DIGITAL PATHOLOGY RESULTS SUMMARY
StatLab KT5+ Advanced Adhesion slides standard deviation
Competitor Adhesion slides standard deviation
RESULTS OF SCAN TIMES
StatLab KT5+ Advanced Adhesion slides Competitor Adhesion slides
Slide count, n=25 for each brand RESULTS OF FILE SIZE
StatLab KT5+ Advanced Adhesion slides Competitor Adhesion slides
Slide count, n=25 for each brand Table 1. Study results comparing digital pathology processes for slides with either ‘moderate-to-severe’ (competitor) or minimal (StatLab KT5+) background staining.
using two adhesion slide brands processed with identical GMS staining protocols. Slides were visually categorised as having either ‘moderate-to-severe’ or ‘minimal’ background staining and scanned at 20x on a high-resolution WSI system with automated tissue detection. Slides exhibiting moderate-to-severe
staining took six to seven times longer to scan and produced file sizes four times larger than slides with minimal background staining. These findings highlight how stain optimisation and slide quality directly affect laboratory throughput, scanner utilisation, and digital storage demands (Table 1). These inefficiencies translate into increased labour hours, equipment utilisation, data storage requirements and cost, as well as slower diagnostic turnaround.4
AI impact Machine learning algorithms are dependent on data quality. AI models trained to detect cancer subtypes, quantify biomarker expression, or grade tumours rely on high-resolution, artefact-free image datasets to detect paterns and subtle changes in tissue morphology. These models are trained on large datasets of digitised slides, and the ability to make reliable predictions is based on the consistency and clarity of
those images. Even minor imperfections and inconsistencies, like blurred sections or air bubbles, introduce visual noise into training datasets and mislead AI models. Unlike human pathologists, algorithms
cannot distinguish biologically meaningful structures from optical artefacts. Studies evaluating AI in pathology have shown that variability in staining, slide preparation, and imaging conditions can significantly impact AI model accuracy and transferability across facilities.5,6
As
a result, training datasets contaminated with imperfections may lead to reduced diagnostic accuracy and increased false positives or negatives. In response to this, researchers at the
University of Stavanger developed deep learning algorithms designed to identify and exclude slide artefacts from tissue analysis. Their work demonstrated that automated artefact detection significantly improves segmentation accuracy and diagnostic reliability in computational pathology workflows.7 However, while downstream artefact
removal helps mitigate quality issues, it introduces additional computational overhead and complexity. A more sustainable approach is preventing artefact introduction at the source starting with superior glass quality and optimised slide preparation.
Standardisation and quality control Despite the growing adoption of digital pathology across healthcare systems, there is a lack of global standards governing:
Glass slide quality Cover glass specifications Staining protocols Mounting procedures Scanning parameters Image file formats AI validation methodologies.
As a result, laboratories operate within highly variable ecosystems. Slides that perform optimally on one scanner may produce suboptimal images on another. Algorithms trained on one dataset may fail when deployed elsewhere, limiting scalability and interoperability. The College of American Pathologists (CAP) has published guidelines addressing validation of whole slide imaging systems for diagnostic use, emphasising the need for rigorous quality control across the digital workflow. However, explicit standards for slide and cover glass optical performance remain underdeveloped.8 International organisations such
Even minor imperfections and
inconsistencies, like blurred sections or air bubbles, introduce visual noise into training datasets and mislead AI models
as the International Organization for Standardization (ISO), Clinical and Laboratory Standards Institute (CLSI), and the Digital Pathology Association are increasingly advocating for harmonised products that address preanalytical and material variables. Establishing formal slide quality specifications, similar to those used in clinical imaging, represents a critical next step.
Training for the digital lens The effectiveness of digital pathology not only depends on the quality of slides and global standardisation, but also on
June 2026
WWW.PATHOLOGYINPRACTICE.COM 49
Average scan time <4min/slide
0.096331969
27 min/slide 0.360806551
0-5 minutes 20 0
0GB–.3GB 22 0
Average file size 0.23GB
.1GB
1.07GB .18GB
5-15 minutes 5 0
.3GB–.6GB 3 0
15-25 minutes 0
13
.6GB–1GB 0
13 $535
25+ minutes 0
12
1.0GB+ 0
12
Est. storage cost for 10,000 slides ($0.05/GB) $115
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56
