Age prediction by deep learning applied to Greenland halibut (Reinhardtius hippoglossoides) otolith images
Otoliths (ear-stones) in the inner ears of vertebrates containing visible year zones are used extensively to determine fish age. Analysis of otoliths is a time-consuming and difficult task that requires the education of human experts. Human age estimates are inconsistent, as several readings by the...
Published in: | PLOS ONE |
---|---|
Main Authors: | , , |
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
Public Library of Science
2022
|
Subjects: | |
Online Access: | https://hdl.handle.net/10037/27289 https://doi.org/10.1371/journal.pone.0277244 |
id |
ftunivtroemsoe:oai:munin.uit.no:10037/27289 |
---|---|
record_format |
openpolar |
spelling |
ftunivtroemsoe:oai:munin.uit.no:10037/27289 2023-05-15T16:28:07+02:00 Age prediction by deep learning applied to Greenland halibut (Reinhardtius hippoglossoides) otolith images Martinsen, Ivar Harbitz, Alf Bianchi, Filippo Maria 2022-11-04 https://hdl.handle.net/10037/27289 https://doi.org/10.1371/journal.pone.0277244 eng eng Public Library of Science PLOS ONE Martinsen I, Harbitz, Bianchi. Age prediction by deep learning applied to Greenland halibut (Reinhardtius hippoglossoides) otolith images. PLOS ONE. 2022;17(11) FRIDAID 2069451 doi:10.1371/journal.pone.0277244 1932-6203 https://hdl.handle.net/10037/27289 Attribution 4.0 International (CC BY 4.0) openAccess Copyright 2022 The Author(s) https://creativecommons.org/licenses/by/4.0 CC-BY Journal article Tidsskriftartikkel Peer reviewed publishedVersion 2022 ftunivtroemsoe https://doi.org/10.1371/journal.pone.0277244 2022-11-10T00:01:31Z Otoliths (ear-stones) in the inner ears of vertebrates containing visible year zones are used extensively to determine fish age. Analysis of otoliths is a time-consuming and difficult task that requires the education of human experts. Human age estimates are inconsistent, as several readings by the same human expert might result in different ages assigned to the same otolith, in addition to an inherent bias between readers. To improve efficiency and resolve inconsistent results in the age reading from otolith images by human experts, an automated procedure based on convolutional neural networks (CNNs), a class of deep learning models suitable for image processing, is investigated. We applied CNNs that perform image regression to estimate the age of Greenland halibut (Reinhardtius hippoglossoides) with good results for individual ages as well as the overall age distribution, with an average CV of about 10% relative to the read ages by experts. In addition, the density distribution of predicted ages resembles the density distribution of the ground truth. By using k*lfold cross-validation, we test all available samples, and we show that the results are rather sensitive to the choice of test set. Finally, we apply explanation techniques to analyze the decision process of deep learning models. In particular, we produce heatmaps indicating which input features that are the most important in the computation of predicted age. Article in Journal/Newspaper Greenland University of Tromsø: Munin Open Research Archive Greenland PLOS ONE 17 11 e0277244 |
institution |
Open Polar |
collection |
University of Tromsø: Munin Open Research Archive |
op_collection_id |
ftunivtroemsoe |
language |
English |
description |
Otoliths (ear-stones) in the inner ears of vertebrates containing visible year zones are used extensively to determine fish age. Analysis of otoliths is a time-consuming and difficult task that requires the education of human experts. Human age estimates are inconsistent, as several readings by the same human expert might result in different ages assigned to the same otolith, in addition to an inherent bias between readers. To improve efficiency and resolve inconsistent results in the age reading from otolith images by human experts, an automated procedure based on convolutional neural networks (CNNs), a class of deep learning models suitable for image processing, is investigated. We applied CNNs that perform image regression to estimate the age of Greenland halibut (Reinhardtius hippoglossoides) with good results for individual ages as well as the overall age distribution, with an average CV of about 10% relative to the read ages by experts. In addition, the density distribution of predicted ages resembles the density distribution of the ground truth. By using k*lfold cross-validation, we test all available samples, and we show that the results are rather sensitive to the choice of test set. Finally, we apply explanation techniques to analyze the decision process of deep learning models. In particular, we produce heatmaps indicating which input features that are the most important in the computation of predicted age. |
format |
Article in Journal/Newspaper |
author |
Martinsen, Ivar Harbitz, Alf Bianchi, Filippo Maria |
spellingShingle |
Martinsen, Ivar Harbitz, Alf Bianchi, Filippo Maria Age prediction by deep learning applied to Greenland halibut (Reinhardtius hippoglossoides) otolith images |
author_facet |
Martinsen, Ivar Harbitz, Alf Bianchi, Filippo Maria |
author_sort |
Martinsen, Ivar |
title |
Age prediction by deep learning applied to Greenland halibut (Reinhardtius hippoglossoides) otolith images |
title_short |
Age prediction by deep learning applied to Greenland halibut (Reinhardtius hippoglossoides) otolith images |
title_full |
Age prediction by deep learning applied to Greenland halibut (Reinhardtius hippoglossoides) otolith images |
title_fullStr |
Age prediction by deep learning applied to Greenland halibut (Reinhardtius hippoglossoides) otolith images |
title_full_unstemmed |
Age prediction by deep learning applied to Greenland halibut (Reinhardtius hippoglossoides) otolith images |
title_sort |
age prediction by deep learning applied to greenland halibut (reinhardtius hippoglossoides) otolith images |
publisher |
Public Library of Science |
publishDate |
2022 |
url |
https://hdl.handle.net/10037/27289 https://doi.org/10.1371/journal.pone.0277244 |
geographic |
Greenland |
geographic_facet |
Greenland |
genre |
Greenland |
genre_facet |
Greenland |
op_relation |
PLOS ONE Martinsen I, Harbitz, Bianchi. Age prediction by deep learning applied to Greenland halibut (Reinhardtius hippoglossoides) otolith images. PLOS ONE. 2022;17(11) FRIDAID 2069451 doi:10.1371/journal.pone.0277244 1932-6203 https://hdl.handle.net/10037/27289 |
op_rights |
Attribution 4.0 International (CC BY 4.0) openAccess Copyright 2022 The Author(s) https://creativecommons.org/licenses/by/4.0 |
op_rightsnorm |
CC-BY |
op_doi |
https://doi.org/10.1371/journal.pone.0277244 |
container_title |
PLOS ONE |
container_volume |
17 |
container_issue |
11 |
container_start_page |
e0277244 |
_version_ |
1766017740272828416 |