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:
| Main Author: | |
|---|---|
| Other Authors: | , , , |
| 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 |