Image classification optimization of high resolution tissue images

M. Kozlovszky, K. Hegedus, G. Windisch, L. Kovács, G. Pintér

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

Abstract

Generic image classification methods are not performing well on tissue images. Such software solutions are producing high number of false negative and positive results, which prevents their clinical usage. We have created the MorphCeck high resolution tissue image processing framework, which enables us to collect morphological and morphometrical parameter values of the examined tissues. Size of such tissue images can easily reach the order of 100 MB-1 GB. Therefore, the image processing speed and effectiveness is an important factor. Our main goal is to accurately evaluate high resolution H-E (hematoxilin-eozin) stained colon tissue sample images, and based on the parameters classify the images into differentiated sets according to the structure and the surface manifestation of the tissues. We have interfaced our MorphCheck tissue image measurement software framework with the WND-CHARM general purpose image classifier and tried to classify high resolution tissue images with this combined software solution. The classification is by default initiated with a large training set and three main classes (healthy, adenoma, carcinoma), however the new image classification process' wall-clock time was intolerably high on single core PC. The processing time is depending on the size/resolution of the image and the size of the training set. Due to the tissue specific image parameters the classification effectiveness was promising. So we have started a development process to decrease the processing time and further increase the accuracy of the classification. We have developed a workflow based parallel version of the MorphCheck and WND-CHARM classifier software. In collaboration with the MTA SZTAKI Application Porting Centre the WND-CHARM has been ported to some distributed computing infrastructure (DCI). The paper introduces the steps that were taken to optimize WND-CHARM applications running faster using DCIs and some performance results of the tissue image classification process.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
PublisherSpringer Verlag
Pages532-539
Number of pages8
Volume8353 LNCS
ISBN (Print)9783662438794
DOIs
Publication statusPublished - 2014
Event9th International Conference on Large-Scale Scientific Computations, LSSC 2013 - Sozopol, Bulgaria
Duration: Jun 3 2013Jun 7 2013

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume8353 LNCS
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other9th International Conference on Large-Scale Scientific Computations, LSSC 2013
CountryBulgaria
CitySozopol
Period6/3/136/7/13

Fingerprint

Image classification
Image Classification
High Resolution
Tissue
Optimization
Software
Image Processing
Classify
Classifier
Image processing
Classifiers
Distributed Computing
Distributed computer systems
Development Process
Processing
Work Flow
Clocks
Infrastructure
Optimise
Decrease

Keywords

  • Application porting
  • gUSE
  • HP-SEE
  • Medical image processing workflow
  • MorphCheck
  • Scalability
  • WND-CHARM

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Kozlovszky, M., Hegedus, K., Windisch, G., Kovács, L., & Pintér, G. (2014). Image classification optimization of high resolution tissue images. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 8353 LNCS, pp. 532-539). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8353 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-662-43880-0_61

Image classification optimization of high resolution tissue images. / Kozlovszky, M.; Hegedus, K.; Windisch, G.; Kovács, L.; Pintér, G.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 8353 LNCS Springer Verlag, 2014. p. 532-539 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8353 LNCS).

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

Kozlovszky, M, Hegedus, K, Windisch, G, Kovács, L & Pintér, G 2014, Image classification optimization of high resolution tissue images. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 8353 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 8353 LNCS, Springer Verlag, pp. 532-539, 9th International Conference on Large-Scale Scientific Computations, LSSC 2013, Sozopol, Bulgaria, 6/3/13. https://doi.org/10.1007/978-3-662-43880-0_61
Kozlovszky M, Hegedus K, Windisch G, Kovács L, Pintér G. Image classification optimization of high resolution tissue images. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 8353 LNCS. Springer Verlag. 2014. p. 532-539. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-662-43880-0_61
Kozlovszky, M. ; Hegedus, K. ; Windisch, G. ; Kovács, L. ; Pintér, G. / Image classification optimization of high resolution tissue images. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 8353 LNCS Springer Verlag, 2014. pp. 532-539 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{2b9bd0aa3fda42bdb009d547a2b34424,
title = "Image classification optimization of high resolution tissue images",
abstract = "Generic image classification methods are not performing well on tissue images. Such software solutions are producing high number of false negative and positive results, which prevents their clinical usage. We have created the MorphCeck high resolution tissue image processing framework, which enables us to collect morphological and morphometrical parameter values of the examined tissues. Size of such tissue images can easily reach the order of 100 MB-1 GB. Therefore, the image processing speed and effectiveness is an important factor. Our main goal is to accurately evaluate high resolution H-E (hematoxilin-eozin) stained colon tissue sample images, and based on the parameters classify the images into differentiated sets according to the structure and the surface manifestation of the tissues. We have interfaced our MorphCheck tissue image measurement software framework with the WND-CHARM general purpose image classifier and tried to classify high resolution tissue images with this combined software solution. The classification is by default initiated with a large training set and three main classes (healthy, adenoma, carcinoma), however the new image classification process' wall-clock time was intolerably high on single core PC. The processing time is depending on the size/resolution of the image and the size of the training set. Due to the tissue specific image parameters the classification effectiveness was promising. So we have started a development process to decrease the processing time and further increase the accuracy of the classification. We have developed a workflow based parallel version of the MorphCheck and WND-CHARM classifier software. In collaboration with the MTA SZTAKI Application Porting Centre the WND-CHARM has been ported to some distributed computing infrastructure (DCI). The paper introduces the steps that were taken to optimize WND-CHARM applications running faster using DCIs and some performance results of the tissue image classification process.",
keywords = "Application porting, gUSE, HP-SEE, Medical image processing workflow, MorphCheck, Scalability, WND-CHARM",
author = "M. Kozlovszky and K. Hegedus and G. Windisch and L. Kov{\'a}cs and G. Pint{\'e}r",
year = "2014",
doi = "10.1007/978-3-662-43880-0_61",
language = "English",
isbn = "9783662438794",
volume = "8353 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "532--539",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - GEN

T1 - Image classification optimization of high resolution tissue images

AU - Kozlovszky, M.

AU - Hegedus, K.

AU - Windisch, G.

AU - Kovács, L.

AU - Pintér, G.

PY - 2014

Y1 - 2014

N2 - Generic image classification methods are not performing well on tissue images. Such software solutions are producing high number of false negative and positive results, which prevents their clinical usage. We have created the MorphCeck high resolution tissue image processing framework, which enables us to collect morphological and morphometrical parameter values of the examined tissues. Size of such tissue images can easily reach the order of 100 MB-1 GB. Therefore, the image processing speed and effectiveness is an important factor. Our main goal is to accurately evaluate high resolution H-E (hematoxilin-eozin) stained colon tissue sample images, and based on the parameters classify the images into differentiated sets according to the structure and the surface manifestation of the tissues. We have interfaced our MorphCheck tissue image measurement software framework with the WND-CHARM general purpose image classifier and tried to classify high resolution tissue images with this combined software solution. The classification is by default initiated with a large training set and three main classes (healthy, adenoma, carcinoma), however the new image classification process' wall-clock time was intolerably high on single core PC. The processing time is depending on the size/resolution of the image and the size of the training set. Due to the tissue specific image parameters the classification effectiveness was promising. So we have started a development process to decrease the processing time and further increase the accuracy of the classification. We have developed a workflow based parallel version of the MorphCheck and WND-CHARM classifier software. In collaboration with the MTA SZTAKI Application Porting Centre the WND-CHARM has been ported to some distributed computing infrastructure (DCI). The paper introduces the steps that were taken to optimize WND-CHARM applications running faster using DCIs and some performance results of the tissue image classification process.

AB - Generic image classification methods are not performing well on tissue images. Such software solutions are producing high number of false negative and positive results, which prevents their clinical usage. We have created the MorphCeck high resolution tissue image processing framework, which enables us to collect morphological and morphometrical parameter values of the examined tissues. Size of such tissue images can easily reach the order of 100 MB-1 GB. Therefore, the image processing speed and effectiveness is an important factor. Our main goal is to accurately evaluate high resolution H-E (hematoxilin-eozin) stained colon tissue sample images, and based on the parameters classify the images into differentiated sets according to the structure and the surface manifestation of the tissues. We have interfaced our MorphCheck tissue image measurement software framework with the WND-CHARM general purpose image classifier and tried to classify high resolution tissue images with this combined software solution. The classification is by default initiated with a large training set and three main classes (healthy, adenoma, carcinoma), however the new image classification process' wall-clock time was intolerably high on single core PC. The processing time is depending on the size/resolution of the image and the size of the training set. Due to the tissue specific image parameters the classification effectiveness was promising. So we have started a development process to decrease the processing time and further increase the accuracy of the classification. We have developed a workflow based parallel version of the MorphCheck and WND-CHARM classifier software. In collaboration with the MTA SZTAKI Application Porting Centre the WND-CHARM has been ported to some distributed computing infrastructure (DCI). The paper introduces the steps that were taken to optimize WND-CHARM applications running faster using DCIs and some performance results of the tissue image classification process.

KW - Application porting

KW - gUSE

KW - HP-SEE

KW - Medical image processing workflow

KW - MorphCheck

KW - Scalability

KW - WND-CHARM

UR - http://www.scopus.com/inward/record.url?scp=84904087622&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84904087622&partnerID=8YFLogxK

U2 - 10.1007/978-3-662-43880-0_61

DO - 10.1007/978-3-662-43880-0_61

M3 - Conference contribution

SN - 9783662438794

VL - 8353 LNCS

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

SP - 532

EP - 539

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

PB - Springer Verlag

ER -