works.
- Specifications that matter in practice and how to read them.
- Typical failure modes and the symptoms you see on screen.
- Placement, throughput classes, and QA/maintenance habits for reliable scanners.
Chapter 5: File formats, tiling, and storage
- Why multiple WSI formats exist and how proprietary and open options compare.
- DICOM basics for digital pathology and interoperability.
- Tiling and pyramids as the backbone of viewer performance.
- Storage planning, archiving tiers, and ways to reduce vendor lock-in risk.
Chapter 6: Viewers, workstations, and performance
- Viewer features pathologists rely on and how responsiveness shapes adoption.
- Displays and ergonomics as the new “microscope” environment.
- Network performance and remote reporting considerations with practical troubleshooting.
Chapter 7: Validation and quality assurance
- Why validation is mandatory and what counts as a “system” requiring re-validation.
- Designing concordance studies with scope, washout, and challenging cases.
- Ongoing QA/monitoring after go-live plus governance and documentation that keep the service safe.
Portfolio and next steps
The diagrams, workflow maps, and roadmap sketches you create in Unit 2 can be turned into tangible portfolio artifacts:
- A high-level “glass to gigabytes” diagram for your lab or an imagined lab.
- A short narrative of a proposed digital workflow for one subspecialty.
- A simple risk and dependency map for a phased rollout.
These can later connect to QuPath projects in DCP 602 and to more advanced data and deployment plans in later informatics/ML modules.