Automatic image quality assessment for digital pathology

Ali R.N. Avanaki, Kathryn S. Espig, Albert Xthona, Christian Lanciault, Tom R.L. Kimpe

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Citations (Scopus)

Abstract

Slide quality is an important factor in pathology workflow and diagnosis. We examine the extent of quality variations in digitized hematoxylin-eosin (H&E) slides due to variations and errors in staining and/or scanning (e.g., outof- focus blur & stitching). We propose two automatic quality estimators by adapting image quality assessment (IQA) methods that are originally developed for natural images. For the first estimator, we assume a gold-standard reference digital pathology slide is available. Quality of a given slide is estimated by comparing the slide to such a reference using a full-reference perceptual IQA method such as VIF (visual information fidelity) or SSIM (structural similarity metric). Our second estimator is based on IL-NIQE (integrated local natural image quality evaluator), a no-reference IQA, which we train using a set of artifact-free H&E high-power images (20× or 40×) from breast tissue. The first estimator (referenced) predicts marked quality reduction of images with simulated blurring as compared to the artifact-free originals used as references. The histograms of scores by the second estimator (no-reference) for images with artifact (blur, stitching, folded tissue, or air bubble artifacts) and for artifact-free images are highly separable. Moreover, the scores by the second estimator are correlated with the ratings given by a pathologist. We conclude that our approach is promising and further research is outlined for developing robust automatic quality estimators.

Original languageEnglish (US)
Title of host publicationBreast Imaging - 13th International Workshop, IWDM 2016, Proceedings
EditorsKristina Lang, Anders Tingberg, Pontus Timberg
PublisherSpringer Verlag
Pages431-438
Number of pages8
ISBN (Print)9783319415451
DOIs
StatePublished - Jan 1 2016
Event13th International Workshop on Breast Imaging, IWDM 2016 - Malmo, Sweden
Duration: Jun 19 2016Jun 22 2016

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume9699
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference13th International Workshop on Breast Imaging, IWDM 2016
CountrySweden
CityMalmo
Period6/19/166/22/16

Fingerprint

Image Quality Assessment
Pathology
Image quality
Estimator
Stitching
Tissue
Gold
Structural Similarity
Scanning
Image Quality
High Power
Fidelity
Histogram
Bubble
Work Flow
Air
Metric
Predict

Keywords

  • Whole slide imaging (WSI)

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Avanaki, A. R. N., Espig, K. S., Xthona, A., Lanciault, C., & Kimpe, T. R. L. (2016). Automatic image quality assessment for digital pathology. In K. Lang, A. Tingberg, & P. Timberg (Eds.), Breast Imaging - 13th International Workshop, IWDM 2016, Proceedings (pp. 431-438). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9699). Springer Verlag. https://doi.org/10.1007/978-3-319-41546-8_54

Automatic image quality assessment for digital pathology. / Avanaki, Ali R.N.; Espig, Kathryn S.; Xthona, Albert; Lanciault, Christian; Kimpe, Tom R.L.

Breast Imaging - 13th International Workshop, IWDM 2016, Proceedings. ed. / Kristina Lang; Anders Tingberg; Pontus Timberg. Springer Verlag, 2016. p. 431-438 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9699).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Avanaki, ARN, Espig, KS, Xthona, A, Lanciault, C & Kimpe, TRL 2016, Automatic image quality assessment for digital pathology. in K Lang, A Tingberg & P Timberg (eds), Breast Imaging - 13th International Workshop, IWDM 2016, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9699, Springer Verlag, pp. 431-438, 13th International Workshop on Breast Imaging, IWDM 2016, Malmo, Sweden, 6/19/16. https://doi.org/10.1007/978-3-319-41546-8_54
Avanaki ARN, Espig KS, Xthona A, Lanciault C, Kimpe TRL. Automatic image quality assessment for digital pathology. In Lang K, Tingberg A, Timberg P, editors, Breast Imaging - 13th International Workshop, IWDM 2016, Proceedings. Springer Verlag. 2016. p. 431-438. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-41546-8_54
Avanaki, Ali R.N. ; Espig, Kathryn S. ; Xthona, Albert ; Lanciault, Christian ; Kimpe, Tom R.L. / Automatic image quality assessment for digital pathology. Breast Imaging - 13th International Workshop, IWDM 2016, Proceedings. editor / Kristina Lang ; Anders Tingberg ; Pontus Timberg. Springer Verlag, 2016. pp. 431-438 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{9b1a9544ef9e4bf1b51f15f131a87223,
title = "Automatic image quality assessment for digital pathology",
abstract = "Slide quality is an important factor in pathology workflow and diagnosis. We examine the extent of quality variations in digitized hematoxylin-eosin (H&E) slides due to variations and errors in staining and/or scanning (e.g., outof- focus blur & stitching). We propose two automatic quality estimators by adapting image quality assessment (IQA) methods that are originally developed for natural images. For the first estimator, we assume a gold-standard reference digital pathology slide is available. Quality of a given slide is estimated by comparing the slide to such a reference using a full-reference perceptual IQA method such as VIF (visual information fidelity) or SSIM (structural similarity metric). Our second estimator is based on IL-NIQE (integrated local natural image quality evaluator), a no-reference IQA, which we train using a set of artifact-free H&E high-power images (20× or 40×) from breast tissue. The first estimator (referenced) predicts marked quality reduction of images with simulated blurring as compared to the artifact-free originals used as references. The histograms of scores by the second estimator (no-reference) for images with artifact (blur, stitching, folded tissue, or air bubble artifacts) and for artifact-free images are highly separable. Moreover, the scores by the second estimator are correlated with the ratings given by a pathologist. We conclude that our approach is promising and further research is outlined for developing robust automatic quality estimators.",
keywords = "Whole slide imaging (WSI)",
author = "Avanaki, {Ali R.N.} and Espig, {Kathryn S.} and Albert Xthona and Christian Lanciault and Kimpe, {Tom R.L.}",
year = "2016",
month = "1",
day = "1",
doi = "10.1007/978-3-319-41546-8_54",
language = "English (US)",
isbn = "9783319415451",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "431--438",
editor = "Kristina Lang and Anders Tingberg and Pontus Timberg",
booktitle = "Breast Imaging - 13th International Workshop, IWDM 2016, Proceedings",
address = "Germany",

}

TY - GEN

T1 - Automatic image quality assessment for digital pathology

AU - Avanaki, Ali R.N.

AU - Espig, Kathryn S.

AU - Xthona, Albert

AU - Lanciault, Christian

AU - Kimpe, Tom R.L.

PY - 2016/1/1

Y1 - 2016/1/1

N2 - Slide quality is an important factor in pathology workflow and diagnosis. We examine the extent of quality variations in digitized hematoxylin-eosin (H&E) slides due to variations and errors in staining and/or scanning (e.g., outof- focus blur & stitching). We propose two automatic quality estimators by adapting image quality assessment (IQA) methods that are originally developed for natural images. For the first estimator, we assume a gold-standard reference digital pathology slide is available. Quality of a given slide is estimated by comparing the slide to such a reference using a full-reference perceptual IQA method such as VIF (visual information fidelity) or SSIM (structural similarity metric). Our second estimator is based on IL-NIQE (integrated local natural image quality evaluator), a no-reference IQA, which we train using a set of artifact-free H&E high-power images (20× or 40×) from breast tissue. The first estimator (referenced) predicts marked quality reduction of images with simulated blurring as compared to the artifact-free originals used as references. The histograms of scores by the second estimator (no-reference) for images with artifact (blur, stitching, folded tissue, or air bubble artifacts) and for artifact-free images are highly separable. Moreover, the scores by the second estimator are correlated with the ratings given by a pathologist. We conclude that our approach is promising and further research is outlined for developing robust automatic quality estimators.

AB - Slide quality is an important factor in pathology workflow and diagnosis. We examine the extent of quality variations in digitized hematoxylin-eosin (H&E) slides due to variations and errors in staining and/or scanning (e.g., outof- focus blur & stitching). We propose two automatic quality estimators by adapting image quality assessment (IQA) methods that are originally developed for natural images. For the first estimator, we assume a gold-standard reference digital pathology slide is available. Quality of a given slide is estimated by comparing the slide to such a reference using a full-reference perceptual IQA method such as VIF (visual information fidelity) or SSIM (structural similarity metric). Our second estimator is based on IL-NIQE (integrated local natural image quality evaluator), a no-reference IQA, which we train using a set of artifact-free H&E high-power images (20× or 40×) from breast tissue. The first estimator (referenced) predicts marked quality reduction of images with simulated blurring as compared to the artifact-free originals used as references. The histograms of scores by the second estimator (no-reference) for images with artifact (blur, stitching, folded tissue, or air bubble artifacts) and for artifact-free images are highly separable. Moreover, the scores by the second estimator are correlated with the ratings given by a pathologist. We conclude that our approach is promising and further research is outlined for developing robust automatic quality estimators.

KW - Whole slide imaging (WSI)

UR - http://www.scopus.com/inward/record.url?scp=84977595916&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84977595916&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-41546-8_54

DO - 10.1007/978-3-319-41546-8_54

M3 - Conference contribution

AN - SCOPUS:84977595916

SN - 9783319415451

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 431

EP - 438

BT - Breast Imaging - 13th International Workshop, IWDM 2016, Proceedings

A2 - Lang, Kristina

A2 - Tingberg, Anders

A2 - Timberg, Pontus

PB - Springer Verlag

ER -