Meeting Abstract Color Management in Digital Pathology

Meeting AbstractColor Management in Digital Pathology

W. Craig Revie,1 Mike Shires,2 Pete Jackson,2 David Brettle,3

Ravinder Cochrane,1 and Darren Treanor2,3

1FFEI Ltd., Hemel Hempstead HP2 7DF, UK2University of Leeds, Leeds LS2 9JT, UK3Leeds Teaching Hospitals NHS Trust, Leeds LS14 6UH, UK

Correspondence should be addressed to W. Craig Revie; [email protected]

Received 2 September 2014; Accepted 2 September 2014

Copyright © 2014 W. Craig Revie et al.This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

In digital microscopes and whole slide imaging systems,images of slides are captured, transmitted, and reproduced ona computer display. In order to allow pathologists to interpretthese images accurately and efficiently, it is important thatcolors from the slides are displayed in a consistent and reliablefashion.

The final color of the image presented to the viewingpathologist depends on several steps through the imag-ing pathway, including sample illumination, magnification,image capture, compression, storage, and reproduction onthe computer display.There aremany possible system designsand, within a single system, different setup options which canaffect the final image leading to significant variation in imageappearances.

This paper summarizes recent work by members of theInternational Color Consortium Medical Imaging WorkingGroup to develop test materials and methods for the assess-ment of color calibration of digital microscope systems. Thiswork includes sharing of ideas on device calibration andimage processing and display.

The paper further discusses the challenges encounteredin the development of a suitable color target that includes aset of patches with spectra similar to those encountered whenviewing pathology slides with stained tissue samples.

Background

Staining of tissue sections on glass slides is the foundation onwhich diagnosis and prognosis in pathology are based.

In digital microscopes and whole slide imaging (WSI)systems, images of slides are captured, transmitted, andreproduced on a computer display. In order to allow pathol-ogists to interpret these images accurately and efficiently, itis important that colors from the slides are displayed in aconsistent and reliable fashion.

The final color of the image presented to the viewingpathologist depends on several steps through the imag-ing pathway, including sample illumination, magnification,image capture, compression, storage, and reproduction onthe computer display as shown in Figure 1. There are manypossible system designs and, within a single system, differentsetup options which can affect the final image leading tosignificant variation in image appearances.

We have proposed an objective measure for the colorperformance of digital microscope systems which allows theoverall result of this processing sequence to be assessed.

Color Calibration Assessment for Pathology

Recognizing the importance of color reproduction, sev-eral digital pathology vendors and research groups haveattempted to develop color calibration systems for pathologyimages.

Silverstein et al. developed ICC color calibration profilesfor three displays focusing on the display end of the imagingworkflow [1] using an on-screen color target. The resultingprofiles can be used to compare display performance, butas they do not include measurement of the color variationintroduced by the imaging device (WSI instrument or digital

Hindawi Publishing CorporationAnalytical Cellular PathologyVolume 2014, Article ID 652757, 2 pageshttp://dx.doi.org/10.1155/2014/652757

2 Analytical Cellular Pathology

Ligh

t sou

rce

Mic

rosc

ope s

lide

Imag

ing

optic

s

RGB

sens

ors

Proc

essin

g

RGB

imag

eof

slid

e

View

erso

ftwar

e

Disp

lay an

dvi

ewin

g en

viro

nmen

t

Imag

eob

serv

er

Display moduleImage capture module

Figure 1: WSI image processing.

microscope), they do not allow for end-to-end color calibra-tion.

Others have used color calibration targets based onsynthetic targets such as photographic transparency of theMacbeth color target. Such film based methods involveimaging the target with the imaging device and then using theresulting image to develop a color correction profile specificto that device [2, 3]. Although these methods almost cer-tainly produce improvements in color consistency betweenscanners, they are limited by the fact that the colors areproduced by combinations of cyan, magenta, and yellow filmdyes which have substantially different spectra from those ofhistopathology stains. This difference is very likely to lead tosubstantial errors.

To address these limitations, tissue based color targetshave been proposed such as a section of mouse embryostained in a standard way. Although such targets have theadvantage of accurately representing the target material to beimaged, producing useful color phantoms based on stainedtissue is difficult due to the significant variation betweenstained slides.

A fundamental problem when assessing the way inwhich a digital microscope system handles color is that ofmetamerism. This is a well-known effect present in manysituations where two colors appear identical in one situationbut appear to be different in another.

Calibration Assessment Target

Wehave developed a novel color calibration assessment targetfor digital pathology as shown in Figure 2 which allowsthe color calibration of digital microscope systems to bemeasured objectively. The target includes a representative setof colored patches of biopolymer that have been stained withthe pathology stains.

Limitations and Future Work

The choice of stains used in this work reflects the UK’shistopathology and cytopathology practice. This is assumedto be broadly representative of worldwide pathology practice.There may be significant international variability in both thestains used and the exact chemical composition of the dyesused—for example, we analyzed both Meyer’s and Harris’shaematoxylin, but other methods may be in use which havenot been measured in this work. Although the pilot work

H&E stain area

(1) Haematoxylin(2) Neutral red(3) Light green FS(4) PAS(5) Methyl green

(6) Eosin

1 2 3

654

7 8 9

(7) Ponceau fuchsin(8) Aniline blue(9) Tartrazine

Sier

ra re

fere

nce s

lide

http

s://s

ierr

a.ffei

.co.u

kID

: 2014

.d.0001

Figure 2: WSI image processing.

described above produced a color calibration assessmenttarget with suitable characteristics for use, currently under-going testing, further development will likely be needed tomanufacture this device on a large scale.

Acknowledgments

The project described was in part funded by the UK Technol-ogy Strategy Board and Leeds Teaching Hospitals CharitableTrustees.

References

[1] L. D. Silverstein, S. F. Hashmi, K. Lang, and E. A. Krupinski,“Paradigm for achieving color-reproduction accuracy in LCDsfor medical imaging,” Journal of the Society for InformationDisplay, vol. 20, no. 1, pp. 53–62, 2012.

[2] P. Shrestha and B. Hulsken, “Color accuracy and reproducibilityin whole slide imaging scanners,” in Medical Imaging 2014:Digital Pathology, vol. 9041 of Proceedings of SPIE, 2014.

[3] P. Bautista, N. Hashimoto, and Y. Yagi, “Color standardizationin whole slide imaging using a color calibration slide,” Journalof Pathology Informatics, vol. 5, no. 1, article 4, 2014.

Submit your manuscripts athttp://www.hindawi.com

Stem CellsInternational

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014


MEDIATORSINFLAMMATION

of


Behavioural Neurology

EndocrinologyInternational Journal of



Disease Markers


BioMed Research International

OncologyJournal of



Oxidative Medicine and Cellular Longevity


PPAR Research

The Scientific World JournalHindawi Publishing Corporation http://www.hindawi.com Volume 2014

Immunology ResearchHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Journal of

ObesityJournal of



Computational and Mathematical Methods in Medicine

OphthalmologyJournal of


Diabetes ResearchJournal of



Research and TreatmentAIDS


Gastroenterology Research and Practice


Parkinson’s Disease

Evidence-Based Complementary and Alternative Medicine

Volume 2014Hindawi Publishing Corporationhttp://www.hindawi.com

Meeting Abstract Color Management in Digital Pathology

Documents

Transcript of Meeting Abstract Color Management in Digital Pathology