Machine learning for detection of stenoses and aneurysms: application in a physiologically realistic virtual patient database

Jones, G.; Parr, J.; Nithiarasu, Perumal; Pant, Sanjay

doi:10.1007/s10237-021-01497-7

Journal article 731 views 85 downloads

Machine learning for detection of stenoses and aneurysms: application in a physiologically realistic virtual patient database

G. Jones, J. Parr, Perumal Nithiarasu

, Sanjay Pant

Biomechanics and Modeling in Mechanobiology, Volume: 20, Issue: 6, Pages: 2097 - 2146

Swansea University Authors: Perumal Nithiarasu , Sanjay Pant

PDF | Version of Record

© The Author(s) 2021. This article is licensed under a Creative Commons Attribution 4.0 International License
Download (2.52MB)

Check full text

DOI (Published version): 10.1007/s10237-021-01497-7

Abstract

This study presents an application of machine learning (ML) methods for detecting the presence of stenoses and aneurysms in the human arterial system. Four major forms of arterial disease—carotid artery stenosis (CAS), subclavian artery stenosis (SAS), peripheral arterial disease (PAD), and abdomina...

Full description

Published in:	Biomechanics and Modeling in Mechanobiology
ISSN:	1617-7959 1617-7940
Published:	Springer Science and Business Media LLC 2021
Online Access:	Check full text
URI:	https://cronfa.swan.ac.uk/Record/cronfa57307
Tags:	Add Tag No Tags, Be the first to tag this record!

first_indexed	2021-07-12T12:22:49Z
last_indexed	2022-06-24T03:14:45Z
id	cronfa57307
recordtype	SURis
fullrecord	<?xml version="1.0"?><rfc1807><datestamp>2022-06-23T14:56:52.0125725</datestamp><bib-version>v2</bib-version><id>57307</id><entry>2021-07-12</entry><title>Machine learning for detection of stenoses and aneurysms: application in a physiologically realistic virtual patient database</title><swanseaauthors><author><sid>3b28bf59358fc2b9bd9a46897dbfc92d</sid><ORCID>0000-0002-4901-2980</ORCID><firstname>Perumal</firstname><surname>Nithiarasu</surname><name>Perumal Nithiarasu</name><active>true</active><ethesisStudent>false</ethesisStudent></author><author><sid>43b388e955511a9d1b86b863c2018a9f</sid><ORCID>0000-0002-2081-308X</ORCID><firstname>Sanjay</firstname><surname>Pant</surname><name>Sanjay Pant</name><active>true</active><ethesisStudent>false</ethesisStudent></author></swanseaauthors><date>2021-07-12</date><deptcode>CIVL</deptcode><abstract>This study presents an application of machine learning (ML) methods for detecting the presence of stenoses and aneurysms in the human arterial system. Four major forms of arterial disease—carotid artery stenosis (CAS), subclavian artery stenosis (SAS), peripheral arterial disease (PAD), and abdominal aortic aneurysms (AAA)—are considered. The ML methods are trained and tested on a physiologically realistic virtual patient database (VPD) containing 28,868 healthy subjects, adapted from the authors previous work and augmented to include disease. It is found that the tree-based methods of Random Forest and Gradient Boosting outperform other approaches. The performance of ML methods is quantified through the F1 score and computation of sensitivities and specificities. When using six haemodynamic measurements (pressure in the common carotid, brachial, and radial arteries; and flow-rate in the common carotid, brachial, and femoral arteries), it is found that maximum F1 scores larger than 0.9 are achieved for CAS and PAD, larger than 0.85 for SAS, and larger than 0.98 for both low- and high-severity AAAs. Corresponding sensitivities and specificities are larger than 90% for CAS and PAD, larger than 85% for SAS, and larger than 98% for both low- and high-severity AAAs. When reducing the number of measurements, performance is degraded by less than 5% when three measurements are used, and less than 10% when only two measurements are used for classification. For AAA, it is shown that F1 scores larger than 0.85 and corresponding sensitivities and specificities larger than 85% are achievable when using only a single measurement. The results are encouraging to pursue AAA monitoring and screening through wearable devices which can reliably measure pressure or flow-rates.</abstract><type>Journal Article</type><journal>Biomechanics and Modeling in Mechanobiology</journal><volume>20</volume><journalNumber>6</journalNumber><paginationStart>2097</paginationStart><paginationEnd>2146</paginationEnd><publisher>Springer Science and Business Media LLC</publisher><placeOfPublication/><isbnPrint/><isbnElectronic/><issnPrint>1617-7959</issnPrint><issnElectronic>1617-7940</issnElectronic><keywords>Virtual patients; Stenosis; Aneurysm; Pulse wave haemodynamics; Screening; Machine learning</keywords><publishedDay>1</publishedDay><publishedMonth>12</publishedMonth><publishedYear>2021</publishedYear><publishedDate>2021-12-01</publishedDate><doi>10.1007/s10237-021-01497-7</doi><url/><notes/><college>COLLEGE NANME</college><department>Civil Engineering</department><CollegeCode>COLLEGE CODE</CollegeCode><DepartmentCode>CIVL</DepartmentCode><institution>Swansea University</institution><apcterm>SU Library paid the OA fee (TA Institutional Deal)</apcterm><funders>This work is supported by an EPSRC studentship ref. EP/N509553/1 and an EPSRC grant ref. EP/R010811/1.</funders><lastEdited>2022-06-23T14:56:52.0125725</lastEdited><Created>2021-07-12T13:18:46.5307670</Created><path><level id="1">Faculty of Science and Engineering</level><level id="2">School of Aerospace, Civil, Electrical, General and Mechanical Engineering - Civil Engineering</level></path><authors><author><firstname>G.</firstname><surname>Jones</surname><order>1</order></author><author><firstname>J.</firstname><surname>Parr</surname><order>2</order></author><author><firstname>Perumal</firstname><surname>Nithiarasu</surname><orcid>0000-0002-4901-2980</orcid><order>3</order></author><author><firstname>Sanjay</firstname><surname>Pant</surname><orcid>0000-0002-2081-308X</orcid><order>4</order></author></authors><documents><document><filename>57307__20526__d0ae8982775e43798ceeb05fbfdd5c58.pdf</filename><originalFilename>57307.pdf</originalFilename><uploaded>2021-08-03T16:09:26.8703898</uploaded><type>Output</type><contentLength>2647455</contentLength><contentType>application/pdf</contentType><version>Version of Record</version><cronfaStatus>true</cronfaStatus><documentNotes>© The Author(s) 2021. This article is licensed under a Creative Commons Attribution 4.0 International License</documentNotes><copyrightCorrect>true</copyrightCorrect><language>eng</language><licence>http://creativecommons.org/licenses/by/4.0/</licence></document></documents><OutputDurs/></rfc1807>
spelling	2022-06-23T14:56:52.0125725 v2 57307 2021-07-12 Machine learning for detection of stenoses and aneurysms: application in a physiologically realistic virtual patient database 3b28bf59358fc2b9bd9a46897dbfc92d 0000-0002-4901-2980 Perumal Nithiarasu Perumal Nithiarasu true false 43b388e955511a9d1b86b863c2018a9f 0000-0002-2081-308X Sanjay Pant Sanjay Pant true false 2021-07-12 CIVL This study presents an application of machine learning (ML) methods for detecting the presence of stenoses and aneurysms in the human arterial system. Four major forms of arterial disease—carotid artery stenosis (CAS), subclavian artery stenosis (SAS), peripheral arterial disease (PAD), and abdominal aortic aneurysms (AAA)—are considered. The ML methods are trained and tested on a physiologically realistic virtual patient database (VPD) containing 28,868 healthy subjects, adapted from the authors previous work and augmented to include disease. It is found that the tree-based methods of Random Forest and Gradient Boosting outperform other approaches. The performance of ML methods is quantified through the F1 score and computation of sensitivities and specificities. When using six haemodynamic measurements (pressure in the common carotid, brachial, and radial arteries; and flow-rate in the common carotid, brachial, and femoral arteries), it is found that maximum F1 scores larger than 0.9 are achieved for CAS and PAD, larger than 0.85 for SAS, and larger than 0.98 for both low- and high-severity AAAs. Corresponding sensitivities and specificities are larger than 90% for CAS and PAD, larger than 85% for SAS, and larger than 98% for both low- and high-severity AAAs. When reducing the number of measurements, performance is degraded by less than 5% when three measurements are used, and less than 10% when only two measurements are used for classification. For AAA, it is shown that F1 scores larger than 0.85 and corresponding sensitivities and specificities larger than 85% are achievable when using only a single measurement. The results are encouraging to pursue AAA monitoring and screening through wearable devices which can reliably measure pressure or flow-rates. Journal Article Biomechanics and Modeling in Mechanobiology 20 6 2097 2146 Springer Science and Business Media LLC 1617-7959 1617-7940 Virtual patients; Stenosis; Aneurysm; Pulse wave haemodynamics; Screening; Machine learning 1 12 2021 2021-12-01 10.1007/s10237-021-01497-7 COLLEGE NANME Civil Engineering COLLEGE CODE CIVL Swansea University SU Library paid the OA fee (TA Institutional Deal) This work is supported by an EPSRC studentship ref. EP/N509553/1 and an EPSRC grant ref. EP/R010811/1. 2022-06-23T14:56:52.0125725 2021-07-12T13:18:46.5307670 Faculty of Science and Engineering School of Aerospace, Civil, Electrical, General and Mechanical Engineering - Civil Engineering G. Jones 1 J. Parr 2 Perumal Nithiarasu 0000-0002-4901-2980 3 Sanjay Pant 0000-0002-2081-308X 4 57307__20526__d0ae8982775e43798ceeb05fbfdd5c58.pdf 57307.pdf 2021-08-03T16:09:26.8703898 Output 2647455 application/pdf Version of Record true © The Author(s) 2021. This article is licensed under a Creative Commons Attribution 4.0 International License true eng http://creativecommons.org/licenses/by/4.0/
title	Machine learning for detection of stenoses and aneurysms: application in a physiologically realistic virtual patient database
spellingShingle	Machine learning for detection of stenoses and aneurysms: application in a physiologically realistic virtual patient database Perumal Nithiarasu Sanjay Pant
title_short	Machine learning for detection of stenoses and aneurysms: application in a physiologically realistic virtual patient database
title_full	Machine learning for detection of stenoses and aneurysms: application in a physiologically realistic virtual patient database
title_fullStr	Machine learning for detection of stenoses and aneurysms: application in a physiologically realistic virtual patient database
title_full_unstemmed	Machine learning for detection of stenoses and aneurysms: application in a physiologically realistic virtual patient database
title_sort	Machine learning for detection of stenoses and aneurysms: application in a physiologically realistic virtual patient database
author_id_str_mv	3b28bf59358fc2b9bd9a46897dbfc92d 43b388e955511a9d1b86b863c2018a9f
author_id_fullname_str_mv	3b28bf59358fc2b9bd9a46897dbfc92d_*_Perumal Nithiarasu 43b388e955511a9d1b86b863c2018a9f_*_Sanjay Pant
author	Perumal Nithiarasu Sanjay Pant
author2	G. Jones J. Parr Perumal Nithiarasu Sanjay Pant
format	Journal article
container_title	Biomechanics and Modeling in Mechanobiology
container_volume	20
container_issue	6
container_start_page	2097
publishDate	2021
institution	Swansea University
issn	1617-7959 1617-7940
doi_str_mv	10.1007/s10237-021-01497-7
publisher	Springer Science and Business Media LLC
college_str	Faculty of Science and Engineering
hierarchytype
hierarchy_top_id	facultyofscienceandengineering
hierarchy_top_title	Faculty of Science and Engineering
hierarchy_parent_id	facultyofscienceandengineering
hierarchy_parent_title	Faculty of Science and Engineering
department_str	School of Aerospace, Civil, Electrical, General and Mechanical Engineering - Civil Engineering{{{_:::_}}}Faculty of Science and Engineering{{{_:::_}}}School of Aerospace, Civil, Electrical, General and Mechanical Engineering - Civil Engineering
document_store_str	1
active_str	0
description	This study presents an application of machine learning (ML) methods for detecting the presence of stenoses and aneurysms in the human arterial system. Four major forms of arterial disease—carotid artery stenosis (CAS), subclavian artery stenosis (SAS), peripheral arterial disease (PAD), and abdominal aortic aneurysms (AAA)—are considered. The ML methods are trained and tested on a physiologically realistic virtual patient database (VPD) containing 28,868 healthy subjects, adapted from the authors previous work and augmented to include disease. It is found that the tree-based methods of Random Forest and Gradient Boosting outperform other approaches. The performance of ML methods is quantified through the F1 score and computation of sensitivities and specificities. When using six haemodynamic measurements (pressure in the common carotid, brachial, and radial arteries; and flow-rate in the common carotid, brachial, and femoral arteries), it is found that maximum F1 scores larger than 0.9 are achieved for CAS and PAD, larger than 0.85 for SAS, and larger than 0.98 for both low- and high-severity AAAs. Corresponding sensitivities and specificities are larger than 90% for CAS and PAD, larger than 85% for SAS, and larger than 98% for both low- and high-severity AAAs. When reducing the number of measurements, performance is degraded by less than 5% when three measurements are used, and less than 10% when only two measurements are used for classification. For AAA, it is shown that F1 scores larger than 0.85 and corresponding sensitivities and specificities larger than 85% are achievable when using only a single measurement. The results are encouraging to pursue AAA monitoring and screening through wearable devices which can reliably measure pressure or flow-rates.
published_date	2021-12-01T04:12:57Z
_version_	1763753881525813248
score	11.016704

Machine learning for detection of stenoses and aneurysms: application in a physiologically realistic virtual patient database

Similar Items