Training echo state networks for rotation-invariant bone marrow cell classification
dc.contributor.author
Kainz, Philipp
dc.contributor.author
Burgsteiner, Harald
dc.contributor.author
Asslaber, Martin
dc.contributor.author
Ahammer, Helmut
dc.date.accessioned
2017-10-13T13:13:21Z
dc.date.available
2017-10-06T02:44:57Z
dc.date.available
2017-10-13T13:13:21Z
dc.date.issued
2017-06
dc.identifier.issn
1433-3058
dc.identifier.issn
0941-0643
dc.identifier.other
10.1007/s00521-016-2609-9
en_US
dc.identifier.uri
http://hdl.handle.net/20.500.11850/191020
dc.identifier.doi
10.3929/ethz-b-000191020
dc.description.abstract
The main principle of diagnostic pathology is the reliable interpretation of individual cells in context of the tissue architecture. Especially a confident examination of bone marrow specimen is dependent on a valid classification of myeloid cells. In this work, we propose a novel rotation-invariant learning scheme for multi-class echo state networks (ESNs), which achieves very high performance in automated bone marrow cell classification. Based on representing static images as temporal sequence of rotations, we show how ESNs robustly recognize cells of arbitrary rotations by taking advantage of their short-term memory capacity. The performance of our approach is compared to a classification random forest that learns rotation-invariance in a conventional way by exhaustively training on multiple rotations of individual samples. The methods were evaluated on a human bone marrow image database consisting of granulopoietic and erythropoietic cells in different maturation stages. Our ESN approach to cell classification does not rely on segmentation of cells or manual feature extraction and can therefore directly be applied to image data.
en_US
dc.format
application/pdf
dc.language.iso
en
en_US
dc.publisher
Springer
en_US
dc.rights.uri
http://creativecommons.org/licenses/by/4.0/
dc.subject
Computer-assisted pathology
en_US
dc.subject
Histopathological image analysis
en_US
dc.subject
Bone marrow cell classification
en_US
dc.subject
Echo state networks
en_US
dc.subject
Reservoir computing
en_US
dc.title
Training echo state networks for rotation-invariant bone marrow cell classification
en_US
dc.type
Journal Article
dc.rights.license
Creative Commons Attribution 4.0 International
dc.date.published
2016-09-21
ethz.journal.title
Neural Computing and Applications
ethz.journal.volume
28
en_US
ethz.journal.issue
6
en_US
ethz.journal.abbreviated
Neural Comput & Applic
ethz.pages.start
1277
en_US
ethz.pages.end
1292
en_US
ethz.version.deposit
publishedVersion
en_US
ethz.identifier.wos
ethz.identifier.scopus
ethz.publication.place
London
en_US
ethz.publication.status
published
en_US
ethz.date.deposited
2017-10-06T02:45:14Z
ethz.source
WOS
ethz.eth
yes
en_US
ethz.availability
Open access
en_US
ethz.rosetta.installDate
2017-10-13T13:13:24Z
ethz.rosetta.lastUpdated
2021-02-14T19:20:39Z
ethz.rosetta.versionExported
true
ethz.COinS
ctx_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.atitle=Training%20echo%20state%20networks%20for%20rotation-invariant%20bone%20marrow%20cell%20classification&rft.jtitle=Neural%20Computing%20and%20Applications&rft.date=2017-06&rft.volume=28&rft.issue=6&rft.spage=1277&rft.epage=1292&rft.issn=1433-3058&0941-0643&rft.au=Kainz,%20Philipp&Burgsteiner,%20Harald&Asslaber,%20Martin&Ahammer,%20Helmut&rft.genre=article&rft_id=info:doi/10.1007/s00521-016-2609-9&
Files in this item
Publication type
-
Journal Article [133719]