Cargando…
A tool for federated training of segmentation models on whole slide images
The largest bottleneck to the development of convolutional neural network (CNN) models in the computational pathology domain is the collection and curation of diverse training datasets. Training CNNs requires large cohorts of image data, and model generalizability is dependent on training data heter...
Autores principales: | , , , , , , |
---|---|
Formato: | Online Artículo Texto |
Lenguaje: | English |
Publicado: |
Elsevier
2022
|
Materias: | |
Acceso en línea: | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9326476/ https://www.ncbi.nlm.nih.gov/pubmed/35910077 http://dx.doi.org/10.1016/j.jpi.2022.100101 |
_version_ | 1784757295006089216 |
---|---|
author | Lutnick, Brendon Manthey, David Becker, Jan U. Zuckerman, Jonathan E. Rodrigues, Luis Jen, Kuang-Yu Sarder, Pinaki |
author_facet | Lutnick, Brendon Manthey, David Becker, Jan U. Zuckerman, Jonathan E. Rodrigues, Luis Jen, Kuang-Yu Sarder, Pinaki |
author_sort | Lutnick, Brendon |
collection | PubMed |
description | The largest bottleneck to the development of convolutional neural network (CNN) models in the computational pathology domain is the collection and curation of diverse training datasets. Training CNNs requires large cohorts of image data, and model generalizability is dependent on training data heterogeneity. Including data from multiple centers enhances the generalizability of CNN-based models, but this is hindered by the logistical challenges of sharing medical data. In this paper, we explore the feasibility of training our recently developed cloud-based segmentation tool (Histo-Cloud) using federated learning. Using a dataset of renal tissue biopsies we show that federated training to segment interstitial fibrosis and tubular atrophy (IFTA) using datasets from three institutions is not found to be different from a training by pooling the data on one server when tested on a fourth (holdout) institution’s data. Further, training a model to segment glomeruli for a federated dataset (split by staining) demonstrates similar performance. |
format | Online Article Text |
id | pubmed-9326476 |
institution | National Center for Biotechnology Information |
language | English |
publishDate | 2022 |
publisher | Elsevier |
record_format | MEDLINE/PubMed |
spelling | pubmed-93264762022-07-28 A tool for federated training of segmentation models on whole slide images Lutnick, Brendon Manthey, David Becker, Jan U. Zuckerman, Jonathan E. Rodrigues, Luis Jen, Kuang-Yu Sarder, Pinaki J Pathol Inform Technical Note The largest bottleneck to the development of convolutional neural network (CNN) models in the computational pathology domain is the collection and curation of diverse training datasets. Training CNNs requires large cohorts of image data, and model generalizability is dependent on training data heterogeneity. Including data from multiple centers enhances the generalizability of CNN-based models, but this is hindered by the logistical challenges of sharing medical data. In this paper, we explore the feasibility of training our recently developed cloud-based segmentation tool (Histo-Cloud) using federated learning. Using a dataset of renal tissue biopsies we show that federated training to segment interstitial fibrosis and tubular atrophy (IFTA) using datasets from three institutions is not found to be different from a training by pooling the data on one server when tested on a fourth (holdout) institution’s data. Further, training a model to segment glomeruli for a federated dataset (split by staining) demonstrates similar performance. Elsevier 2022-05-21 /pmc/articles/PMC9326476/ /pubmed/35910077 http://dx.doi.org/10.1016/j.jpi.2022.100101 Text en © 2022 The Authors https://creativecommons.org/licenses/by-nc-nd/4.0/This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). |
spellingShingle | Technical Note Lutnick, Brendon Manthey, David Becker, Jan U. Zuckerman, Jonathan E. Rodrigues, Luis Jen, Kuang-Yu Sarder, Pinaki A tool for federated training of segmentation models on whole slide images |
title | A tool for federated training of segmentation models on whole slide images |
title_full | A tool for federated training of segmentation models on whole slide images |
title_fullStr | A tool for federated training of segmentation models on whole slide images |
title_full_unstemmed | A tool for federated training of segmentation models on whole slide images |
title_short | A tool for federated training of segmentation models on whole slide images |
title_sort | tool for federated training of segmentation models on whole slide images |
topic | Technical Note |
url | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9326476/ https://www.ncbi.nlm.nih.gov/pubmed/35910077 http://dx.doi.org/10.1016/j.jpi.2022.100101 |
work_keys_str_mv | AT lutnickbrendon atoolforfederatedtrainingofsegmentationmodelsonwholeslideimages AT mantheydavid atoolforfederatedtrainingofsegmentationmodelsonwholeslideimages AT beckerjanu atoolforfederatedtrainingofsegmentationmodelsonwholeslideimages AT zuckermanjonathane atoolforfederatedtrainingofsegmentationmodelsonwholeslideimages AT rodriguesluis atoolforfederatedtrainingofsegmentationmodelsonwholeslideimages AT jenkuangyu atoolforfederatedtrainingofsegmentationmodelsonwholeslideimages AT sarderpinaki atoolforfederatedtrainingofsegmentationmodelsonwholeslideimages AT lutnickbrendon toolforfederatedtrainingofsegmentationmodelsonwholeslideimages AT mantheydavid toolforfederatedtrainingofsegmentationmodelsonwholeslideimages AT beckerjanu toolforfederatedtrainingofsegmentationmodelsonwholeslideimages AT zuckermanjonathane toolforfederatedtrainingofsegmentationmodelsonwholeslideimages AT rodriguesluis toolforfederatedtrainingofsegmentationmodelsonwholeslideimages AT jenkuangyu toolforfederatedtrainingofsegmentationmodelsonwholeslideimages AT sarderpinaki toolforfederatedtrainingofsegmentationmodelsonwholeslideimages |