No Cover Image

Journal article 417 views 35 downloads

Digital solution for detection of undiagnosed diabetes using machine learning-based retinal image analysis

Benny Zee Orcid Logo, Jack Lee, Maria Lai, Peter Chee, Jim Rafferty Orcid Logo, Becky Thomas, David Owens Orcid Logo

BMJ Open Diabetes Research and Care, Volume: 10, Issue: 6, Start page: e002914

Swansea University Authors: Jim Rafferty Orcid Logo, Becky Thomas, David Owens Orcid Logo

  • 62304_VoR.pdf

    PDF | Version of Record

    This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license

    Download (695.33KB)

Check full text

DOI (Published version): 10.1136/bmjdrc-2022-002914

Abstract

Introduction Undiagnosed diabetes is a global health issue. Previous studies have estimated that about 24.1%–75.1% of all diabetes cases are undiagnosed, leading to more diabetic complications and inducing huge healthcare costs. Many current methods for diabetes diagnosis rely on metabolic indices a...

Full description

Published in: BMJ Open Diabetes Research and Care
ISSN: 2052-4897
Published: BMJ 2022
Online Access: Check full text

URI: https://cronfa.swan.ac.uk/Record/cronfa62304
Tags: Add Tag
No Tags, Be the first to tag this record!
first_indexed 2023-01-30T13:21:00Z
last_indexed 2023-01-31T04:18:48Z
id cronfa62304
recordtype SURis
fullrecord <?xml version="1.0"?><rfc1807><datestamp>2023-01-30T13:23:25.3806654</datestamp><bib-version>v2</bib-version><id>62304</id><entry>2023-01-11</entry><title>Digital solution for detection of undiagnosed diabetes using machine learning-based retinal image analysis</title><swanseaauthors><author><sid>52effe759a718bd36eb12cdd10fe1a09</sid><ORCID>0000-0002-1667-7265</ORCID><firstname>Jim</firstname><surname>Rafferty</surname><name>Jim Rafferty</name><active>true</active><ethesisStudent>false</ethesisStudent></author><author><sid>e83b45ec71428bd748ce201048f43d6a</sid><ORCID/><firstname>Becky</firstname><surname>Thomas</surname><name>Becky Thomas</name><active>true</active><ethesisStudent>false</ethesisStudent></author><author><sid>2fd4b7c3f82c6d3bd546eff61ff944e9</sid><ORCID>0000-0003-1002-1238</ORCID><firstname>David</firstname><surname>Owens</surname><name>David Owens</name><active>true</active><ethesisStudent>false</ethesisStudent></author></swanseaauthors><date>2023-01-11</date><deptcode>HDAT</deptcode><abstract>Introduction Undiagnosed diabetes is a global health issue. Previous studies have estimated that about 24.1%&#x2013;75.1% of all diabetes cases are undiagnosed, leading to more diabetic complications and inducing huge healthcare costs. Many current methods for diabetes diagnosis rely on metabolic indices and are subject to considerable variability. In contrast, a digital approach based on retinal image represents a stable marker of overall glycemic status.Research design and methods Our study involves 2221 subjects for developing a classification model, with 945 subjects with diabetes and 1276 controls. The training data included 70% and the testing data 30% of the subjects. All subjects had their retinal images taken using a non-mydriatic fundus camera. Two separate data sets were used for external validation. The Hong Kong testing data contain 734 controls without diabetes and 660 subjects with diabetes, and the UK testing data have 1682 subjects with diabetes.Results The 10-fold cross-validation using the support vector machine approach has a sensitivity of 92% and a specificity of 96.2%. The separate testing data from Hong Kong provided a sensitivity of 99.5% and a specificity of 91.1%. For the UK testing data, the sensitivity is 98.0%. The accuracy of the Caucasian retinal images is comparable with that of the Asian data. It implies that the digital method can be applied globally. Those with diabetes complications in both Hong Kong and UK data have a higher probability of risk of diabetes compared with diabetes subjects without complications.Conclusions A digital machine learning-based method to estimate the risk of diabetes based on retinal images has been developed and validated using both Asian and Caucasian data. Retinal image analysis is a fast, convenient, and non-invasive technique for community health applications. In addition, it is an ideal solution for undiagnosed diabetes prescreening.</abstract><type>Journal Article</type><journal>BMJ Open Diabetes Research and Care</journal><volume>10</volume><journalNumber>6</journalNumber><paginationStart>e002914</paginationStart><paginationEnd/><publisher>BMJ</publisher><placeOfPublication/><isbnPrint/><isbnElectronic/><issnPrint/><issnElectronic>2052-4897</issnElectronic><keywords/><publishedDay>22</publishedDay><publishedMonth>12</publishedMonth><publishedYear>2022</publishedYear><publishedDate>2022-12-22</publishedDate><doi>10.1136/bmjdrc-2022-002914</doi><url/><notes/><college>COLLEGE NANME</college><department>Health Data Science</department><CollegeCode>COLLEGE CODE</CollegeCode><DepartmentCode>HDAT</DepartmentCode><institution>Swansea University</institution><apcterm/><funders>This study was supported by the Hong Kong Innovation and Technology Fund - Midstream Research Programme (MRP/037/17X).</funders><projectreference/><lastEdited>2023-01-30T13:23:25.3806654</lastEdited><Created>2023-01-11T14:51:06.0810257</Created><path><level id="1">Faculty of Medicine, Health and Life Sciences</level><level id="2">Swansea University Medical School - Medicine</level></path><authors><author><firstname>Benny</firstname><surname>Zee</surname><orcid>0000-0002-7238-845x</orcid><order>1</order></author><author><firstname>Jack</firstname><surname>Lee</surname><order>2</order></author><author><firstname>Maria</firstname><surname>Lai</surname><order>3</order></author><author><firstname>Peter</firstname><surname>Chee</surname><order>4</order></author><author><firstname>Jim</firstname><surname>Rafferty</surname><orcid>0000-0002-1667-7265</orcid><order>5</order></author><author><firstname>Becky</firstname><surname>Thomas</surname><orcid/><order>6</order></author><author><firstname>David</firstname><surname>Owens</surname><orcid>0000-0003-1002-1238</orcid><order>7</order></author></authors><documents><document><filename>62304__26431__e982f6eae0324ffe895385bf1fb58ecb.pdf</filename><originalFilename>62304_VoR.pdf</originalFilename><uploaded>2023-01-30T13:21:24.8103236</uploaded><type>Output</type><contentLength>712016</contentLength><contentType>application/pdf</contentType><version>Version of Record</version><cronfaStatus>true</cronfaStatus><documentNotes>This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license</documentNotes><copyrightCorrect>true</copyrightCorrect><language>eng</language><licence>http://creativecommons.org/licenses/by-nc/4.0/</licence></document></documents><OutputDurs/></rfc1807>
spelling 2023-01-30T13:23:25.3806654 v2 62304 2023-01-11 Digital solution for detection of undiagnosed diabetes using machine learning-based retinal image analysis 52effe759a718bd36eb12cdd10fe1a09 0000-0002-1667-7265 Jim Rafferty Jim Rafferty true false e83b45ec71428bd748ce201048f43d6a Becky Thomas Becky Thomas true false 2fd4b7c3f82c6d3bd546eff61ff944e9 0000-0003-1002-1238 David Owens David Owens true false 2023-01-11 HDAT Introduction Undiagnosed diabetes is a global health issue. Previous studies have estimated that about 24.1%–75.1% of all diabetes cases are undiagnosed, leading to more diabetic complications and inducing huge healthcare costs. Many current methods for diabetes diagnosis rely on metabolic indices and are subject to considerable variability. In contrast, a digital approach based on retinal image represents a stable marker of overall glycemic status.Research design and methods Our study involves 2221 subjects for developing a classification model, with 945 subjects with diabetes and 1276 controls. The training data included 70% and the testing data 30% of the subjects. All subjects had their retinal images taken using a non-mydriatic fundus camera. Two separate data sets were used for external validation. The Hong Kong testing data contain 734 controls without diabetes and 660 subjects with diabetes, and the UK testing data have 1682 subjects with diabetes.Results The 10-fold cross-validation using the support vector machine approach has a sensitivity of 92% and a specificity of 96.2%. The separate testing data from Hong Kong provided a sensitivity of 99.5% and a specificity of 91.1%. For the UK testing data, the sensitivity is 98.0%. The accuracy of the Caucasian retinal images is comparable with that of the Asian data. It implies that the digital method can be applied globally. Those with diabetes complications in both Hong Kong and UK data have a higher probability of risk of diabetes compared with diabetes subjects without complications.Conclusions A digital machine learning-based method to estimate the risk of diabetes based on retinal images has been developed and validated using both Asian and Caucasian data. Retinal image analysis is a fast, convenient, and non-invasive technique for community health applications. In addition, it is an ideal solution for undiagnosed diabetes prescreening. Journal Article BMJ Open Diabetes Research and Care 10 6 e002914 BMJ 2052-4897 22 12 2022 2022-12-22 10.1136/bmjdrc-2022-002914 COLLEGE NANME Health Data Science COLLEGE CODE HDAT Swansea University This study was supported by the Hong Kong Innovation and Technology Fund - Midstream Research Programme (MRP/037/17X). 2023-01-30T13:23:25.3806654 2023-01-11T14:51:06.0810257 Faculty of Medicine, Health and Life Sciences Swansea University Medical School - Medicine Benny Zee 0000-0002-7238-845x 1 Jack Lee 2 Maria Lai 3 Peter Chee 4 Jim Rafferty 0000-0002-1667-7265 5 Becky Thomas 6 David Owens 0000-0003-1002-1238 7 62304__26431__e982f6eae0324ffe895385bf1fb58ecb.pdf 62304_VoR.pdf 2023-01-30T13:21:24.8103236 Output 712016 application/pdf Version of Record true This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license true eng http://creativecommons.org/licenses/by-nc/4.0/
title Digital solution for detection of undiagnosed diabetes using machine learning-based retinal image analysis
spellingShingle Digital solution for detection of undiagnosed diabetes using machine learning-based retinal image analysis
Jim Rafferty
Becky Thomas
David Owens
title_short Digital solution for detection of undiagnosed diabetes using machine learning-based retinal image analysis
title_full Digital solution for detection of undiagnosed diabetes using machine learning-based retinal image analysis
title_fullStr Digital solution for detection of undiagnosed diabetes using machine learning-based retinal image analysis
title_full_unstemmed Digital solution for detection of undiagnosed diabetes using machine learning-based retinal image analysis
title_sort Digital solution for detection of undiagnosed diabetes using machine learning-based retinal image analysis
author_id_str_mv 52effe759a718bd36eb12cdd10fe1a09
e83b45ec71428bd748ce201048f43d6a
2fd4b7c3f82c6d3bd546eff61ff944e9
author_id_fullname_str_mv 52effe759a718bd36eb12cdd10fe1a09_***_Jim Rafferty
e83b45ec71428bd748ce201048f43d6a_***_Becky Thomas
2fd4b7c3f82c6d3bd546eff61ff944e9_***_David Owens
author Jim Rafferty
Becky Thomas
David Owens
author2 Benny Zee
Jack Lee
Maria Lai
Peter Chee
Jim Rafferty
Becky Thomas
David Owens
format Journal article
container_title BMJ Open Diabetes Research and Care
container_volume 10
container_issue 6
container_start_page e002914
publishDate 2022
institution Swansea University
issn 2052-4897
doi_str_mv 10.1136/bmjdrc-2022-002914
publisher BMJ
college_str Faculty of Medicine, Health and Life Sciences
hierarchytype
hierarchy_top_id facultyofmedicinehealthandlifesciences
hierarchy_top_title Faculty of Medicine, Health and Life Sciences
hierarchy_parent_id facultyofmedicinehealthandlifesciences
hierarchy_parent_title Faculty of Medicine, Health and Life Sciences
department_str Swansea University Medical School - Medicine{{{_:::_}}}Faculty of Medicine, Health and Life Sciences{{{_:::_}}}Swansea University Medical School - Medicine
document_store_str 1
active_str 0
description Introduction Undiagnosed diabetes is a global health issue. Previous studies have estimated that about 24.1%–75.1% of all diabetes cases are undiagnosed, leading to more diabetic complications and inducing huge healthcare costs. Many current methods for diabetes diagnosis rely on metabolic indices and are subject to considerable variability. In contrast, a digital approach based on retinal image represents a stable marker of overall glycemic status.Research design and methods Our study involves 2221 subjects for developing a classification model, with 945 subjects with diabetes and 1276 controls. The training data included 70% and the testing data 30% of the subjects. All subjects had their retinal images taken using a non-mydriatic fundus camera. Two separate data sets were used for external validation. The Hong Kong testing data contain 734 controls without diabetes and 660 subjects with diabetes, and the UK testing data have 1682 subjects with diabetes.Results The 10-fold cross-validation using the support vector machine approach has a sensitivity of 92% and a specificity of 96.2%. The separate testing data from Hong Kong provided a sensitivity of 99.5% and a specificity of 91.1%. For the UK testing data, the sensitivity is 98.0%. The accuracy of the Caucasian retinal images is comparable with that of the Asian data. It implies that the digital method can be applied globally. Those with diabetes complications in both Hong Kong and UK data have a higher probability of risk of diabetes compared with diabetes subjects without complications.Conclusions A digital machine learning-based method to estimate the risk of diabetes based on retinal images has been developed and validated using both Asian and Caucasian data. Retinal image analysis is a fast, convenient, and non-invasive technique for community health applications. In addition, it is an ideal solution for undiagnosed diabetes prescreening.
published_date 2022-12-22T04:21:48Z
_version_ 1763754439010680832
score 11.012678

Similar Items