Deep learning-based diatom taxonomy on virtual slides
Abstract Deep convolutional neural networks are emerging as the state of the art method for supervised classification of images also in the context of taxonomic identification. Different morphologies and imaging technologies applied across organismal groups lead to highly specific image domains, whi...
Published in: | Scientific Reports |
---|---|
Main Authors: | , , , , |
Other Authors: | , |
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
Springer Science and Business Media LLC
2020
|
Subjects: | |
Online Access: | http://dx.doi.org/10.1038/s41598-020-71165-w https://www.nature.com/articles/s41598-020-71165-w.pdf https://www.nature.com/articles/s41598-020-71165-w |
id |
crspringernat:10.1038/s41598-020-71165-w |
---|---|
record_format |
openpolar |
spelling |
crspringernat:10.1038/s41598-020-71165-w 2023-05-15T18:25:35+02:00 Deep learning-based diatom taxonomy on virtual slides Kloster, Michael Langenkämper, Daniel Zurowietz, Martin Beszteri, Bánk Nattkemper, Tim W. Deutsche Forschungsgemeinschaft Projekt DEAL 2020 http://dx.doi.org/10.1038/s41598-020-71165-w https://www.nature.com/articles/s41598-020-71165-w.pdf https://www.nature.com/articles/s41598-020-71165-w en eng Springer Science and Business Media LLC https://creativecommons.org/licenses/by/4.0 https://creativecommons.org/licenses/by/4.0 CC-BY Scientific Reports volume 10, issue 1 ISSN 2045-2322 Multidisciplinary journal-article 2020 crspringernat https://doi.org/10.1038/s41598-020-71165-w 2022-01-14T15:43:18Z Abstract Deep convolutional neural networks are emerging as the state of the art method for supervised classification of images also in the context of taxonomic identification. Different morphologies and imaging technologies applied across organismal groups lead to highly specific image domains, which need customization of deep learning solutions. Here we provide an example using deep convolutional neural networks (CNNs) for taxonomic identification of the morphologically diverse microalgal group of diatoms. Using a combination of high-resolution slide scanning microscopy, web-based collaborative image annotation and diatom-tailored image analysis, we assembled a diatom image database from two Southern Ocean expeditions. We use these data to investigate the effect of CNN architecture, background masking, data set size and possible concept drift upon image classification performance. Surprisingly, VGG16, a relatively old network architecture, showed the best performance and generalizing ability on our images. Different from a previous study, we found that background masking slightly improved performance. In general, training only a classifier on top of convolutional layers pre-trained on extensive, but not domain-specific image data showed surprisingly high performance (F1 scores around 97%) with already relatively few (100–300) examples per class, indicating that domain adaptation to a novel taxonomic group can be feasible with a limited investment of effort. Article in Journal/Newspaper Southern Ocean Springer Nature (via Crossref) Southern Ocean Scientific Reports 10 1 |
institution |
Open Polar |
collection |
Springer Nature (via Crossref) |
op_collection_id |
crspringernat |
language |
English |
topic |
Multidisciplinary |
spellingShingle |
Multidisciplinary Kloster, Michael Langenkämper, Daniel Zurowietz, Martin Beszteri, Bánk Nattkemper, Tim W. Deep learning-based diatom taxonomy on virtual slides |
topic_facet |
Multidisciplinary |
description |
Abstract Deep convolutional neural networks are emerging as the state of the art method for supervised classification of images also in the context of taxonomic identification. Different morphologies and imaging technologies applied across organismal groups lead to highly specific image domains, which need customization of deep learning solutions. Here we provide an example using deep convolutional neural networks (CNNs) for taxonomic identification of the morphologically diverse microalgal group of diatoms. Using a combination of high-resolution slide scanning microscopy, web-based collaborative image annotation and diatom-tailored image analysis, we assembled a diatom image database from two Southern Ocean expeditions. We use these data to investigate the effect of CNN architecture, background masking, data set size and possible concept drift upon image classification performance. Surprisingly, VGG16, a relatively old network architecture, showed the best performance and generalizing ability on our images. Different from a previous study, we found that background masking slightly improved performance. In general, training only a classifier on top of convolutional layers pre-trained on extensive, but not domain-specific image data showed surprisingly high performance (F1 scores around 97%) with already relatively few (100–300) examples per class, indicating that domain adaptation to a novel taxonomic group can be feasible with a limited investment of effort. |
author2 |
Deutsche Forschungsgemeinschaft Projekt DEAL |
format |
Article in Journal/Newspaper |
author |
Kloster, Michael Langenkämper, Daniel Zurowietz, Martin Beszteri, Bánk Nattkemper, Tim W. |
author_facet |
Kloster, Michael Langenkämper, Daniel Zurowietz, Martin Beszteri, Bánk Nattkemper, Tim W. |
author_sort |
Kloster, Michael |
title |
Deep learning-based diatom taxonomy on virtual slides |
title_short |
Deep learning-based diatom taxonomy on virtual slides |
title_full |
Deep learning-based diatom taxonomy on virtual slides |
title_fullStr |
Deep learning-based diatom taxonomy on virtual slides |
title_full_unstemmed |
Deep learning-based diatom taxonomy on virtual slides |
title_sort |
deep learning-based diatom taxonomy on virtual slides |
publisher |
Springer Science and Business Media LLC |
publishDate |
2020 |
url |
http://dx.doi.org/10.1038/s41598-020-71165-w https://www.nature.com/articles/s41598-020-71165-w.pdf https://www.nature.com/articles/s41598-020-71165-w |
geographic |
Southern Ocean |
geographic_facet |
Southern Ocean |
genre |
Southern Ocean |
genre_facet |
Southern Ocean |
op_source |
Scientific Reports volume 10, issue 1 ISSN 2045-2322 |
op_rights |
https://creativecommons.org/licenses/by/4.0 https://creativecommons.org/licenses/by/4.0 |
op_rightsnorm |
CC-BY |
op_doi |
https://doi.org/10.1038/s41598-020-71165-w |
container_title |
Scientific Reports |
container_volume |
10 |
container_issue |
1 |
_version_ |
1766207131862695936 |