Cargando…

Efficient, high-performance semantic segmentation using multi-scale feature extraction

The success of deep learning in recent years has arguably been driven by the availability of large datasets for training powerful predictive algorithms. In medical applications however, the sensitive nature of the data limits the collection and exchange of large-scale datasets. Privacy-preserving an...

Descripción completa

Detalles Bibliográficos
Autores principales: Knolle, Moritz, Kaissis, Georgios, Jungmann, Friederike, Ziegelmayer, Sebastian, Sasse, Daniel, Makowski, Marcus, Rueckert, Daniel, Braren, Rickmer
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Public Library of Science 2021
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8375977/
https://www.ncbi.nlm.nih.gov/pubmed/34411138
http://dx.doi.org/10.1371/journal.pone.0255397
_version_ 1783740410929086464
author Knolle, Moritz
Kaissis, Georgios
Jungmann, Friederike
Ziegelmayer, Sebastian
Sasse, Daniel
Makowski, Marcus
Rueckert, Daniel
Braren, Rickmer
author_facet Knolle, Moritz
Kaissis, Georgios
Jungmann, Friederike
Ziegelmayer, Sebastian
Sasse, Daniel
Makowski, Marcus
Rueckert, Daniel
Braren, Rickmer
author_sort Knolle, Moritz
collection PubMed
description The success of deep learning in recent years has arguably been driven by the availability of large datasets for training powerful predictive algorithms. In medical applications however, the sensitive nature of the data limits the collection and exchange of large-scale datasets. Privacy-preserving and collaborative learning systems can enable the successful application of machine learning in medicine. However, collaborative protocols such as federated learning require the frequent transfer of parameter updates over a network. To enable the deployment of such protocols to a wide range of systems with varying computational performance, efficient deep learning architectures for resource-constrained environments are required. Here we present MoNet, a small, highly optimized neural-network-based segmentation algorithm leveraging efficient multi-scale image features. MoNet is a shallow, U-Net-like architecture based on repeated, dilated convolutions with decreasing dilation rates. We apply and test our architecture on the challenging clinical tasks of pancreatic segmentation in computed tomography (CT) images as well as brain tumor segmentation in magnetic resonance imaging (MRI) data. We assess our model’s segmentation performance and demonstrate that it provides performance on par with compared architectures while providing superior out-of-sample generalization performance, outperforming larger architectures on an independent validation set, while utilizing significantly fewer parameters. We furthermore confirm the suitability of our architecture for federated learning applications by demonstrating a substantial reduction in serialized model storage requirement as a surrogate for network data transfer. Finally, we evaluate MoNet’s inference latency on the central processing unit (CPU) to determine its utility in environments without access to graphics processing units. Our implementation is publicly available as free and open-source software.
format Online
Article
Text
id pubmed-8375977
institution National Center for Biotechnology Information
language English
publishDate 2021
publisher Public Library of Science
record_format MEDLINE/PubMed
spelling pubmed-83759772021-08-20 Efficient, high-performance semantic segmentation using multi-scale feature extraction Knolle, Moritz Kaissis, Georgios Jungmann, Friederike Ziegelmayer, Sebastian Sasse, Daniel Makowski, Marcus Rueckert, Daniel Braren, Rickmer PLoS One Research Article The success of deep learning in recent years has arguably been driven by the availability of large datasets for training powerful predictive algorithms. In medical applications however, the sensitive nature of the data limits the collection and exchange of large-scale datasets. Privacy-preserving and collaborative learning systems can enable the successful application of machine learning in medicine. However, collaborative protocols such as federated learning require the frequent transfer of parameter updates over a network. To enable the deployment of such protocols to a wide range of systems with varying computational performance, efficient deep learning architectures for resource-constrained environments are required. Here we present MoNet, a small, highly optimized neural-network-based segmentation algorithm leveraging efficient multi-scale image features. MoNet is a shallow, U-Net-like architecture based on repeated, dilated convolutions with decreasing dilation rates. We apply and test our architecture on the challenging clinical tasks of pancreatic segmentation in computed tomography (CT) images as well as brain tumor segmentation in magnetic resonance imaging (MRI) data. We assess our model’s segmentation performance and demonstrate that it provides performance on par with compared architectures while providing superior out-of-sample generalization performance, outperforming larger architectures on an independent validation set, while utilizing significantly fewer parameters. We furthermore confirm the suitability of our architecture for federated learning applications by demonstrating a substantial reduction in serialized model storage requirement as a surrogate for network data transfer. Finally, we evaluate MoNet’s inference latency on the central processing unit (CPU) to determine its utility in environments without access to graphics processing units. Our implementation is publicly available as free and open-source software. Public Library of Science 2021-08-19 /pmc/articles/PMC8375977/ /pubmed/34411138 http://dx.doi.org/10.1371/journal.pone.0255397 Text en © 2021 Knolle et al https://creativecommons.org/licenses/by/4.0/This is an open access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/) , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
spellingShingle Research Article
Knolle, Moritz
Kaissis, Georgios
Jungmann, Friederike
Ziegelmayer, Sebastian
Sasse, Daniel
Makowski, Marcus
Rueckert, Daniel
Braren, Rickmer
Efficient, high-performance semantic segmentation using multi-scale feature extraction
title Efficient, high-performance semantic segmentation using multi-scale feature extraction
title_full Efficient, high-performance semantic segmentation using multi-scale feature extraction
title_fullStr Efficient, high-performance semantic segmentation using multi-scale feature extraction
title_full_unstemmed Efficient, high-performance semantic segmentation using multi-scale feature extraction
title_short Efficient, high-performance semantic segmentation using multi-scale feature extraction
title_sort efficient, high-performance semantic segmentation using multi-scale feature extraction
topic Research Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8375977/
https://www.ncbi.nlm.nih.gov/pubmed/34411138
http://dx.doi.org/10.1371/journal.pone.0255397
work_keys_str_mv AT knollemoritz efficienthighperformancesemanticsegmentationusingmultiscalefeatureextraction
AT kaissisgeorgios efficienthighperformancesemanticsegmentationusingmultiscalefeatureextraction
AT jungmannfriederike efficienthighperformancesemanticsegmentationusingmultiscalefeatureextraction
AT ziegelmayersebastian efficienthighperformancesemanticsegmentationusingmultiscalefeatureextraction
AT sassedaniel efficienthighperformancesemanticsegmentationusingmultiscalefeatureextraction
AT makowskimarcus efficienthighperformancesemanticsegmentationusingmultiscalefeatureextraction
AT rueckertdaniel efficienthighperformancesemanticsegmentationusingmultiscalefeatureextraction
AT brarenrickmer efficienthighperformancesemanticsegmentationusingmultiscalefeatureextraction