Recent developments in digital pathology enable the rapid scanning of microscope slides at high resolution, making the digitisation of histopathology slides for routine diagnosis purposes feasible. An important initial step in the wider adoption of this technology is the establishment of validation data assessing how effective pathologists are using digital workstations in comparison to conventional light microscopes and glass slides when examining cases for primary diagnosis. I will report on the first study sufficiently powered to demonstrate a statistically valid equivalent (i.e. non-inferior) performance of digital pathology (DP) against standard glass slide (GS) microscopy. This study examined a total of 3,017 cases were included, generating 10,138 slides, which when scanned resulted in a digital archive of 2.45 terabytes. As well as demonstrating non-inferiority of digital in comparison to glass slides the study was useful in establishing rules for slide scanning and identifying areas where digital pathology has limitations and needs to be used with caution. Finally the presentation covers the impact adopting digital pathology will have on diagnostic laboratories, the economics of these changes and where these changes are most likely to benefit patients.

1. Digital pathology in routine
diagnostic histopathology
Dr David Snead
University Hospitals of Coventry and
Warwickshire NHS Trust and
Centre of Excellence for Digital Pathology
Coventry, UK.

2. Conflict of interests
• Omnyx funding for validation trial
• Omnyx/GE funding UHCW Digital Pathology
Centre of Excellence

3. Introduction
• 2011 UHCW entered an engagement with Omnyx
Digital pathology
• Whole slide imaging solution for diagnostic
histopathology
• Based on high throughput digital slide scanners and
networked diagnostic workstations
• Beta system tested in 2012
• Full system designed around the UHCW workload
2012

5. Requirements of a digital pathology
solution
• Rapid scanning
• Integration with the laboratory LIMS
• Stable
• Fast data transfer for real time reporting
• Validation - proven equity to light microscopy
• International standard for digital archiving

7. Why digital pathology at UHCW?
• Innovation
• Local pathology network needs
• Home working and remote reporting
• Academic potential
• Synergy with the University of Warwick computer
science department
• Enthusiastic consultant workforce
• Training opportunities

8. Immediate challenges
• Cost and return on investment
• Validation
• FDA decreed DP is not class 1 exempt. Pre-market
testing is required
• CPA require validation against existing technology
• None inferiority study designed
• Audit meeting variations used as benchmark of
internal variation

9. Slide Review
36.0% (1:56:13)
Other
16.0% (0:51:43)
Reporting
34.6%
(1:51:38)
Organizing Cases 24.1%
(0:10:25)
Querying for Cases 18.5%
(0:07:59)
Waiting for Delivery 11.2%
(0:04:49)
Matching 10.5%
(0:04:32)
Searching for Cases 9.4%
(0:04:04)
Transporting Cases 9.2%
(0:03:58)
Other 17.0%
(0:07:21)
Workflow Opportunities
100%
(0:43:09)
13.4
%
Pathologist T&M Study Results
Breakdown of Workflow Opportunities

11. Pre-allocation of specimens
Push system

12. Improved workflow efficiency
Pull system

13. CWPS MDT review

14. Validation study power calculation

15. Validation study design
• Double reporting
• Glass first digital second
• Minimum of 3 week washout period
• Compare reports to detect differences
• Steering group meets fortnightly to assess and classify
differences
• “Ground truth” assigned to one or other platform
• Study closed when3014 cases were double reported

16. Validation study methods
• Sequential cases in all subspecialties selected from
filing
• Slides received
• Cleaned / re-coverslipped
• Scanned
• Released to individual pathologists work bench and
subspecialty benches
• 14 pathologists involved
• 1/3 cases reported by the same pathologist 2/3 by
different pathologists

29. Cases 3,017
Slides scanned @ x40 (0.274um/pixel) 10,138
Slides scanned @ x60 (0.137um/pixel) 1,384
Data 2.45 TB
Estimated annual slide archive size 22TB
Scan speed per slide 90
seconds

30. 96 97 98 99 100
Percentage
All data
Same pathologists
Different pathologists
All data
Same pathologists
Different pathologists
3017
1009
2008
3017
1009
2008
99.3 (99, 99.6)
99.1 (98.5, 99.7)
99.4 (99.1, 99.7)
97.6 (97.1, 98.2)
97.2 (96.2, 98.2)
97.8 (97.2, 98.4)
Data used n Percentage (95% Confidence interval)
Completete concordance or no clinical difference
Completete concordance

32. X60 (0.137um/pixel) X40 (0.274um/pixel)

34. Problems
• Speed of streaming
• Tiles out of focus
• Colour reproduction with DPAS stains
• Screen fatigue

35. Challenges for routine practice
• Front and back end interface with LIMS needed
• Develop scanning rules
• Re-work laboratory protocols
• Improve section quality and tissue mounting
• Maintain streaming speed within the departmental security
protocol
• Some things will still need glass
• Polarisation
• Cytology
• Over sized blocks
• Low grade dysplasia
• X100 oil (scanty organisms)

36. Positives
• Scan speed excellent mean around 90 seconds
per slide
• Image quality
• Workflow software
• Very easy to use system
• Fits well in laboratory workflow
• Stable
• Excellent support

37. What does digital pathology offer?
• Economic advantages
• Increase efficiency of pathologists
• Reduce turn around time to report cases
• Improved review of cases including MDT/Tumour board review
• Quality advantages
• Reduced error rate
• Increased subspecialisation
• IHC scoring and indexing
• Tumour grading / dysplasia grading
• Cancer finder

38. Remote reporting
• RAS token remote login
• Ultra and Omnyx accessed through VRN
• Dragon voice recognition installed
• Backlogged cases available to report
• Report entered in and authorised
• Additional requests made via Ultra

39. Flexible workforce
• 39,000 surgicals
• 17 consultants (2,300 per wte)
• 14 in post (12.5 wte) (3,120 per wte)
• Outsourcing backlog to locums
• £30 per case
• Avoids employment costs i.e. PDP, appraisal,
prospective cover, sick leave, maternity leave etc.

40. Algorithms in development
• Improved accuracy and patient safety
• Cancer grading tool prostate, breast, and bladder
cancers
• Cancer finding tool, region of interest alert
• Alerts for slides or tissue samples not examined
• Overlay tool intelligently identifies regions of interest in
sequentially cut sections
• Automation downstream quantitative ICC e.g. ER, PR,
Ki67, HER2
• Quantification of tumour volume for molecular analysis

41. Digital pathology centre of excellence
• Mitotic count tool 3rd
AMIDA Grand Challenge Nagoya 2013
• Nuclear grading tool 1st
MITOS-Atypia 2014 Challenge
• Gland segmentation competition (GlaS) MICCAI Munich Conference
Oct 2015
• Tumour grading tool
• Cancer finding tool
• IHC slides with quantitative scores
• Resection margin, depth of invasion exported directly to report
Korsuk
SirinukunwattanaNasir Rajpoot Adnan Mujahid
Violeta
Kovacheva Nick Trahearn

43. Acknowledgements
• Aisha Meskiri
• Yee Wah Tsang
• Klaus Chen
• Bidisa Sinha
• Sari Suortamo
• Yen Yeo
• Elaine Blessing
• Shatrugan Sah
• Kishore Goparlakrishnan
• Emma Simmons
• Hesham El Daly
• Emma Simmons
• Sarah Read Jones
• Ian Cree
• Peter Kimani
• Ric Crossman