Context based interpolation of coarse deep learning prediction maps for the segmentation of fine structures in immunofluorescence images

Nicolas Brieu; Christos Gavriel; David James Harrison; Peter David Caie; Guenter Schmidt

doi:10.1117/12.2292794

Context based interpolation of coarse deep learning prediction maps for the segmentation of fine structures in immunofluorescence images

Nicolas Brieu, Christos Gavriel, David James Harrison, Peter David Caie, Guenter Schmidt

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

Abstract

The automatic analysis of digital pathology images is becoming of increasing interest for the development of novel therapeutic drugs and of the associated companion diagnostic tests in oncology. A precise quantification of the tumor microenvironment and therefore an accurate segmentation of the tumor extend are critical in this context. In this paper, we present a new approach based on visual context Random Forest to generate high resolution segmentation maps from Deep Learning coarse segmentation maps. Through an example inimmunofluorescence, we show that this method enables an accurate and fast detection of the tumor structures in terms of qualitative and quantitative evaluation against three baseline approaches. For the method to be resilient to the high variability of staining intensity, a novel locally adaptive normalization algorithm is moreover introduced.

Original language	English
Title of host publication	Medical Imaging 2018
Subtitle of host publication	Digital Pathology
Editors	John E. Tomaszewski, Metin N. Gurcan
Publisher	SPIE
Number of pages	6
DOIs	https://doi.org/10.1117/12.2292794
Publication status	Published - 6 Mar 2018
Event	Symposium: SPIE Medical Imaging : Digital Pathology - Marriott Marquis Houston, Houston, United States Duration: 10 Feb 2018 → 15 Feb 2018 Conference number: 10581 https://spie.org/MI/conferencedetails/digital-pathology

Publication series

Name	Proceedings of SPIE
Publisher	Society of Photo-optical Instrumentation Engineers
Volume	10581
ISSN (Print)	0277-786X
ISSN (Electronic)	1996-756X

Conference

Conference	Symposium: SPIE Medical Imaging
Country/Territory	United States
City	Houston
Period	10/02/18 → 15/02/18
Internet address	https://spie.org/MI/conferencedetails/digital-pathology

Keywords

Digital pathology
Whole slide imaging (WSI)
Immunofluorescence (IF)
Deep learning
Random Forest
Interpolation
Semantic segmentation

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1117/12.2292794

Brieu-2018-Context-based-SPIE-105810P
© 2018, SPIE. This work has been made available online in accordance with the publisher’s policies. This is the final published version of the work, which was originally published at https://doi.org/10.1117/12.2292794
Final published version, 1.12 MB

Cite this

Brieu, N., Gavriel, C., Harrison, D. J., Caie, P. D., & Schmidt, G. (2018). Context based interpolation of coarse deep learning prediction maps for the segmentation of fine structures in immunofluorescence images. In J. E. Tomaszewski, & M. N. Gurcan (Eds.), Medical Imaging 2018: Digital Pathology Article 105810P (Proceedings of SPIE; Vol. 10581). SPIE. https://doi.org/10.1117/12.2292794

@inproceedings{e43dbf8403d34284a6b6763218d2f1c5,

title = "Context based interpolation of coarse deep learning prediction maps for the segmentation of fine structures in immunofluorescence images",

abstract = "The automatic analysis of digital pathology images is becoming of increasing interest for the development of novel therapeutic drugs and of the associated companion diagnostic tests in oncology. A precise quantification of the tumor microenvironment and therefore an accurate segmentation of the tumor extend are critical in this context. In this paper, we present a new approach based on visual context Random Forest to generate high resolution segmentation maps from Deep Learning coarse segmentation maps. Through an example inimmunofluorescence, we show that this method enables an accurate and fast detection of the tumor structures in terms of qualitative and quantitative evaluation against three baseline approaches. For the method to be resilient to the high variability of staining intensity, a novel locally adaptive normalization algorithm is moreover introduced.",

keywords = "Digital pathology, Whole slide imaging (WSI), Immunofluorescence (IF), Deep learning, Random Forest, Interpolation, Semantic segmentation",

author = "Nicolas Brieu and Christos Gavriel and Harrison, {David James} and Caie, {Peter David} and Guenter Schmidt",

year = "2018",

month = mar,

day = "6",

doi = "10.1117/12.2292794",

language = "English",

series = "Proceedings of SPIE",

publisher = "SPIE",

editor = "Tomaszewski, {John E.} and Gurcan, {Metin N.}",

booktitle = "Medical Imaging 2018",

address = "United States",

note = "Symposium: SPIE Medical Imaging : Digital Pathology ; Conference date: 10-02-2018 Through 15-02-2018",

url = "https://spie.org/MI/conferencedetails/digital-pathology",

}

Brieu, N, Gavriel, C, Harrison, DJ, Caie, PD & Schmidt, G 2018, Context based interpolation of coarse deep learning prediction maps for the segmentation of fine structures in immunofluorescence images. in JE Tomaszewski & MN Gurcan (eds), Medical Imaging 2018: Digital Pathology., 105810P, Proceedings of SPIE, vol. 10581, SPIE, Symposium: SPIE Medical Imaging , Houston, Texas, United States, 10/02/18. https://doi.org/10.1117/12.2292794

Context based interpolation of coarse deep learning prediction maps for the segmentation of fine structures in immunofluorescence images. / Brieu, Nicolas; Gavriel, Christos; Harrison, David James et al.
Medical Imaging 2018: Digital Pathology. ed. / John E. Tomaszewski; Metin N. Gurcan. SPIE, 2018. 105810P (Proceedings of SPIE; Vol. 10581).

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

TY - GEN

T1 - Context based interpolation of coarse deep learning prediction maps for the segmentation of fine structures in immunofluorescence images

AU - Brieu, Nicolas

AU - Gavriel, Christos

AU - Harrison, David James

AU - Caie, Peter David

AU - Schmidt, Guenter

N1 - Conference code: 10581

PY - 2018/3/6

Y1 - 2018/3/6

N2 - The automatic analysis of digital pathology images is becoming of increasing interest for the development of novel therapeutic drugs and of the associated companion diagnostic tests in oncology. A precise quantification of the tumor microenvironment and therefore an accurate segmentation of the tumor extend are critical in this context. In this paper, we present a new approach based on visual context Random Forest to generate high resolution segmentation maps from Deep Learning coarse segmentation maps. Through an example inimmunofluorescence, we show that this method enables an accurate and fast detection of the tumor structures in terms of qualitative and quantitative evaluation against three baseline approaches. For the method to be resilient to the high variability of staining intensity, a novel locally adaptive normalization algorithm is moreover introduced.

AB - The automatic analysis of digital pathology images is becoming of increasing interest for the development of novel therapeutic drugs and of the associated companion diagnostic tests in oncology. A precise quantification of the tumor microenvironment and therefore an accurate segmentation of the tumor extend are critical in this context. In this paper, we present a new approach based on visual context Random Forest to generate high resolution segmentation maps from Deep Learning coarse segmentation maps. Through an example inimmunofluorescence, we show that this method enables an accurate and fast detection of the tumor structures in terms of qualitative and quantitative evaluation against three baseline approaches. For the method to be resilient to the high variability of staining intensity, a novel locally adaptive normalization algorithm is moreover introduced.

KW - Digital pathology

KW - Whole slide imaging (WSI)

KW - Immunofluorescence (IF)

KW - Deep learning

KW - Random Forest

KW - Interpolation

KW - Semantic segmentation

U2 - 10.1117/12.2292794

DO - 10.1117/12.2292794

M3 - Conference contribution

T3 - Proceedings of SPIE

BT - Medical Imaging 2018

A2 - Tomaszewski, John E.

A2 - Gurcan, Metin N.

PB - SPIE

T2 - Symposium: SPIE Medical Imaging

Y2 - 10 February 2018 through 15 February 2018

ER -

Context based interpolation of coarse deep learning prediction maps for the segmentation of fine structures in immunofluorescence images

Abstract

Publication series

Conference

Keywords

UN SDGs

Access to Document

Fingerprint

Cite this