liu.seSearch for publications in DiVA
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • oxford
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Training Nuclei Detection Algorithms with Simple Annotations
Fraunhofer Mevis.
Fraunhofer Mevis.
Sectra AB.
Linköping University, Department of Science and Technology, Media and Information Technology. Linköping University, Center for Medical Image Science and Visualization (CMIV). Linköping University, Faculty of Science & Engineering.ORCID iD: 0000-0002-9368-0177
Show others and affiliations
2017 (English)In: Journal of Pathology Informatics, ISSN 2229-5089, E-ISSN 2153-3539, Vol. 8Article in journal (Refereed) Published
Abstract [en]

Background:

Generating good training datasets is essential for machine learning-based nuclei detection methods. However, creating exhaustive nuclei contour annotations, to derive optimal training data from, is often infeasible.

Methods:

We compared different approaches for training nuclei detection methods solely based on nucleus center markers. Such markers contain less accurate information, especially with regard to nuclear boundaries, but can be produced much easier and in greater quantities. The approaches use different automated sample extraction methods to derive image positions and class labels from nucleus center markers. In addition, the approaches use different automated sample selection methods to improve the detection quality of the classification algorithm and reduce the run time of the training process. We evaluated the approaches based on a previously published generic nuclei detection algorithm and a set of Ki-67-stained breast cancer images.

Results:

A Voronoi tessellation-based sample extraction method produced the best performing training sets. However, subsampling of the extracted training samples was crucial. Even simple class balancing improved the detection quality considerably. The incorporation of active learning led to a further increase in detection quality.

Conclusions:

With appropriate sample extraction and selection methods, nuclei detection algorithms trained on the basis of simple center marker annotations can produce comparable quality to algorithms trained on conventionally created training sets.

Place, publisher, year, edition, pages
2017. Vol. 8
Keywords [en]
Active learning, machine learning, nuclei detection, training set generation
National Category
Signal Processing
Identifiers
URN: urn:nbn:se:liu:diva-140749DOI: 10.4103/jpi.jpi_3_17OAI: oai:DiVA.org:liu-140749DiVA, id: diva2:1140183
Funder
VINNOVA, 2014-04257Available from: 2017-09-11 Created: 2017-09-11 Last updated: 2017-10-09

Open Access in DiVA

fulltext(2205 kB)17 downloads
File information
File name FULLTEXT01.pdfFile size 2205 kBChecksum SHA-512
4e780425513c53b0821e004082bbb88a88a1599d3e009bfdc2f48e9056045fa6765dca6fb1a89839278824192ac206071dc81104817a238ea85f6949fbb7a346
Type fulltextMimetype application/pdf

Other links

Publisher's full text

Search in DiVA

By author/editor
Lundström, Claes
By organisation
Media and Information TechnologyCenter for Medical Image Science and Visualization (CMIV)Faculty of Science & Engineering
In the same journal
Journal of Pathology Informatics
Signal Processing

Search outside of DiVA

GoogleGoogle Scholar
Total: 17 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

doi
urn-nbn

Altmetric score

doi
urn-nbn
Total: 93 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • oxford
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf