Illustration of the proposed sampling protocol.

<p><i>k</i> random points (green dots) are drawn from the sampling set (blue dots). The subset of sampled data points is defined by the data points that are closest to each drawn point (orange stars). Red squares represent the remaining data points that were not selected.</p>

Saved in:
Bibliographic Details
Main Author: Matheus Viana da Silva (21433572) (author)
Other Authors: Natália de Carvalho Santos (21433575) (author), Julie Ouellette (17041536) (author), Baptiste Lacoste (10536857) (author), Cesar H. Comin (3724186) (author)
Published: 2025
Subjects:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1852019988412497920
author Matheus Viana da Silva (21433572)
author2 Natália de Carvalho Santos (21433575)
Julie Ouellette (17041536)
Baptiste Lacoste (10536857)
Cesar H. Comin (3724186)
author2_role author
author
author
author
author_facet Matheus Viana da Silva (21433572)
Natália de Carvalho Santos (21433575)
Julie Ouellette (17041536)
Baptiste Lacoste (10536857)
Cesar H. Comin (3724186)
author_role author
dc.creator.none.fl_str_mv Matheus Viana da Silva (21433572)
Natália de Carvalho Santos (21433575)
Julie Ouellette (17041536)
Baptiste Lacoste (10536857)
Cesar H. Comin (3724186)
dc.date.none.fl_str_mv 2025-05-27T18:17:33Z
dc.identifier.none.fl_str_mv 10.1371/journal.pone.0322048.g005
dc.relation.none.fl_str_mv https://figshare.com/articles/figure/Illustration_of_the_proposed_sampling_protocol_/29160090
dc.rights.none.fl_str_mv CC BY 4.0
info:eu-repo/semantics/openAccess
dc.subject.none.fl_str_mv Cell Biology
Genetics
Biotechnology
Cancer
Space Science
Biological Sciences not elsewhere classified
Information Systems not elsewhere classified
xlink "> creating
poor inference quality
available metadata information
training sets lead
neural networks trained
datasets traditionally used
selecting relevant samples
image acquisition process
good contrast leads
base dataset using
training samples results
training samples used
neural network
four samples
considering samples
vessmap dataset
usually required
single image
several hours
segmentation algorithms
possible changes
new dataset
might affect
manual annotation
large non
introduce vessmap
intensity variability
imaged tissues
image annotation
highly distinct
generalization capability
even outlier
especially true
distribution shifts
different conditions
dice scores
dice score
demanding task
carefully selected
careful selection
assorted set
dc.title.none.fl_str_mv Illustration of the proposed sampling protocol.
dc.type.none.fl_str_mv Image
Figure
info:eu-repo/semantics/publishedVersion
image
description <p><i>k</i> random points (green dots) are drawn from the sampling set (blue dots). The subset of sampled data points is defined by the data points that are closest to each drawn point (orange stars). Red squares represent the remaining data points that were not selected.</p>
eu_rights_str_mv openAccess
id Manara_052d65a389fa2916c96072e7a4d85c07
identifier_str_mv 10.1371/journal.pone.0322048.g005
network_acronym_str Manara
network_name_str ManaraRepo
oai_identifier_str oai:figshare.com:article/29160090
publishDate 2025
repository.mail.fl_str_mv
repository.name.fl_str_mv
repository_id_str
rights_invalid_str_mv CC BY 4.0
spelling Illustration of the proposed sampling protocol.Matheus Viana da Silva (21433572)Natália de Carvalho Santos (21433575)Julie Ouellette (17041536)Baptiste Lacoste (10536857)Cesar H. Comin (3724186)Cell BiologyGeneticsBiotechnologyCancerSpace ScienceBiological Sciences not elsewhere classifiedInformation Systems not elsewhere classifiedxlink "> creatingpoor inference qualityavailable metadata informationtraining sets leadneural networks traineddatasets traditionally usedselecting relevant samplesimage acquisition processgood contrast leadsbase dataset usingtraining samples resultstraining samples usedneural networkfour samplesconsidering samplesvessmap datasetusually requiredsingle imageseveral hourssegmentation algorithmspossible changesnew datasetmight affectmanual annotationlarge nonintroduce vessmapintensity variabilityimaged tissuesimage annotationhighly distinctgeneralization capabilityeven outlierespecially truedistribution shiftsdifferent conditionsdice scoresdice scoredemanding taskcarefully selectedcareful selectionassorted set<p><i>k</i> random points (green dots) are drawn from the sampling set (blue dots). The subset of sampled data points is defined by the data points that are closest to each drawn point (orange stars). Red squares represent the remaining data points that were not selected.</p>2025-05-27T18:17:33ZImageFigureinfo:eu-repo/semantics/publishedVersionimage10.1371/journal.pone.0322048.g005https://figshare.com/articles/figure/Illustration_of_the_proposed_sampling_protocol_/29160090CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/291600902025-05-27T18:17:33Z
spellingShingle Illustration of the proposed sampling protocol.
Matheus Viana da Silva (21433572)
Cell Biology
Genetics
Biotechnology
Cancer
Space Science
Biological Sciences not elsewhere classified
Information Systems not elsewhere classified
xlink "> creating
poor inference quality
available metadata information
training sets lead
neural networks trained
datasets traditionally used
selecting relevant samples
image acquisition process
good contrast leads
base dataset using
training samples results
training samples used
neural network
four samples
considering samples
vessmap dataset
usually required
single image
several hours
segmentation algorithms
possible changes
new dataset
might affect
manual annotation
large non
introduce vessmap
intensity variability
imaged tissues
image annotation
highly distinct
generalization capability
even outlier
especially true
distribution shifts
different conditions
dice scores
dice score
demanding task
carefully selected
careful selection
assorted set
status_str publishedVersion
title Illustration of the proposed sampling protocol.
title_full Illustration of the proposed sampling protocol.
title_fullStr Illustration of the proposed sampling protocol.
title_full_unstemmed Illustration of the proposed sampling protocol.
title_short Illustration of the proposed sampling protocol.
title_sort Illustration of the proposed sampling protocol.
topic Cell Biology
Genetics
Biotechnology
Cancer
Space Science
Biological Sciences not elsewhere classified
Information Systems not elsewhere classified
xlink "> creating
poor inference quality
available metadata information
training sets lead
neural networks trained
datasets traditionally used
selecting relevant samples
image acquisition process
good contrast leads
base dataset using
training samples results
training samples used
neural network
four samples
considering samples
vessmap dataset
usually required
single image
several hours
segmentation algorithms
possible changes
new dataset
might affect
manual annotation
large non
introduce vessmap
intensity variability
imaged tissues
image annotation
highly distinct
generalization capability
even outlier
especially true
distribution shifts
different conditions
dice scores
dice score
demanding task
carefully selected
careful selection
assorted set