Digital Tumor-Collagen Proximity Signature Predicts Survival in Diffuse Large B-Cell Lymphoma

Talha Qaiser; Matthew Pugh; Sandra Margielewska; Robert Hollows; Paul Murray; Nasir Rajpoot

doi:10.1007/978-3-030-23937-4_19

Digital Tumor-Collagen Proximity Signature Predicts Survival in Diffuse Large B-Cell Lymphoma

Talha Qaiser
, Matthew Pugh
, Sandra Margielewska
, Robert Hollows
, Paul Murray
, Nasir Rajpoot

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

Abstract

Diffuse large B-cell lymphoma (DLBCL) is a heterogeneous tumor that originates from normal B-cells. A limited number of studies have investigated the role of acellular stromal microenvironment on outcome in DLBCL. Here, we propose a novel digital proximity signature (DPS) for predicting overall survival (OS) in DLBCL patients. We propose a novel end-to-end multi-task deep learning model for cell detection and classification and investigate the spatial proximity of collagen (type VI) and tumor cells for estimating the DPS. To the best of our knowledge, this is the first study that performs automated analysis of tumor and collagen on DLBCL to identify potential prognostic factors. Experimental results favor our cell classification algorithm over conventional approaches. In addition, our pilot results show that strongly associated tumor-collagen regions are statistically significant (p = 0.03) in predicting OS in DLBCL patients.

Original language	English
Title of host publication	Digital Pathology - 15th European Congress, ECDP 2019, Proceedings
Editors	Constantino Carlos Reyes-Aldasoro, Andrew Janowczyk, Mitko Veta, Peter Bankhead, Korsuk Sirinukunwattana
Publisher	Springer Verlag
Pages	163-171
Number of pages	9
ISBN (Print)	9783030239367
DOIs	https://doi.org/10.1007/978-3-030-23937-4_19
Publication status	Published - 2019
Event	15th European Congress on Digital Pathology, ECDP 2019 - Warwick, United Kingdom Duration: 10 Apr 2019 → 13 Apr 2019

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	11435 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	15th European Congress on Digital Pathology, ECDP 2019
Country/Territory	United Kingdom
City	Warwick
Period	10/04/19 → 13/04/19

UN SDGs

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

SDG 3 Good Health and Well-being

Keywords

Computational pathology
Deep learning
Survival analysis

Access to Document

10.1007/978-3-030-23937-4_19

Cite this

Qaiser, T., Pugh, M., Margielewska, S., Hollows, R., Murray, P., & Rajpoot, N. (2019). Digital Tumor-Collagen Proximity Signature Predicts Survival in Diffuse Large B-Cell Lymphoma. In C. C. Reyes-Aldasoro, A. Janowczyk, M. Veta, P. Bankhead, & K. Sirinukunwattana (Eds.), Digital Pathology - 15th European Congress, ECDP 2019, Proceedings (pp. 163-171). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11435 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-030-23937-4_19

Qaiser, Talha ; Pugh, Matthew ; Margielewska, Sandra et al. / Digital Tumor-Collagen Proximity Signature Predicts Survival in Diffuse Large B-Cell Lymphoma. Digital Pathology - 15th European Congress, ECDP 2019, Proceedings. editor / Constantino Carlos Reyes-Aldasoro ; Andrew Janowczyk ; Mitko Veta ; Peter Bankhead ; Korsuk Sirinukunwattana. Springer Verlag, 2019. pp. 163-171 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{18a36c6f72c341f8b78a0820f580b9c8,

title = "Digital Tumor-Collagen Proximity Signature Predicts Survival in Diffuse Large B-Cell Lymphoma",

abstract = "Diffuse large B-cell lymphoma (DLBCL) is a heterogeneous tumor that originates from normal B-cells. A limited number of studies have investigated the role of acellular stromal microenvironment on outcome in DLBCL. Here, we propose a novel digital proximity signature (DPS) for predicting overall survival (OS) in DLBCL patients. We propose a novel end-to-end multi-task deep learning model for cell detection and classification and investigate the spatial proximity of collagen (type VI) and tumor cells for estimating the DPS. To the best of our knowledge, this is the first study that performs automated analysis of tumor and collagen on DLBCL to identify potential prognostic factors. Experimental results favor our cell classification algorithm over conventional approaches. In addition, our pilot results show that strongly associated tumor-collagen regions are statistically significant (p = 0.03) in predicting OS in DLBCL patients.",

keywords = "Computational pathology, Deep learning, Survival analysis",

author = "Talha Qaiser and Matthew Pugh and Sandra Margielewska and Robert Hollows and Paul Murray and Nasir Rajpoot",

note = "Publisher Copyright: {\textcopyright} 2019, Springer Nature Switzerland AG.; 15th European Congress on Digital Pathology, ECDP 2019 ; Conference date: 10-04-2019 Through 13-04-2019",

year = "2019",

doi = "10.1007/978-3-030-23937-4\_19",

language = "English",

isbn = "9783030239367",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "163--171",

editor = "Reyes-Aldasoro, \{Constantino Carlos\} and Andrew Janowczyk and Mitko Veta and Peter Bankhead and Korsuk Sirinukunwattana",

booktitle = "Digital Pathology - 15th European Congress, ECDP 2019, Proceedings",

}

Qaiser, T, Pugh, M, Margielewska, S, Hollows, R, Murray, P & Rajpoot, N 2019, Digital Tumor-Collagen Proximity Signature Predicts Survival in Diffuse Large B-Cell Lymphoma. in CC Reyes-Aldasoro, A Janowczyk, M Veta, P Bankhead & K Sirinukunwattana (eds), Digital Pathology - 15th European Congress, ECDP 2019, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11435 LNCS, Springer Verlag, pp. 163-171, 15th European Congress on Digital Pathology, ECDP 2019, Warwick, United Kingdom, 10/04/19. https://doi.org/10.1007/978-3-030-23937-4_19

Digital Tumor-Collagen Proximity Signature Predicts Survival in Diffuse Large B-Cell Lymphoma. / Qaiser, Talha; Pugh, Matthew; Margielewska, Sandra et al.
Digital Pathology - 15th European Congress, ECDP 2019, Proceedings. ed. / Constantino Carlos Reyes-Aldasoro; Andrew Janowczyk; Mitko Veta; Peter Bankhead; Korsuk Sirinukunwattana. Springer Verlag, 2019. p. 163-171 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11435 LNCS).

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

TY - GEN

T1 - Digital Tumor-Collagen Proximity Signature Predicts Survival in Diffuse Large B-Cell Lymphoma

AU - Qaiser, Talha

AU - Pugh, Matthew

AU - Margielewska, Sandra

AU - Hollows, Robert

AU - Murray, Paul

AU - Rajpoot, Nasir

PY - 2019

Y1 - 2019

N2 - Diffuse large B-cell lymphoma (DLBCL) is a heterogeneous tumor that originates from normal B-cells. A limited number of studies have investigated the role of acellular stromal microenvironment on outcome in DLBCL. Here, we propose a novel digital proximity signature (DPS) for predicting overall survival (OS) in DLBCL patients. We propose a novel end-to-end multi-task deep learning model for cell detection and classification and investigate the spatial proximity of collagen (type VI) and tumor cells for estimating the DPS. To the best of our knowledge, this is the first study that performs automated analysis of tumor and collagen on DLBCL to identify potential prognostic factors. Experimental results favor our cell classification algorithm over conventional approaches. In addition, our pilot results show that strongly associated tumor-collagen regions are statistically significant (p = 0.03) in predicting OS in DLBCL patients.

AB - Diffuse large B-cell lymphoma (DLBCL) is a heterogeneous tumor that originates from normal B-cells. A limited number of studies have investigated the role of acellular stromal microenvironment on outcome in DLBCL. Here, we propose a novel digital proximity signature (DPS) for predicting overall survival (OS) in DLBCL patients. We propose a novel end-to-end multi-task deep learning model for cell detection and classification and investigate the spatial proximity of collagen (type VI) and tumor cells for estimating the DPS. To the best of our knowledge, this is the first study that performs automated analysis of tumor and collagen on DLBCL to identify potential prognostic factors. Experimental results favor our cell classification algorithm over conventional approaches. In addition, our pilot results show that strongly associated tumor-collagen regions are statistically significant (p = 0.03) in predicting OS in DLBCL patients.

KW - Computational pathology

KW - Deep learning

KW - Survival analysis

UR - https://www.scopus.com/pages/publications/85069185046

U2 - 10.1007/978-3-030-23937-4_19

DO - 10.1007/978-3-030-23937-4_19

M3 - Conference contribution

AN - SCOPUS:85069185046

SN - 9783030239367

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 163

EP - 171

BT - Digital Pathology - 15th European Congress, ECDP 2019, Proceedings

A2 - Reyes-Aldasoro, Constantino Carlos

A2 - Janowczyk, Andrew

A2 - Veta, Mitko

A2 - Bankhead, Peter

A2 - Sirinukunwattana, Korsuk

PB - Springer Verlag

T2 - 15th European Congress on Digital Pathology, ECDP 2019

Y2 - 10 April 2019 through 13 April 2019

ER -

Qaiser T, Pugh M, Margielewska S, Hollows R, Murray P, Rajpoot N. Digital Tumor-Collagen Proximity Signature Predicts Survival in Diffuse Large B-Cell Lymphoma. In Reyes-Aldasoro CC, Janowczyk A, Veta M, Bankhead P, Sirinukunwattana K, editors, Digital Pathology - 15th European Congress, ECDP 2019, Proceedings. Springer Verlag. 2019. p. 163-171. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-23937-4_19

Digital Tumor-Collagen Proximity Signature Predicts Survival in Diffuse Large B-Cell Lymphoma

Abstract

Publication series

Conference

UN SDGs

Keywords

Access to Document

Other files and links

Fingerprint

Cite this