Deep Learning Models Differentiate Tumor Grades from HE Stained Histology Sections

Mina Khoshdeli; Alexander D Borowsky; Bahram Parvin

doi:10.1109/EMBC.2018.8512357

Deep Learning Models Differentiate Tumor Grades from HE Stained Histology Sections

Mina Khoshdeli, Alexander D Borowsky, Bahram Parvin

Pathology and Laboratory Medicine

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

3 Scopus citations

Abstract

Aberration in tissue architecture is an essential index for cancer diagnosis and tumor grading. Therefore, extracting features of aberrant phenotypes and classification of the histology tissue can provide a model for computer-aided pathology (CAP). As a case study, we investigate the application of convolutional neural networks (CNN)s for tumor grading and decomposing tumor architecture from hematoxylin and eosin (HE) stained histology sections of kidney. The former and latter contribute to CAP and the role of the tumor architecture on the outcome (e.g., survival), respectively. A training set is constructed and sample images are classified into six categories of normal, fat, blood, stroma, low-grade granular tumor, and high-grade clear cell carcinoma. We have compared the performances of a deep versus shallow networks, and shown that the deeper model outperforms the shallow network.

Original language	English (US)
Title of host publication	40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2018
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	620-623
Number of pages	4
Volume	2018-July
ISBN (Electronic)	9781538636466
DOIs	https://doi.org/10.1109/EMBC.2018.8512357
State	Published - Oct 26 2018
Event	40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2018 - Honolulu, United States Duration: Jul 18 2018 → Jul 21 2018

Other

Other	40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2018
Country	United States
City	Honolulu
Period	7/18/18 → 7/21/18

ASJC Scopus subject areas

Signal Processing
Biomedical Engineering
Computer Vision and Pattern Recognition
Health Informatics

Access to Document

10.1109/EMBC.2018.8512357

Fingerprint Dive into the research topics of 'Deep Learning Models Differentiate Tumor Grades from HE Stained Histology Sections'. Together they form a unique fingerprint.

Cite this

Khoshdeli, M., Borowsky, A. D., & Parvin, B. (2018). Deep Learning Models Differentiate Tumor Grades from HE Stained Histology Sections. In 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2018 (Vol. 2018-July, pp. 620-623). [8512357] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/EMBC.2018.8512357

Deep Learning Models Differentiate Tumor Grades from HE Stained Histology Sections. / Khoshdeli, Mina; Borowsky, Alexander D; Parvin, Bahram.

40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2018. Vol. 2018-July Institute of Electrical and Electronics Engineers Inc., 2018. p. 620-623 8512357.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Khoshdeli, M, Borowsky, AD & Parvin, B 2018, Deep Learning Models Differentiate Tumor Grades from HE Stained Histology Sections. in 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2018. vol. 2018-July, 8512357, Institute of Electrical and Electronics Engineers Inc., pp. 620-623, 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2018, Honolulu, United States, 7/18/18. https://doi.org/10.1109/EMBC.2018.8512357

@inproceedings{54682b9bb6104b47846ec87aa63a246f,

title = "Deep Learning Models Differentiate Tumor Grades from HE Stained Histology Sections",

abstract = "Aberration in tissue architecture is an essential index for cancer diagnosis and tumor grading. Therefore, extracting features of aberrant phenotypes and classification of the histology tissue can provide a model for computer-aided pathology (CAP). As a case study, we investigate the application of convolutional neural networks (CNN)s for tumor grading and decomposing tumor architecture from hematoxylin and eosin (HE) stained histology sections of kidney. The former and latter contribute to CAP and the role of the tumor architecture on the outcome (e.g., survival), respectively. A training set is constructed and sample images are classified into six categories of normal, fat, blood, stroma, low-grade granular tumor, and high-grade clear cell carcinoma. We have compared the performances of a deep versus shallow networks, and shown that the deeper model outperforms the shallow network.",

author = "Mina Khoshdeli and Borowsky, {Alexander D} and Bahram Parvin",

year = "2018",

month = oct,

day = "26",

doi = "10.1109/EMBC.2018.8512357",

language = "English (US)",

volume = "2018-July",

pages = "620--623",

booktitle = "40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2018",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

address = "United States",

note = "40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2018 ; Conference date: 18-07-2018 Through 21-07-2018",

}

TY - GEN

T1 - Deep Learning Models Differentiate Tumor Grades from HE Stained Histology Sections

AU - Khoshdeli, Mina

AU - Borowsky, Alexander D

AU - Parvin, Bahram

PY - 2018/10/26

Y1 - 2018/10/26

N2 - Aberration in tissue architecture is an essential index for cancer diagnosis and tumor grading. Therefore, extracting features of aberrant phenotypes and classification of the histology tissue can provide a model for computer-aided pathology (CAP). As a case study, we investigate the application of convolutional neural networks (CNN)s for tumor grading and decomposing tumor architecture from hematoxylin and eosin (HE) stained histology sections of kidney. The former and latter contribute to CAP and the role of the tumor architecture on the outcome (e.g., survival), respectively. A training set is constructed and sample images are classified into six categories of normal, fat, blood, stroma, low-grade granular tumor, and high-grade clear cell carcinoma. We have compared the performances of a deep versus shallow networks, and shown that the deeper model outperforms the shallow network.

AB - Aberration in tissue architecture is an essential index for cancer diagnosis and tumor grading. Therefore, extracting features of aberrant phenotypes and classification of the histology tissue can provide a model for computer-aided pathology (CAP). As a case study, we investigate the application of convolutional neural networks (CNN)s for tumor grading and decomposing tumor architecture from hematoxylin and eosin (HE) stained histology sections of kidney. The former and latter contribute to CAP and the role of the tumor architecture on the outcome (e.g., survival), respectively. A training set is constructed and sample images are classified into six categories of normal, fat, blood, stroma, low-grade granular tumor, and high-grade clear cell carcinoma. We have compared the performances of a deep versus shallow networks, and shown that the deeper model outperforms the shallow network.

UR - http://www.scopus.com/inward/record.url?scp=85056650427&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85056650427&partnerID=8YFLogxK

U2 - 10.1109/EMBC.2018.8512357

DO - 10.1109/EMBC.2018.8512357

M3 - Conference contribution

C2 - 30440473

AN - SCOPUS:85056650427

VL - 2018-July

SP - 620

EP - 623

BT - 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2018

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBC 2018

Y2 - 18 July 2018 through 21 July 2018

ER -