Viable and Necrotic Tumor Assessment from Whole Slide Images of Osteosarcoma Using Machine-Learning and Deep-Learning Models

Simple item page
dc.contributor.ORCID0000-0001-8143-4107 (Arunachalam, HB)
dc.contributor.ORCID0000-0002-0278-4174 (Daescu, O)
dc.contributor.VIAF172508805 (Daescu, O)
dc.contributor.authorArunachalam, Harish Babu
dc.contributor.authorMishra, Rashika
dc.contributor.authorDaescu, Ovidiu
dc.contributor.authorCederberg, K.
dc.contributor.authorRakheja, D.
dc.contributor.authorSengupta, A.
dc.contributor.authorLeonard, D.
dc.contributor.authorHallac, R.
dc.contributor.authorLeavey, P.
dc.contributor.utdAuthorArunachalam, Harish Babu
dc.contributor.utdAuthorMishra, Rashika
dc.contributor.utdAuthorDaescu, Ovidiu
dc.date.accessioned2019-10-31T15:17:32Z
dc.date.available2019-10-31T15:17:32Z
dc.date.created2019-04-17
dc.description.abstractPathological estimation of tumor necrosis after chemotherapy is essential for patients with osteosarcoma. This study reports the first fully automated tool to assess viable and necrotic tumor in osteosarcoma, employing advances in histopathology digitization and automated learning. We selected 40 digitized whole slide images representing the heterogeneity of osteosarcoma and chemotherapy response. With the goal of labeling the diverse regions of the digitized tissue into viable tumor, necrotic tumor, and non-tumor, we trained 13 machine-learning models and selected the top performing one (a Support Vector Machine) based on reported accuracy. We also developed a deep-learning architecture and trained it on the same data set. We computed the receiver-operator characteristic for discrimination of non-tumor from tumor followed by conditional discrimination of necrotic from viable tumor and found our models performing exceptionally well. We then used the trained models to identify regions of interest on image-tiles generated from test whole slide images. The classification output is visualized as a tumor-prediction map, displaying the extent of viable and necrotic tumor in the slide image. Thus, we lay the foundation for a complete tumor assessment pipeline from original histology images to tumor-prediction map generation. The proposed pipeline can also be adopted for other types of tumor. ©2019 The Authors
dc.description.departmentErik Jonsson School of Engineering and Computer Science
dc.description.sponsorshipCPRIT (Cancer Prevention and Research Institute of Texas) award RP150164
dc.identifier.bibliographicCitationArunachalam, H. B., R. Mishra, O. Daescu, K. Cederberg, et al. 2019. "Viable and necrotic tumor assessment from whole slide images of osteosarcoma using machine-learning and deep-learning models." PLOS One 14(4), doi: 10.1371/journal.pone.0210706
dc.identifier.issn1932-6203
dc.identifier.issue4
dc.identifier.urihttps://hdl.handle.net/10735.1/7056
dc.identifier.volume14
dc.language.isoen
dc.publisherPublic Library of Science
dc.relation.urihttps://dx.doi.org/10.1371/journal.pone.0210706
dc.rightsCC BY 4.0 (Attribution)
dc.rights©2019 The Authors
dc.rights.urihttp://creativecommons.org/licenses/by/4.0/
dc.source.journalPLOS One
dc.subjectOsteosarcoma
dc.subjectHistology
dc.subjectPathologists
dc.subjectNecrosis
dc.subjectMagnetic resonance imaging
dc.subjectNeural networks (Computer science)
dc.subjectSupport vector machines
dc.subjectNecrosis
dc.titleViable and Necrotic Tumor Assessment from Whole Slide Images of Osteosarcoma Using Machine-Learning and Deep-Learning Models
dc.type.genrearticle

Files

Original bundle

Now showing 1 - 1 of 1
Thumbnail Image
Name:
JECS-7047-260434.94.pdf
Size:
2.29 MB
Format:
Adobe Portable Document Format
Description:
Article
Download

Collections

Daescu, Ovidiu