Automatic Generation of Interpretable Lung Cancer Scoring Models from Chest X-Ray Images

Horry, MJ; Chakraborty, S; Pradhan, B; Paul, M; Gomes, DPS; Ul-Haq, A

Automatic Generation of Interpretable Lung Cancer Scoring Models from Chest X-Ray Images

Horry, MJ Chakraborty, S

Pradhan, B Paul, M Gomes, DPS Ul-Haq, A

Permalink

Publication Type:: Journal Article

Open Access

Copyright Clearance Process

Recently Added
In Progress
Open Access

This item is open access.

Adobe PDF

Download full textAdobe PDF (769.16 kB)

View statistics

Full metadata record

Field	Value	Language
dc.contributor.author	Horry, MJ
dc.contributor.author	Chakraborty, S https://orcid.org/0000-0002-0102-5424
dc.contributor.author	Pradhan, B
dc.contributor.author	Paul, M
dc.contributor.author	Gomes, DPS
dc.contributor.author	Ul-Haq, A
dc.date.accessioned	2021-02-07T22:52:38Z
dc.date.available	2021-02-07T22:52:38Z
dc.identifier.uri	http://hdl.handle.net/10453/145908
dc.description.abstract	Lung cancer is the leading cause of cancer death worldwide with early detection being the key to a positive patient prognosis. Although a multitude of studies have demonstrated that machine learning, and particularly deep learning, techniques are effective at automatically diagnosing lung cancer, these techniques have yet to be clinically approved and adopted by the medical community. Most research in this field is focused on the narrow task of nodule detection to provide an artificial radiological second reading. We instead focus on extracting, from chest X-ray images, a wider range of pathologies associated with lung cancer using a computer vision model trained on a large dataset. We then find the set of best fit decision trees against an independent, smaller dataset for which lung cancer malignancy metadata is provided. For this small inferencing dataset, our best model achieves sensitivity and specificity of 85% and 75% respectively with a positive predictive value of 85% which is comparable to the performance of human radiologists. Furthermore, the decision trees created by this method may be considered as a starting point for refinement by medical experts into clinically usable multi-variate lung cancer scoring and diagnostic models.
dc.language	en
dc.rights	info:eu-repo/semantics/openAccess
dc.title	Automatic Generation of Interpretable Lung Cancer Scoring Models from Chest X-Ray Images
dc.type	Journal Article
pubs.organisational-group	/University of Technology Sydney/Faculty of Engineering and Information Technology
pubs.organisational-group	/University of Technology Sydney/Faculty of Engineering and Information Technology/School of Information, Systems and Modelling
pubs.organisational-group	/University of Technology Sydney
utslib.copyright.status	open_access	*
dc.date.updated	2021-02-07T22:52:14Z

Abstract:

Lung cancer is the leading cause of cancer death worldwide with early detection being the key to a positive patient prognosis. Although a multitude of studies have demonstrated that machine learning, and particularly deep learning, techniques are effective at automatically diagnosing lung cancer, these techniques have yet to be clinically approved and adopted by the medical community. Most research in this field is focused on the narrow task of nodule detection to provide an artificial radiological second reading. We instead focus on extracting, from chest X-ray images, a wider range of pathologies associated with lung cancer using a computer vision model trained on a large dataset. We then find the set of best fit decision trees against an independent, smaller dataset for which lung cancer malignancy metadata is provided. For this small inferencing dataset, our best model achieves sensitivity and specificity of 85% and 75% respectively with a positive predictive value of 85% which is comparable to the performance of human radiologists. Furthermore, the decision trees created by this method may be considered as a starting point for refinement by medical experts into clinically usable multi-variate lung cancer scoring and diagnostic models.

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/145908