The University of Southampton
University of Southampton Institutional Repository

Diagnostic accuracy of a convolutional neural network assessment of solitary pulmonary nodules compared with PET with CT imaging and dynamic contrast-enhanced CT imaging using unenhanced and contrast-enhanced CT imaging

Diagnostic accuracy of a convolutional neural network assessment of solitary pulmonary nodules compared with PET with CT imaging and dynamic contrast-enhanced CT imaging using unenhanced and contrast-enhanced CT imaging
Diagnostic accuracy of a convolutional neural network assessment of solitary pulmonary nodules compared with PET with CT imaging and dynamic contrast-enhanced CT imaging using unenhanced and contrast-enhanced CT imaging

Background: solitary pulmonary nodules (SPNs) measuring 8 to 30 mm in diameter require further workup to determine the likelihood of malignancy. Research Question: What is the diagnostic performance of a lung cancer prediction convolutional neural network (LCP-CNN) in SPNs using unenhanced and contrast-enhanced CT imaging compared with the current clinical workup? 

Study Design and Methods: this was a post hoc analysis of the Single Pulmonary Nodule Investigation: Accuracy and Cost-Effectiveness of Dynamic Contrast Enhanced Computed Tomography in the Characterisation of Solitary Pulmonary Nodules trial, a prospective multicenter study comparing the diagnostic accuracy of dynamic contrast-enhanced (DCE) CT imaging with PET imaging in SPNs. The LCP-CNN was designed and validated in an external cohort. LCP-CNN-generated risk scores were created from the noncontrast and contrast-enhanced CT scan images from the DCE CT imaging. The gold standard was histologic analysis or 2 years of follow-up. The area under the receiver operating characteristic curves (AUC) were calculated using LCP-CNN score, maximum standardized uptake value, and DCE CT scan maximum enhancement and were compared using the DeLong test. 

Results: two hundred seventy participants (mean ± SD age, 68.3 ± 8.8 years; 49% women) underwent PET with CT scan imaging and DCE CT imaging with CT scan data available centrally for LCP-CNN analysis. The accuracy of the LCP-CNN on the noncontrast images (AUC, 0.83; 95% CI, 0.79-0.88) was superior to that of DCE CT imaging (AUC, 0.76; 95% CI, 0.69-0.82; P = .03) and equal to that of PET with CT scan imaging (AUC, 0.86; 95% CI, 0.81-0.90; P = .35). The presence of contrast resulted in a small reduction in diagnostic accuracy, with the AUC falling from 0.83 (95% CI, 0.79-0.88) on the noncontrast images to 0.80 to 0.83 after contrast (P < .05 for 240 s after contrast only). 

Interpretation: an LCP-CNN algorithm provides an AUC equivalent to PET with CT scan imaging in the diagnosis of solitary pulmonary nodules. 

Trial Registration: ClinicalTrials.gov Identifier; No.: NCT02013063

Solitary Pulmonary Nodule, TOMOGRAPHY, computed tomography, diagnostic test accuracy, machine learning, positron emission tomoography, x-ray computed
0012-3692
444-454
Weir-McCall, Jonathan R.
4f1bbde3-9448-46c9-a03a-7900d2787d4d
Debruyn, Elise
3d2c5b4d-d4c2-4192-b3f6-b87719f96bbc
Harris, Scott
19ea097b-df15-4f0f-be19-8ac42c190028
Qureshi, Nagmi R.
269a0e9b-9716-47a1-95db-926991b47bf5
Rintoul, Robert C.
6bc4015b-14ca-4d33-8d03-12311fafa94f
Gleeson, Fergus V.
8650cbb5-04a4-446b-829e-6d58d399d089
Gilbert, Fiona J.
d0021270-f0b9-48ce-a68b-d7643418b8fc
Brindle, Anindo Banerjee Lucy
309e5706-b823-42c7-8900-4dadb308b090
Callister, Matthew
0a1d4b62-4f9e-4375-885d-ce7bdba6d73e
Clegg, Andrew
c9dd5894-8e1c-4a09-9a4c-e88e142ddf2e
Cook, Andrew
ab9c7bb3-974a-4db9-b3c2-9942988005d5
Cozens, Kelly
ad334c90-643c-4e72-a957-7d3d15a938ce
Crosbie, Philip
be9e8533-893d-4370-8d65-59fd6f890e5e
Dizdarevic, Sabina
05112ee6-9888-4c6c-bc00-0936ea3ab61a
Eaton, Rosemary
a22211e5-0dfd-4e7c-a8a8-ca701daef04a
Eichhorst, Kathrin
ee84619d-a332-4390-80ca-bd41beded026
Frew, Anthony
d763283a-879b-44ce-8d8e-c18e5868cc12
Groves, Ashley
40b2d7cc-5bc7-4210-90e0-e07070879de1
Han, Sai
dc569400-0635-4b27-8cfb-b38c334e96a2
Jones, Jeremy
270b303b-6bad-4be7-8ea0-63d0e8015c91
Kankam, Osie
5da30939-2745-4356-a6c8-a4dd9fa0d06e
Karunasaagarar, Kavitasagary
a8f39c0e-2d9d-4a08-bd2f-bf76a1a33e5f
Kurban, Lutfi
d2efc878-e91d-4e33-bb23-ea70b72abdd0
Little, Louisa
41f55532-898e-41a6-97c6-2802a379d684
Madden, Jackie
6aeabad8-f525-4368-b459-1f54ec394c6b
McClement, Chris
f44a15d3-7cc3-42bd-bbe3-d56e179a8f14
Miles, Ken
6bb0e8a9-ebcc-4792-bc74-de61dab2f467
Moate, Patricia
68a93f71-aee5-4406-a321-e1fde2fe791a
Peebles, Charles
2f68f796-caf1-4b5b-993f-611e90648988
Pike, Lucy
693e5934-a3c3-4c22-9477-17d409a4c632
Poon, Fat-Wui
0653650f-26d6-4c5a-9fa9-2ca3508a4f07
Sinclair, Donald
fd805c59-86a6-4e78-a293-844ff5d0fc7a
Shah, Andrew
8ddd4aa8-8831-4afa-94ff-61ec713998b8
Vale, Luke
cfda2de4-2f8d-4977-bc06-83c413b9e773
George, Steve
6ed0817d-dadd-44a1-9ab7-6a13c45ab306
Riley, Richard
10997621-4f9e-43cd-95d9-cd40cc8ff39c
Lodge, Andrea
55a6689d-a536-4c60-8d02-f04c76f7be95
Buscombe, John
2f33e916-900e-43ff-8549-e991193b5ff1
Green, Theresa
54ff353e-6a2a-4166-bdb3-ebe62efb298a
Stone, Amanda
880b9b2f-c034-43ed-ac6d-1098224f256b
Navani, Neal
905198d0-88ec-472d-8505-83d30c68ca14
Shortman, Robert
55d5f577-34f8-47f7-8ced-f732a30e4e94
Azzopardi, Gabriella
f87bf9b1-aefe-4cad-a9d2-205e43bcfbbc
Doffman, Sarah
37392e1a-9419-4a7b-bdc5-95a018bfa5cb
Bush, Janice
d6b61436-d71a-426b-910c-64ab0b3fa1f2
Lyttle, Jane
5f5c0b3e-0488-407b-a10d-c645bd26d7fb
Jacob, Kenneth
b9e9edca-e3aa-41a8-acf5-6e296606f72a
Horst, Joris van der
1252bb7a-a28b-42e1-a929-39d8e5bc7c01
Sarvesvaran, Joseph
37ad4508-86d5-41a3-a522-b6937d6f7645
McLaren, Barbara
85bd9591-5628-41e8-b4e3-df22ae551e95
Gomersall, Lesley
163038d8-68d4-4f27-980f-7316e0f0b52e
Weir-McCall, Jonathan R.
4f1bbde3-9448-46c9-a03a-7900d2787d4d
Debruyn, Elise
3d2c5b4d-d4c2-4192-b3f6-b87719f96bbc
Harris, Scott
19ea097b-df15-4f0f-be19-8ac42c190028
Qureshi, Nagmi R.
269a0e9b-9716-47a1-95db-926991b47bf5
Rintoul, Robert C.
6bc4015b-14ca-4d33-8d03-12311fafa94f
Gleeson, Fergus V.
8650cbb5-04a4-446b-829e-6d58d399d089
Gilbert, Fiona J.
d0021270-f0b9-48ce-a68b-d7643418b8fc
Brindle, Anindo Banerjee Lucy
309e5706-b823-42c7-8900-4dadb308b090
Callister, Matthew
0a1d4b62-4f9e-4375-885d-ce7bdba6d73e
Clegg, Andrew
c9dd5894-8e1c-4a09-9a4c-e88e142ddf2e
Cook, Andrew
ab9c7bb3-974a-4db9-b3c2-9942988005d5
Cozens, Kelly
ad334c90-643c-4e72-a957-7d3d15a938ce
Crosbie, Philip
be9e8533-893d-4370-8d65-59fd6f890e5e
Dizdarevic, Sabina
05112ee6-9888-4c6c-bc00-0936ea3ab61a
Eaton, Rosemary
a22211e5-0dfd-4e7c-a8a8-ca701daef04a
Eichhorst, Kathrin
ee84619d-a332-4390-80ca-bd41beded026
Frew, Anthony
d763283a-879b-44ce-8d8e-c18e5868cc12
Groves, Ashley
40b2d7cc-5bc7-4210-90e0-e07070879de1
Han, Sai
dc569400-0635-4b27-8cfb-b38c334e96a2
Jones, Jeremy
270b303b-6bad-4be7-8ea0-63d0e8015c91
Kankam, Osie
5da30939-2745-4356-a6c8-a4dd9fa0d06e
Karunasaagarar, Kavitasagary
a8f39c0e-2d9d-4a08-bd2f-bf76a1a33e5f
Kurban, Lutfi
d2efc878-e91d-4e33-bb23-ea70b72abdd0
Little, Louisa
41f55532-898e-41a6-97c6-2802a379d684
Madden, Jackie
6aeabad8-f525-4368-b459-1f54ec394c6b
McClement, Chris
f44a15d3-7cc3-42bd-bbe3-d56e179a8f14
Miles, Ken
6bb0e8a9-ebcc-4792-bc74-de61dab2f467
Moate, Patricia
68a93f71-aee5-4406-a321-e1fde2fe791a
Peebles, Charles
2f68f796-caf1-4b5b-993f-611e90648988
Pike, Lucy
693e5934-a3c3-4c22-9477-17d409a4c632
Poon, Fat-Wui
0653650f-26d6-4c5a-9fa9-2ca3508a4f07
Sinclair, Donald
fd805c59-86a6-4e78-a293-844ff5d0fc7a
Shah, Andrew
8ddd4aa8-8831-4afa-94ff-61ec713998b8
Vale, Luke
cfda2de4-2f8d-4977-bc06-83c413b9e773
George, Steve
6ed0817d-dadd-44a1-9ab7-6a13c45ab306
Riley, Richard
10997621-4f9e-43cd-95d9-cd40cc8ff39c
Lodge, Andrea
55a6689d-a536-4c60-8d02-f04c76f7be95
Buscombe, John
2f33e916-900e-43ff-8549-e991193b5ff1
Green, Theresa
54ff353e-6a2a-4166-bdb3-ebe62efb298a
Stone, Amanda
880b9b2f-c034-43ed-ac6d-1098224f256b
Navani, Neal
905198d0-88ec-472d-8505-83d30c68ca14
Shortman, Robert
55d5f577-34f8-47f7-8ced-f732a30e4e94
Azzopardi, Gabriella
f87bf9b1-aefe-4cad-a9d2-205e43bcfbbc
Doffman, Sarah
37392e1a-9419-4a7b-bdc5-95a018bfa5cb
Bush, Janice
d6b61436-d71a-426b-910c-64ab0b3fa1f2
Lyttle, Jane
5f5c0b3e-0488-407b-a10d-c645bd26d7fb
Jacob, Kenneth
b9e9edca-e3aa-41a8-acf5-6e296606f72a
Horst, Joris van der
1252bb7a-a28b-42e1-a929-39d8e5bc7c01
Sarvesvaran, Joseph
37ad4508-86d5-41a3-a522-b6937d6f7645
McLaren, Barbara
85bd9591-5628-41e8-b4e3-df22ae551e95
Gomersall, Lesley
163038d8-68d4-4f27-980f-7316e0f0b52e

Weir-McCall, Jonathan R., Debruyn, Elise, Harris, Scott, Qureshi, Nagmi R., Rintoul, Robert C., Gleeson, Fergus V., Gilbert, Fiona J., Brindle, Anindo Banerjee Lucy, Callister, Matthew, Clegg, Andrew, Cook, Andrew, Cozens, Kelly, Crosbie, Philip, Dizdarevic, Sabina, Eaton, Rosemary, Eichhorst, Kathrin, Frew, Anthony, Groves, Ashley, Han, Sai, Jones, Jeremy, Kankam, Osie, Karunasaagarar, Kavitasagary, Kurban, Lutfi, Little, Louisa, Madden, Jackie, McClement, Chris, Miles, Ken, Moate, Patricia, Peebles, Charles, Pike, Lucy, Poon, Fat-Wui, Sinclair, Donald, Shah, Andrew, Vale, Luke, George, Steve, Riley, Richard, Lodge, Andrea, Buscombe, John, Green, Theresa, Stone, Amanda, Navani, Neal, Shortman, Robert, Azzopardi, Gabriella, Doffman, Sarah, Bush, Janice, Lyttle, Jane, Jacob, Kenneth, Horst, Joris van der, Sarvesvaran, Joseph, McLaren, Barbara and Gomersall, Lesley (2023) Diagnostic accuracy of a convolutional neural network assessment of solitary pulmonary nodules compared with PET with CT imaging and dynamic contrast-enhanced CT imaging using unenhanced and contrast-enhanced CT imaging. Chest, 163 (2), 444-454. (doi:10.1016/j.chest.2022.08.2227).

Record type: Article

Abstract

Background: solitary pulmonary nodules (SPNs) measuring 8 to 30 mm in diameter require further workup to determine the likelihood of malignancy. Research Question: What is the diagnostic performance of a lung cancer prediction convolutional neural network (LCP-CNN) in SPNs using unenhanced and contrast-enhanced CT imaging compared with the current clinical workup? 

Study Design and Methods: this was a post hoc analysis of the Single Pulmonary Nodule Investigation: Accuracy and Cost-Effectiveness of Dynamic Contrast Enhanced Computed Tomography in the Characterisation of Solitary Pulmonary Nodules trial, a prospective multicenter study comparing the diagnostic accuracy of dynamic contrast-enhanced (DCE) CT imaging with PET imaging in SPNs. The LCP-CNN was designed and validated in an external cohort. LCP-CNN-generated risk scores were created from the noncontrast and contrast-enhanced CT scan images from the DCE CT imaging. The gold standard was histologic analysis or 2 years of follow-up. The area under the receiver operating characteristic curves (AUC) were calculated using LCP-CNN score, maximum standardized uptake value, and DCE CT scan maximum enhancement and were compared using the DeLong test. 

Results: two hundred seventy participants (mean ± SD age, 68.3 ± 8.8 years; 49% women) underwent PET with CT scan imaging and DCE CT imaging with CT scan data available centrally for LCP-CNN analysis. The accuracy of the LCP-CNN on the noncontrast images (AUC, 0.83; 95% CI, 0.79-0.88) was superior to that of DCE CT imaging (AUC, 0.76; 95% CI, 0.69-0.82; P = .03) and equal to that of PET with CT scan imaging (AUC, 0.86; 95% CI, 0.81-0.90; P = .35). The presence of contrast resulted in a small reduction in diagnostic accuracy, with the AUC falling from 0.83 (95% CI, 0.79-0.88) on the noncontrast images to 0.80 to 0.83 after contrast (P < .05 for 240 s after contrast only). 

Interpretation: an LCP-CNN algorithm provides an AUC equivalent to PET with CT scan imaging in the diagnosis of solitary pulmonary nodules. 

Trial Registration: ClinicalTrials.gov Identifier; No.: NCT02013063

This record has no associated files available for download.

More information

e-pub ahead of print date: 8 September 2022
Published date: 7 February 2023
Additional Information: Funding Information: Author contributions: J. R. W.-M. was involved in the design of the study, delivery of the study, interpretation of the results, and writing of the report. F. J. G. was involved in the design of the study, delivery of the study, interpretation of the results, and writing of the report. S. H. was involved in the design of the study, delivery of the study, statistical analysis, interpretation of the results, and writing of the report. E. D. was involved in the delivery of the study and interpretation of the results. N. R. Q. was involved in the design of the study, delivery of the study, and interpretation of the results. R. C. R. was involved in the design of the study, delivery of the study, and interpretation of the results. F. V. G. was involved in the design of the study and was responsible for recruiting participants. All authors reviewed the final report. Role of sponsors: The views expressed are those of the authors and not necessarily those of the NIHR or the Department of Health and Social Care. Beyond the provision of and training in the software, no Optellum employees were involved the study's design, conduct, data analysis or interpretation. ∗ SPUtNIk Investigators: Anindo Banerjee, Lucy Brindle, Matthew Callister, Andrew Clegg, Andrew Cook, Kelly Cozens, Philip Crosbie, Sabina Dizdarevic, Rosemary Eaton, Kathrin Eichhorst, Anthony Frew, Ashley Groves, Sai Han, Jeremy Jones, Osie Kankam, Kavitasagary Karunasaagarar, Lutfi Kurban, Louisa Little, Jackie Madden, Chris McClement, Ken Miles, Patricia Moate, Charles Peebles, Lucy Pike, Fat-Wui Poon, Donald Sinclair, Andrew Shah, Luke Vale, Steve George, Richard Riley, Andrea Lodge, John Buscombe, Theresa Green, Amanda Stone, Neal Navani, Robert Shortman, Gabriella Azzopardi, Sarah Doffman, Janice Bush, Jane Lyttle, Kenneth Jacob, Joris van der Horst, Joseph Sarvesvaran, Barbara McLaren, Lesley Gomersall, Ravi Sharma, Kathleen Collie, Steve O'Hickey, Jayne Tyler, Sue King, John O'Brien, Rajiv Srivastava, Hugh Lloyd-Jones, Sandra Beech, Andrew Scarsbrook, Victoria Ashford-Turner, Elaine Smith, Susan Mbale, Nick Adams, and Gail Pottinger. Data sharing: Individual participant data from the SPUtNIk trial will be made available, including data dictionaries, for approved data sharing requests. Individual participant data will be shared that underlie the results reported in this article, after de-identification and normalization of information (text, tables, figures, and appendixes). The study protocol and statistical analysis plan also will be available. Anonymous data will be available for request from 3 months after publication of the article to researchers who provide a completed data sharing request form that describes a methodologically sound proposal for the purpose of the approved proposal and, if appropriate, have signed a data sharing agreement. Data will be shared after all parties have signed relevant data sharing documentation, covering Southampton Clinical Trials Unit conditions for sharing and, if required, an additional data sharing agreement from the sponsor. Proposals should be directed to ctu@soton.ac.uk. Additional information: The e-Figures and e-Table are available online under “Supplementary Data.” Funding Information: The trial was funded by the National Institute for Health and Care Research Health Technology Assessment Program [Grant 09/22/117] and is being run by Southampton Clinical Trials Unit, which is funded in part by Cancer Research United Kingdom. F. J. G. is an NIHR Senior Investigator. R. C. R. is funded in part by the NIHR Cambridge Biomedical Research Centre, Cancer Research UK Cambridge Centre [Grant BRC-1215-20014], and the Cancer Research Network: Eastern. N. R. Q. is funded in part by the Cambridge Biomedical Research Centre. This research was supported by the NIHR Cambridge Biomedical Research Centre [Grant BRC-1215-20014]. Access to software to generate LCP-CNN scores was provided by Optellum Ltd. Publisher Copyright: © 2022 American College of Chest Physicians
Keywords: Solitary Pulmonary Nodule, TOMOGRAPHY, computed tomography, diagnostic test accuracy, machine learning, positron emission tomoography, x-ray computed

Identifiers

Local EPrints ID: 480093
URI: http://eprints.soton.ac.uk/id/eprint/480093
ISSN: 0012-3692
PURE UUID: 24aef6fa-d28a-481b-83eb-d5dfe77cbd18
ORCID for Andrew Cook: ORCID iD orcid.org/0000-0002-6680-439X
ORCID for Kelly Cozens: ORCID iD orcid.org/0000-0001-9592-9100

Catalogue record

Date deposited: 01 Aug 2023 16:47
Last modified: 18 Mar 2024 03:05

Export record

Altmetrics

Contributors

Author: Jonathan R. Weir-McCall
Author: Elise Debruyn
Author: Scott Harris
Author: Nagmi R. Qureshi
Author: Robert C. Rintoul
Author: Fergus V. Gleeson
Author: Fiona J. Gilbert
Author: Anindo Banerjee Lucy Brindle
Author: Matthew Callister
Author: Andrew Clegg
Author: Andrew Cook ORCID iD
Author: Kelly Cozens ORCID iD
Author: Philip Crosbie
Author: Sabina Dizdarevic
Author: Rosemary Eaton
Author: Kathrin Eichhorst
Author: Anthony Frew
Author: Ashley Groves
Author: Sai Han
Author: Jeremy Jones
Author: Osie Kankam
Author: Kavitasagary Karunasaagarar
Author: Lutfi Kurban
Author: Louisa Little
Author: Jackie Madden
Author: Chris McClement
Author: Ken Miles
Author: Patricia Moate
Author: Charles Peebles
Author: Lucy Pike
Author: Fat-Wui Poon
Author: Donald Sinclair
Author: Andrew Shah
Author: Luke Vale
Author: Steve George
Author: Richard Riley
Author: Andrea Lodge
Author: John Buscombe
Author: Theresa Green
Author: Amanda Stone
Author: Neal Navani
Author: Robert Shortman
Author: Gabriella Azzopardi
Author: Sarah Doffman
Author: Janice Bush
Author: Jane Lyttle
Author: Kenneth Jacob
Author: Joris van der Horst
Author: Joseph Sarvesvaran
Author: Barbara McLaren
Author: Lesley Gomersall

Download statistics

Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.

View more statistics

Atom RSS 1.0 RSS 2.0

Contact ePrints Soton: eprints@soton.ac.uk

ePrints Soton supports OAI 2.0 with a base URL of http://eprints.soton.ac.uk/cgi/oai2

This repository has been built using EPrints software, developed at the University of Southampton, but available to everyone to use.

We use cookies to ensure that we give you the best experience on our website. If you continue without changing your settings, we will assume that you are happy to receive cookies on the University of Southampton website.

×