Chapter 4: Scanners and image acquisition

This chapter is about how glass slides are turned into reliable whole slide images (WSI) inside a scanner. The goal of the reading pack is not to make you an engineer, but to give you enough mental models to judge scanners, spot problems, and talk confidently to vendors and IT.

4.1 Why scanners matter for clinical practice

What you need to know

Whole slide imaging (WSI) is not “nice to have hardware”; it is the core engine that makes digital pathology possible for primary diagnosis, telepathology, teaching, and computational pathology.
A scanner replaces the optical path of the microscope with a camera plus motion control. If the scanner does a poor job (focus, color, coverage), every downstream step suffers, no matter how good your viewer or AI is.
WSI scanners acquire the entire tissue area at high resolution so that you can safely sign out from the screen without needing glass slides for most cases.
Key clinical reasons to care about scanner quality:
- Safe primary diagnosis (you need to see the same details you would on the microscope).
- Rapid, predictable turnaround time (no bottlenecks in high‑volume labs).
- Reliable teleconsults and tumor boards (no surprises when cases are reviewed remotely).
- Foundation for quantitative image analysis and AI (consistent image quality and metadata).
The same scanner decision will affect many use cases: frozen sections, routine surgical pathology, cytology, research slides, and education. Treat scanner selection as a long‑term infrastructure choice, not a gadget purchase.
When you evaluate a scanner, think about use cases (what you will actually do with it) more than headline technical specs.

Reference
Zarella MD, Bowman D, Aeffner F, et al. A practical guide to whole slide imaging: A white paper from the Digital Pathology Association. Arch Pathol Lab Med. 2019;143(2):222–234. doi:10.5858/arpa.2018-0343-RA. Available at: https://doi.org/10.5858/arpa.2018-0343-RA

4.2 Scanner anatomy and how image acquisition really works

What you need to know

A modern WSI scanner is essentially a motorized microscope plus a high‑resolution camera, wrapped in automation and safety checks.
Main components worth knowing by name:
- Illumination system: light source (often LED) plus optics that need to be stable in intensity and color over time.
- Objectives: usually 20× and/or 40×, with specific numerical aperture (NA) that sets the resolving power; some systems use zoom optics instead of discrete objectives.
- XY stage: moves the slide so the camera can visit each tile position; accuracy and repeatability are key for stitching.
- Focus system (Z‑axis): can be hardware autofocus (separate focus camera), software autofocus (analysis of the main camera image), or hybrid strategies.
- Camera / sensor: determines the raw pixel size, dynamic range, and readout speed; line‑scan and area‑scan architectures exist.
- Barcode / slide ID system: glues the physical slide to its digital identity in the LIS and image manager.
Scanners usually acquire images as many small rectangles (“tiles”) that are stitched into a seamless multi‑resolution pyramid.
Different vendors use different scanning strategies:
- Tile‑based vs line‑scan.
- Fixed focus vs per‑tile focus vs continuous focus tracking.
- Single focal plane vs z‑stacks.
Understanding this anatomy helps you interpret vendor marketing claims and match them to what actually matters (e.g., reliable focus at tissue edges, not just “X megapixels”).

Reference
Patel A, Balis UGJ, Cheng J, et al. Contemporary whole slide imaging devices and their applications within the modern pathology department: A selected hardware review. J Pathol Inform. 2021;12:50. doi:10.4103/jpi.jpi_66_21. Available at: https://doi.org/10.4103/jpi.jpi_66_21

4.3 Scanner specifications that really matter (and how to read them)

What you need to know

Scanner brochures list many numbers; focus on the ones that affect diagnostic safety and workflow:
- Microns per pixel (mpp): links digital resolution to tissue; typical values are around 0.25 µm/px (40×) or 0.5 µm/px (20×).
- Objective and NA: higher NA improves fine detail but can shrink depth of field and increase sensitivity to coverslip thickness and tissue thickness.
- Throughput: slides per hour or per day; beware that headline numbers often assume “ideal” slides and limited tissue area.
- Duty cycle and reliability: how long the scanner can run at or near capacity without overheating or frequent jams.
- Color management: how stable and reproducible the color is across time, scanners, and stains.
- Regulatory status: CE‑IVD, FDA clearance, or use as “research only,” which affects what you can legally do with the system.
Microns per pixel matters clinically because it determines what details you can comfortably see:
- At ~0.25 µm/px, nuclear detail and mitoses are usually reliable for most H&E work.
- At ~0.5 µm/px, many diagnoses are still safe, but borderline nuclear details are harder.
Depth of field interacts with tissue thickness and coverslipping quality; scanners with very shallow depth of field can struggle with uneven slides.
“Scan speed” numbers are often quoted for very small tissue areas; always ask vendors for realistic benchmarks that match your own case mix.
For practical scanner comparison, it is often more useful to ask:
- How many slides can you reliably scan per day at our target resolution?
- How often do rescans occur because of focus or coverage errors?
- How many staff hours per day are spent loading and troubleshooting?

4.4 Common scanner failure modes and what they look like on screen

What you need to know

Most scanner problems appear to the pathologist as image quality issues rather than obvious hardware errors. You should recognize these patterns:
- Focus problems: whole slide soft, peripheral softening, or alternating in‑focus and out‑of‑focus tiles.
- Coverage problems: missing tissue at slide edges, skipped levels, truncated biopsies.
- Stitching and registration artifacts: visible grid patterns, misaligned structures across tile boundaries, or “ghosting” in line‑scan systems.
- Color and illumination drift: slides from different days look noticeably different in background color and stain saturation.
- Motion artifacts: duplicated nuclei, streaking, or directional blur from stage motion.
Many issues originate from the slides themselves (thick sections, folds, poor coverslips, excessive mounting medium) but are amplified by scanner choices such as focus strategy and depth of field.
Your validation and QA plan should deliberately include “difficult” slides (thick, folded, necrotic, cytology) so you see how the scanner behaves at the edges of real‑world practice.
It is important to link common image artifacts back to their root causes so that you can act on them:
- Change local tissue processing or coverslipping practices.
- Adjust scanner maintenance and cleaning schedules.
- Update scanning protocols (e.g., use z‑stacks for certain slide types).
A good mental model: the scanner is part of your pre‑analytic chain. Treat recurrent scanner artifacts in the same way you would treat recurrent staining issues in histology: with root‑cause analysis and documented corrective actions.

Reference
Jahn SW, Plass M, Moinfar F. Digital pathology: Advantages, limitations and emerging perspectives. J Clin Med. 2020;9(11):3697. doi:10.3390/jcm9113697. Available at: https://doi.org/10.3390/jcm9113697

4.5 Where scanners live, how they fit into the lab, and how to choose a scanner class

What you need to know

Scanner performance in the real world depends heavily on where the scanner lives and how it is integrated into the lab:
- Proximity to histology and coverslipping (minimizing slide transport and handling).
- Environmental stability (temperature, dust, vibration, access for service).
- Network connectivity and bandwidth to your image server and viewer.
There is no single “best scanner”; instead, think in terms of scanner classes:
- High‑throughput central scanners (hundreds of slides per day, often with large autoloaders).
- Mid‑range scanners for subspecialty labs or research cores.
- Small desktop or on‑demand scanners for frozen sections, small hospitals, or backup.
The “right” scanner class depends on:
- Your case volume and daily peaks.
- How many sites you support (main campus, satellites, home reporting).
- Your current and planned clinical use cases (primary diagnosis vs consults vs education vs AI).
Successful digital transformation depends on more than hardware:
- Governance and project management.
- Clear clinical champions.
- Training for pathologists and technologists.
- Investment in image management, storage, and support.
When choosing scanners, combine a technical evaluation with site visits or detailed case studies from similar institutions; pay attention to how they solved mundane issues such as night‑time loading, rescans, and support.

Reference
Hanna MG, Ardon O, Reuter VE, et al. Integrating digital pathology into clinical practice. Mod Pathol. 2022;35(2):152–164. doi:10.1038/s41379-021-00929-0. Available at: https://doi.org/10.1038/s41379-021-00929-0

4.6 Scanner QA, maintenance, and long‑term reliability

What you need to know

Scanners need a structured quality management approach, similar to any diagnostic instrument:
- Defined roles for daily checks, periodic maintenance, and escalation.
- Documented cleaning routines for optics, stages, and barcode readers.
- Logs for downtime, rescans, and service visits.
Validation is not “one and done”:
- After initial validation, you need ongoing monitoring to detect drift in focus, color, and coverage.
- Major changes such as firmware updates, new objectives, or new slide types usually require at least a small re‑validation.
Many guidelines emphasize using test slides and reference cases:
- Keep a panel of cases that are known to stress the system (e.g., thin biopsies, decalcified bone, cytology).
- Use them when scanners are serviced, moved, or updated.
QA for scanners should be linked to broader digital pathology quality frameworks, covering:
- Display performance and workstations.
- Network and storage reliability.
- Viewer behavior and measurement tools.
Written procedures and training for technologists are essential; most real‑world scanner problems are prevented or caught by well‑trained staff rather than by automation alone.

Reference
Cross S, Furness P, Igali L, Snead D, Treanor D. Best practice recommendations for implementing digital pathology. The Royal College of Pathologists; 2018. Available at: https://www.rcpath.org/static/f465d1b3-797b-4297-b7fedc00b4d77e51/Best-practice-recommendations-for-implementing-digital-pathology.pdf

--- title: "Chapter 4: Scanners and image acquisition" format: html: toc: true toc-depth: 3 code-fold: false df-print: paged smooth-scroll: true execute: echo: true warning: false message: false --- # Chapter 4: Scanners and image acquisition This chapter is about how glass slides are turned into reliable whole slide images (WSI) inside a scanner. The goal of the reading pack is not to make you an engineer, but to give you enough mental models to judge scanners, spot problems, and talk confidently to vendors and IT. --- ## 4.1 Why scanners matter for clinical practice **What you need to know** - Whole slide imaging (WSI) is not “nice to have hardware”; it is the core engine that makes digital pathology possible for primary diagnosis, telepathology, teaching, and computational pathology. - A scanner replaces the optical path of the microscope with a camera plus motion control. If the scanner does a poor job (focus, color, coverage), every downstream step suffers, no matter how good your viewer or AI is. - WSI scanners acquire the *entire* tissue area at high resolution so that you can safely sign out from the screen without needing glass slides for most cases. - Key clinical reasons to care about scanner quality: - Safe primary diagnosis (you need to see the same details you would on the microscope). - Rapid, predictable turnaround time (no bottlenecks in high‑volume labs). - Reliable teleconsults and tumor boards (no surprises when cases are reviewed remotely). - Foundation for quantitative image analysis and AI (consistent image quality and metadata). - The same scanner decision will affect many use cases: frozen sections, routine surgical pathology, cytology, research slides, and education. Treat scanner selection as a long‑term infrastructure choice, not a gadget purchase. - When you evaluate a scanner, think about *use cases* (what you will actually do with it) more than headline technical specs. **Reference** Zarella MD, Bowman D, Aeffner F, et al. A practical guide to whole slide imaging: A white paper from the Digital Pathology Association. Arch Pathol Lab Med. 2019;143(2):222–234. doi:10.5858/arpa.2018-0343-RA. Available at: https://doi.org/10.5858/arpa.2018-0343-RA --- ## 4.2 Scanner anatomy and how image acquisition really works **What you need to know** - A modern WSI scanner is essentially a motorized microscope plus a high‑resolution camera, wrapped in automation and safety checks. - Main components worth knowing by name: - **Illumination system**: light source (often LED) plus optics that need to be stable in intensity and color over time. - **Objectives**: usually 20× and/or 40×, with specific numerical aperture (NA) that sets the resolving power; some systems use zoom optics instead of discrete objectives. - **XY stage**: moves the slide so the camera can visit each tile position; accuracy and repeatability are key for stitching. - **Focus system (Z‑axis)**: can be hardware autofocus (separate focus camera), software autofocus (analysis of the main camera image), or hybrid strategies. - **Camera / sensor**: determines the raw pixel size, dynamic range, and readout speed; line‑scan and area‑scan architectures exist. - **Barcode / slide ID system**: glues the physical slide to its digital identity in the LIS and image manager. - Scanners usually acquire images as many small rectangles (“tiles”) that are stitched into a seamless multi‑resolution pyramid. - Different vendors use different scanning strategies: - Tile‑based vs line‑scan. - Fixed focus vs per‑tile focus vs continuous focus tracking. - Single focal plane vs z‑stacks. - Understanding this anatomy helps you interpret vendor marketing claims and match them to what actually matters (e.g., reliable focus at tissue edges, not just “X megapixels”). **Reference** Patel A, Balis UGJ, Cheng J, et al. Contemporary whole slide imaging devices and their applications within the modern pathology department: A selected hardware review. J Pathol Inform. 2021;12:50. doi:10.4103/jpi.jpi_66_21. Available at: https://doi.org/10.4103/jpi.jpi_66_21 --- ## 4.3 Scanner specifications that really matter (and how to read them) **What you need to know** - Scanner brochures list many numbers; focus on the ones that affect diagnostic safety and workflow: - **Microns per pixel (mpp)**: links digital resolution to tissue; typical values are around 0.25 µm/px (40×) or 0.5 µm/px (20×). - **Objective and NA**: higher NA improves fine detail but can shrink depth of field and increase sensitivity to coverslip thickness and tissue thickness. - **Throughput**: slides per hour or per day; beware that headline numbers often assume “ideal” slides and limited tissue area. - **Duty cycle and reliability**: how long the scanner can run at or near capacity without overheating or frequent jams. - **Color management**: how stable and reproducible the color is across time, scanners, and stains. - **Regulatory status**: CE‑IVD, FDA clearance, or use as “research only,” which affects what you can legally do with the system. - Microns per pixel matters clinically because it determines what details you can comfortably see: - At ~0.25 µm/px, nuclear detail and mitoses are usually reliable for most H&E work. - At ~0.5 µm/px, many diagnoses are still safe, but borderline nuclear details are harder. - Depth of field interacts with tissue thickness and coverslipping quality; scanners with very shallow depth of field can struggle with uneven slides. - “Scan speed” numbers are often quoted for very small tissue areas; always ask vendors for realistic benchmarks that match your own case mix. - For practical scanner comparison, it is often more useful to ask: - How many slides can you reliably scan per day at our target resolution? - How often do rescans occur because of focus or coverage errors? - How many staff hours per day are spent loading and troubleshooting? **Reference** Zarella MD, Bowman D, Aeffner F, et al. A practical guide to whole slide imaging: A white paper from the Digital Pathology Association. Arch Pathol Lab Med. 2019;143(2):222–234. doi:10.5858/arpa.2018-0343-RA. Available at: https://doi.org/10.5858/arpa.2018-0343-RA --- ## 4.4 Common scanner failure modes and what they look like on screen **What you need to know** - Most scanner problems appear to the pathologist as image quality issues rather than obvious hardware errors. You should recognize these patterns: - **Focus problems**: whole slide soft, peripheral softening, or alternating in‑focus and out‑of‑focus tiles. - **Coverage problems**: missing tissue at slide edges, skipped levels, truncated biopsies. - **Stitching and registration artifacts**: visible grid patterns, misaligned structures across tile boundaries, or “ghosting” in line‑scan systems. - **Color and illumination drift**: slides from different days look noticeably different in background color and stain saturation. - **Motion artifacts**: duplicated nuclei, streaking, or directional blur from stage motion. - Many issues originate from the slides themselves (thick sections, folds, poor coverslips, excessive mounting medium) but are amplified by scanner choices such as focus strategy and depth of field. - Your validation and QA plan should deliberately include “difficult” slides (thick, folded, necrotic, cytology) so you see how the scanner behaves at the edges of real‑world practice. - It is important to link common image artifacts back to their root causes so that you can act on them: - Change local tissue processing or coverslipping practices. - Adjust scanner maintenance and cleaning schedules. - Update scanning protocols (e.g., use z‑stacks for certain slide types). - A good mental model: the scanner is part of your pre‑analytic chain. Treat recurrent scanner artifacts in the same way you would treat recurrent staining issues in histology: with root‑cause analysis and documented corrective actions. **Reference** Jahn SW, Plass M, Moinfar F. Digital pathology: Advantages, limitations and emerging perspectives. J Clin Med. 2020;9(11):3697. doi:10.3390/jcm9113697. Available at: https://doi.org/10.3390/jcm9113697 --- ## 4.5 Where scanners live, how they fit into the lab, and how to choose a scanner class **What you need to know** - Scanner performance in the real world depends heavily on where the scanner lives and how it is integrated into the lab: - Proximity to histology and coverslipping (minimizing slide transport and handling). - Environmental stability (temperature, dust, vibration, access for service). - Network connectivity and bandwidth to your image server and viewer. - There is no single “best scanner”; instead, think in terms of **scanner classes**: - High‑throughput central scanners (hundreds of slides per day, often with large autoloaders). - Mid‑range scanners for subspecialty labs or research cores. - Small desktop or on‑demand scanners for frozen sections, small hospitals, or backup. - The “right” scanner class depends on: - Your case volume and daily peaks. - How many sites you support (main campus, satellites, home reporting). - Your current and planned clinical use cases (primary diagnosis vs consults vs education vs AI). - Successful digital transformation depends on more than hardware: - Governance and project management. - Clear clinical champions. - Training for pathologists and technologists. - Investment in image management, storage, and support. - When choosing scanners, combine a technical evaluation with site visits or detailed case studies from similar institutions; pay attention to how they solved mundane issues such as night‑time loading, rescans, and support. **Reference** Hanna MG, Ardon O, Reuter VE, et al. Integrating digital pathology into clinical practice. Mod Pathol. 2022;35(2):152–164. doi:10.1038/s41379-021-00929-0. Available at: https://doi.org/10.1038/s41379-021-00929-0 --- ## 4.6 Scanner QA, maintenance, and long‑term reliability **What you need to know** - Scanners need a structured quality management approach, similar to any diagnostic instrument: - Defined roles for daily checks, periodic maintenance, and escalation. - Documented cleaning routines for optics, stages, and barcode readers. - Logs for downtime, rescans, and service visits. - Validation is not “one and done”: - After initial validation, you need ongoing monitoring to detect drift in focus, color, and coverage. - Major changes such as firmware updates, new objectives, or new slide types usually require at least a small re‑validation. - Many guidelines emphasize using **test slides** and **reference cases**: - Keep a panel of cases that are known to stress the system (e.g., thin biopsies, decalcified bone, cytology). - Use them when scanners are serviced, moved, or updated. - QA for scanners should be linked to broader digital pathology quality frameworks, covering: - Display performance and workstations. - Network and storage reliability. - Viewer behavior and measurement tools. - Written procedures and training for technologists are essential; most real‑world scanner problems are prevented or caught by well‑trained staff rather than by automation alone. **Reference** Cross S, Furness P, Igali L, Snead D, Treanor D. Best practice recommendations for implementing digital pathology. The Royal College of Pathologists; 2018. Available at: https://www.rcpath.org/static/f465d1b3-797b-4297-b7fedc00b4d77e51/Best-practice-recommendations-for-implementing-digital-pathology.pdf