Fast, simple calcium imaging segmentation with fully convolutional networks

Aleksander Klibisz; Derek Rose; Matthew Eicholtz; Jay Blundon; Stanislav Zakharenko

doi:10.1007/978-3-319-67558-9_33

Fast, simple calcium imaging segmentation with fully convolutional networks

Aleksander Klibisz, Derek Rose, Matthew Eicholtz, Jay Blundon, Stanislav Zakharenko

Multimodal Sensor Analytics

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

31 Scopus citations

Abstract

Calcium imaging is a technique for observing neuron activity as a series of images showing indicator fluorescence over time. Manually segmenting neurons is time-consuming, leading to research on automated calcium imaging segmentation (ACIS). We evaluated several deep learning models for ACIS on the Neurofinder competition datasets and report our best model: U-Net2DS, a fully convolutional network that operates on 2D mean summary images. U-Net2DS requires minimal domain-specific pre/post-processing and parameter adjustment, and predictions are made on full 512 × 512 images at ≈9K images per minute. It ranks third in the Neurofinder competition (F:1=0.57) and is the best model to exclusively use deep learning. We also demonstrate useful segmentations on data from outside the competition. The model’s simplicity, speed, and quality results make it a practical choice for ACIS and a strong baseline for more complex models in the future.

Original language	English
Title of host publication	Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support - 3rd International Workshop, DLMIA 2017 and 7th International Workshop, ML-CDS 2017 Held in Conjunction with MICCAI 2017, Proceedings
Editors	Tal Arbel, M. Jorge Cardoso
Publisher	Springer Verlag
Pages	285-293
Number of pages	9
ISBN (Print)	9783319675572
DOIs	https://doi.org/10.1007/978-3-319-67558-9_33
State	Published - 2017
Event	3rd International Workshop on Deep Learning in Medical Image Analysis, DLMIA 2017 and 7th International Workshop on Multimodal Learning for Clinical Decision Support, ML-CDS 2017 held in Conjunction with 20th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2017 - Quebec City, Canada Duration: Sep 14 2017 → Sep 14 2017

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	10553 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	3rd International Workshop on Deep Learning in Medical Image Analysis, DLMIA 2017 and 7th International Workshop on Multimodal Learning for Clinical Decision Support, ML-CDS 2017 held in Conjunction with 20th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2017
Country/Territory	Canada
City	Quebec City
Period	09/14/17 → 09/14/17

Funding

This manuscript has been authored by UT-Battelle, LLC under Contract No. DE-AC05-00OR22725 with the U.S. Department of Energy. The United States Government retains and the publisher, by accepting the article for publication, acknowledges that the United States Government retains a non-exclusive, paid-up, irrevocable, worldwide license to publish or reproduce the published form of this manuscript, or allow others to do so, for United States Government purposes. The Department of Energy will provide public access to these results of federally sponsored research in accordance with the DOE Public Access Plan (http://energy.gov/downloads/ doe-public-access-plan).

Keywords

Calcium imaging
Deep learning
Fully convolutional networks
Microscopy segmentation

Access to Document

10.1007/978-3-319-67558-9_33

Cite this

Klibisz, A., Rose, D., Eicholtz, M., Blundon, J., & Zakharenko, S. (2017). Fast, simple calcium imaging segmentation with fully convolutional networks. In T. Arbel, & M. J. Cardoso (Eds.), Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support - 3rd International Workshop, DLMIA 2017 and 7th International Workshop, ML-CDS 2017 Held in Conjunction with MICCAI 2017, Proceedings (pp. 285-293). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10553 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-67558-9_33

Klibisz, Aleksander ; Rose, Derek ; Eicholtz, Matthew et al. / Fast, simple calcium imaging segmentation with fully convolutional networks. Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support - 3rd International Workshop, DLMIA 2017 and 7th International Workshop, ML-CDS 2017 Held in Conjunction with MICCAI 2017, Proceedings. editor / Tal Arbel ; M. Jorge Cardoso. Springer Verlag, 2017. pp. 285-293 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{bc4d538fba8a4b17b9db04dde3eee52d,

title = "Fast, simple calcium imaging segmentation with fully convolutional networks",

abstract = "Calcium imaging is a technique for observing neuron activity as a series of images showing indicator fluorescence over time. Manually segmenting neurons is time-consuming, leading to research on automated calcium imaging segmentation (ACIS). We evaluated several deep learning models for ACIS on the Neurofinder competition datasets and report our best model: U-Net2DS, a fully convolutional network that operates on 2D mean summary images. U-Net2DS requires minimal domain-specific pre/post-processing and parameter adjustment, and predictions are made on full 512 × 512 images at ≈9K images per minute. It ranks third in the Neurofinder competition (F:1=0.57) and is the best model to exclusively use deep learning. We also demonstrate useful segmentations on data from outside the competition. The model{\textquoteright}s simplicity, speed, and quality results make it a practical choice for ACIS and a strong baseline for more complex models in the future.",

keywords = "Calcium imaging, Deep learning, Fully convolutional networks, Microscopy segmentation",

author = "Aleksander Klibisz and Derek Rose and Matthew Eicholtz and Jay Blundon and Stanislav Zakharenko",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing AG 2017.; 3rd International Workshop on Deep Learning in Medical Image Analysis, DLMIA 2017 and 7th International Workshop on Multimodal Learning for Clinical Decision Support, ML-CDS 2017 held in Conjunction with 20th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2017 ; Conference date: 14-09-2017 Through 14-09-2017",

year = "2017",

doi = "10.1007/978-3-319-67558-9_33",

language = "English",

isbn = "9783319675572",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "285--293",

editor = "Tal Arbel and Cardoso, {M. Jorge}",

booktitle = "Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support - 3rd International Workshop, DLMIA 2017 and 7th International Workshop, ML-CDS 2017 Held in Conjunction with MICCAI 2017, Proceedings",

}

Klibisz, A, Rose, D, Eicholtz, M, Blundon, J & Zakharenko, S 2017, Fast, simple calcium imaging segmentation with fully convolutional networks. in T Arbel & MJ Cardoso (eds), Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support - 3rd International Workshop, DLMIA 2017 and 7th International Workshop, ML-CDS 2017 Held in Conjunction with MICCAI 2017, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10553 LNCS, Springer Verlag, pp. 285-293, 3rd International Workshop on Deep Learning in Medical Image Analysis, DLMIA 2017 and 7th International Workshop on Multimodal Learning for Clinical Decision Support, ML-CDS 2017 held in Conjunction with 20th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2017, Quebec City, Canada, 09/14/17. https://doi.org/10.1007/978-3-319-67558-9_33

Fast, simple calcium imaging segmentation with fully convolutional networks. / Klibisz, Aleksander; Rose, Derek; Eicholtz, Matthew et al.
Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support - 3rd International Workshop, DLMIA 2017 and 7th International Workshop, ML-CDS 2017 Held in Conjunction with MICCAI 2017, Proceedings. ed. / Tal Arbel; M. Jorge Cardoso. Springer Verlag, 2017. p. 285-293 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10553 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Fast, simple calcium imaging segmentation with fully convolutional networks

AU - Klibisz, Aleksander

AU - Rose, Derek

AU - Eicholtz, Matthew

AU - Blundon, Jay

AU - Zakharenko, Stanislav

N1 - Publisher Copyright: © Springer International Publishing AG 2017.

PY - 2017

Y1 - 2017

N2 - Calcium imaging is a technique for observing neuron activity as a series of images showing indicator fluorescence over time. Manually segmenting neurons is time-consuming, leading to research on automated calcium imaging segmentation (ACIS). We evaluated several deep learning models for ACIS on the Neurofinder competition datasets and report our best model: U-Net2DS, a fully convolutional network that operates on 2D mean summary images. U-Net2DS requires minimal domain-specific pre/post-processing and parameter adjustment, and predictions are made on full 512 × 512 images at ≈9K images per minute. It ranks third in the Neurofinder competition (F:1=0.57) and is the best model to exclusively use deep learning. We also demonstrate useful segmentations on data from outside the competition. The model’s simplicity, speed, and quality results make it a practical choice for ACIS and a strong baseline for more complex models in the future.

AB - Calcium imaging is a technique for observing neuron activity as a series of images showing indicator fluorescence over time. Manually segmenting neurons is time-consuming, leading to research on automated calcium imaging segmentation (ACIS). We evaluated several deep learning models for ACIS on the Neurofinder competition datasets and report our best model: U-Net2DS, a fully convolutional network that operates on 2D mean summary images. U-Net2DS requires minimal domain-specific pre/post-processing and parameter adjustment, and predictions are made on full 512 × 512 images at ≈9K images per minute. It ranks third in the Neurofinder competition (F:1=0.57) and is the best model to exclusively use deep learning. We also demonstrate useful segmentations on data from outside the competition. The model’s simplicity, speed, and quality results make it a practical choice for ACIS and a strong baseline for more complex models in the future.

KW - Calcium imaging

KW - Deep learning

KW - Fully convolutional networks

KW - Microscopy segmentation

UR - http://www.scopus.com/inward/record.url?scp=85029803352&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-67558-9_33

DO - 10.1007/978-3-319-67558-9_33

M3 - Conference contribution

AN - SCOPUS:85029803352

SN - 9783319675572

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

SP - 285

EP - 293

BT - Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support - 3rd International Workshop, DLMIA 2017 and 7th International Workshop, ML-CDS 2017 Held in Conjunction with MICCAI 2017, Proceedings

A2 - Arbel, Tal

A2 - Cardoso, M. Jorge

PB - Springer Verlag

T2 - 3rd International Workshop on Deep Learning in Medical Image Analysis, DLMIA 2017 and 7th International Workshop on Multimodal Learning for Clinical Decision Support, ML-CDS 2017 held in Conjunction with 20th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2017

Y2 - 14 September 2017 through 14 September 2017

ER -

Klibisz A, Rose D, Eicholtz M, Blundon J, Zakharenko S. Fast, simple calcium imaging segmentation with fully convolutional networks. In Arbel T, Cardoso MJ, editors, Deep Learning in Medical Image Analysis and Multimodal Learning for Clinical Decision Support - 3rd International Workshop, DLMIA 2017 and 7th International Workshop, ML-CDS 2017 Held in Conjunction with MICCAI 2017, Proceedings. Springer Verlag. 2017. p. 285-293. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-67558-9_33

Fast, simple calcium imaging segmentation with fully convolutional networks

Abstract

Publication series

Conference

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this