Towards Detecting High-Uptake Lesions from Lung CT Scans Using Deep Learning

Krzysztof Pawełczyk; Michal Kawulok; Jakub Nalepa; Michael P. Hayball; Sarah J. McQuaid; Vineet Prakash; Balaji Ganeshan

doi:10.1007/978-3-319-68548-9_29

Towards Detecting High-Uptake Lesions from Lung CT Scans Using Deep Learning

Krzysztof Pawełczyk
, Michal Kawulok
, Jakub Nalepa
, Michael P. Hayball
, Sarah J. McQuaid
, Vineet Prakash
, Balaji Ganeshan

Department of Algorithmics and Software (RAU/RAU5)

Future Processing
Silesian University of Technology
Feedback PLC
Royal Surrey County Hospital NHS Foundation Trust
Ashford and St Peter's Hospitals NHS Foundation Trust
University College London

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

13 Citations (Scopus)

Abstract

Automatic detection of lung lesions from computed tomography (CT) and positron emission tomography (PET) is an important task in lung cancer diagnosis. While CT scans make it possible to retrieve structural information, PET images reveal the functional aspects of the tissue, hence combined PET/CT imagery allows for detecting metabolically active lesions. In this paper, we explore how to exploit deep convolutional neural networks to identify the active tumour tissue exclusively from CT scans, which, to the best of our knowledge, has not been attempted yet. Our experimental results are very encouraging and they clearly indicate the possibility of detecting lesions with high glucose uptake, which could increase the utility of CT in lung cancer diagnosis.

Original language	English
Title of host publication	Image Analysis and Processing - ICIAP 2017 - 19th International Conference, Proceedings
Editors	Sebastiano Battiato, Giovanni Gallo, Filippo Stanco, Raimondo Schettini
Publisher	Springer Verlag
Pages	310-320
Number of pages	11
ISBN (Print)	9783319685472
DOIs	https://doi.org/10.1007/978-3-319-68548-9_29
Publication status	Published - 2017
Event	19th International Conference on Image Analysis and Processing, ICIAP 2017 - Catania, Italy Duration: 11 Sept 2017 → 15 Sept 2017

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	10485 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	19th International Conference on Image Analysis and Processing, ICIAP 2017
Country/Territory	Italy
City	Catania
Period	11/09/17 → 15/09/17

UN SDGs

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

SDG 3 Good Health and Well-being

Keywords

Deep neural networks
Lesion detection
PET/CT imaging

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-319-68548-9_29

Cite this

Pawełczyk, K., Kawulok, M., Nalepa, J., Hayball, M. P., McQuaid, S. J., Prakash, V., & Ganeshan, B. (2017). Towards Detecting High-Uptake Lesions from Lung CT Scans Using Deep Learning. In S. Battiato, G. Gallo, F. Stanco, & R. Schettini (Eds.), Image Analysis and Processing - ICIAP 2017 - 19th International Conference, Proceedings (pp. 310-320). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10485 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-68548-9_29

Pawełczyk, Krzysztof ; Kawulok, Michal ; Nalepa, Jakub et al. / Towards Detecting High-Uptake Lesions from Lung CT Scans Using Deep Learning. Image Analysis and Processing - ICIAP 2017 - 19th International Conference, Proceedings. editor / Sebastiano Battiato ; Giovanni Gallo ; Filippo Stanco ; Raimondo Schettini. Springer Verlag, 2017. pp. 310-320 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{215df3698ab948fba34e47b5f95b0af7,

title = "Towards Detecting High-Uptake Lesions from Lung CT Scans Using Deep Learning",

abstract = "Automatic detection of lung lesions from computed tomography (CT) and positron emission tomography (PET) is an important task in lung cancer diagnosis. While CT scans make it possible to retrieve structural information, PET images reveal the functional aspects of the tissue, hence combined PET/CT imagery allows for detecting metabolically active lesions. In this paper, we explore how to exploit deep convolutional neural networks to identify the active tumour tissue exclusively from CT scans, which, to the best of our knowledge, has not been attempted yet. Our experimental results are very encouraging and they clearly indicate the possibility of detecting lesions with high glucose uptake, which could increase the utility of CT in lung cancer diagnosis.",

keywords = "Deep neural networks, Lesion detection, PET/CT imaging",

author = "Krzysztof Pawe{\l}czyk and Michal Kawulok and Jakub Nalepa and Hayball, \{Michael P.\} and McQuaid, \{Sarah J.\} and Vineet Prakash and Balaji Ganeshan",

note = "Publisher Copyright: {\textcopyright} 2017, Springer International Publishing AG.; 19th International Conference on Image Analysis and Processing, ICIAP 2017 ; Conference date: 11-09-2017 Through 15-09-2017",

year = "2017",

doi = "10.1007/978-3-319-68548-9\_29",

language = "English",

isbn = "9783319685472",

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

publisher = "Springer Verlag",

pages = "310--320",

editor = "Sebastiano Battiato and Giovanni Gallo and Filippo Stanco and Raimondo Schettini",

booktitle = "Image Analysis and Processing - ICIAP 2017 - 19th International Conference, Proceedings",

address = "Germany",

}

Pawełczyk, K, Kawulok, M , Nalepa, J, Hayball, MP, McQuaid, SJ, Prakash, V & Ganeshan, B 2017, Towards Detecting High-Uptake Lesions from Lung CT Scans Using Deep Learning. in S Battiato, G Gallo, F Stanco & R Schettini (eds), Image Analysis and Processing - ICIAP 2017 - 19th International Conference, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10485 LNCS, Springer Verlag, pp. 310-320, 19th International Conference on Image Analysis and Processing, ICIAP 2017, Catania, Italy, 11/09/17. https://doi.org/10.1007/978-3-319-68548-9_29

Towards Detecting High-Uptake Lesions from Lung CT Scans Using Deep Learning. / Pawełczyk, Krzysztof; Kawulok, Michal ; Nalepa, Jakub et al.
Image Analysis and Processing - ICIAP 2017 - 19th International Conference, Proceedings. ed. / Sebastiano Battiato; Giovanni Gallo; Filippo Stanco; Raimondo Schettini. Springer Verlag, 2017. p. 310-320 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10485 LNCS).

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

TY - GEN

T1 - Towards Detecting High-Uptake Lesions from Lung CT Scans Using Deep Learning

AU - Pawełczyk, Krzysztof

AU - Kawulok, Michal

AU - Nalepa, Jakub

AU - Hayball, Michael P.

AU - McQuaid, Sarah J.

AU - Prakash, Vineet

AU - Ganeshan, Balaji

PY - 2017

Y1 - 2017

N2 - Automatic detection of lung lesions from computed tomography (CT) and positron emission tomography (PET) is an important task in lung cancer diagnosis. While CT scans make it possible to retrieve structural information, PET images reveal the functional aspects of the tissue, hence combined PET/CT imagery allows for detecting metabolically active lesions. In this paper, we explore how to exploit deep convolutional neural networks to identify the active tumour tissue exclusively from CT scans, which, to the best of our knowledge, has not been attempted yet. Our experimental results are very encouraging and they clearly indicate the possibility of detecting lesions with high glucose uptake, which could increase the utility of CT in lung cancer diagnosis.

AB - Automatic detection of lung lesions from computed tomography (CT) and positron emission tomography (PET) is an important task in lung cancer diagnosis. While CT scans make it possible to retrieve structural information, PET images reveal the functional aspects of the tissue, hence combined PET/CT imagery allows for detecting metabolically active lesions. In this paper, we explore how to exploit deep convolutional neural networks to identify the active tumour tissue exclusively from CT scans, which, to the best of our knowledge, has not been attempted yet. Our experimental results are very encouraging and they clearly indicate the possibility of detecting lesions with high glucose uptake, which could increase the utility of CT in lung cancer diagnosis.

KW - Deep neural networks

KW - Lesion detection

KW - PET/CT imaging

UR - https://www.scopus.com/pages/publications/85032503071

U2 - 10.1007/978-3-319-68548-9_29

DO - 10.1007/978-3-319-68548-9_29

M3 - Conference contribution

AN - SCOPUS:85032503071

SN - 9783319685472

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

SP - 310

EP - 320

BT - Image Analysis and Processing - ICIAP 2017 - 19th International Conference, Proceedings

A2 - Battiato, Sebastiano

A2 - Gallo, Giovanni

A2 - Stanco, Filippo

A2 - Schettini, Raimondo

PB - Springer Verlag

T2 - 19th International Conference on Image Analysis and Processing, ICIAP 2017

Y2 - 11 September 2017 through 15 September 2017

ER -

Pawełczyk K, Kawulok M , Nalepa J, Hayball MP, McQuaid SJ, Prakash V et al. Towards Detecting High-Uptake Lesions from Lung CT Scans Using Deep Learning. In Battiato S, Gallo G, Stanco F, Schettini R, editors, Image Analysis and Processing - ICIAP 2017 - 19th International Conference, Proceedings. Springer Verlag. 2017. p. 310-320. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-68548-9_29

Towards Detecting High-Uptake Lesions from Lung CT Scans Using Deep Learning

Abstract

Publication series

Conference

UN SDGs

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this