Diagnosis and Severity Assessment of COPD Using a Novel Fast-Response Capnometer and Interpretable Machine Learning

Introduction Spirometry is the gold standard for COPD diagnosis and severity determination, but is technique-dependent, nonspecific, and requires administration by a trained healthcare professional. There is a need for a fast, reliable, and precise alternative diagnostic test. This study’s aim was t...

Full description

Saved in:
Bibliographic Details
Main Authors: Leeran Talker, Cihan Dogan, Daniel Neville, Rui Hen Lim, Henry Broomfield, Gabriel Lambert, Ahmed Selim, Thomas Brown, Laura Wiffen, Julian Carter, Helen F. Ashdown, Gail Hayward, Elango Vijaykumar, Scott T. Weiss, Anoop Chauhan, Ameera X. Patel
Format: Article
Language:English
Published: Taylor & Francis Group 2024-12-01
Series:COPD
Subjects:
Online Access:https://www.tandfonline.com/doi/10.1080/15412555.2024.2321379
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846107958584279040
author Leeran Talker
Cihan Dogan
Daniel Neville
Rui Hen Lim
Henry Broomfield
Gabriel Lambert
Ahmed Selim
Thomas Brown
Laura Wiffen
Julian Carter
Helen F. Ashdown
Gail Hayward
Elango Vijaykumar
Scott T. Weiss
Anoop Chauhan
Ameera X. Patel
author_facet Leeran Talker
Cihan Dogan
Daniel Neville
Rui Hen Lim
Henry Broomfield
Gabriel Lambert
Ahmed Selim
Thomas Brown
Laura Wiffen
Julian Carter
Helen F. Ashdown
Gail Hayward
Elango Vijaykumar
Scott T. Weiss
Anoop Chauhan
Ameera X. Patel
author_sort Leeran Talker
collection DOAJ
description Introduction Spirometry is the gold standard for COPD diagnosis and severity determination, but is technique-dependent, nonspecific, and requires administration by a trained healthcare professional. There is a need for a fast, reliable, and precise alternative diagnostic test. This study’s aim was to use interpretable machine learning to diagnose COPD and assess severity using 75-second carbon dioxide (CO2) breath records captured with TidalSense’s N-TidalTM capnometer.Method For COPD diagnosis, machine learning algorithms were trained and evaluated on 294 COPD (including GOLD stages 1–4) and 705 non-COPD participants. A logistic regression model was also trained to distinguish GOLD 1 from GOLD 4 COPD with the output probability used as an index of severity.Results The best diagnostic model achieved an AUROC of 0.890, sensitivity of 0.771, specificity of 0.850 and positive predictive value (PPV) of 0.834. Evaluating performance on all test capnograms that were confidently ruled in or out yielded PPV of 0.930 and NPV of 0.890. The severity determination model yielded an AUROC of 0.980, sensitivity of 0.958, specificity of 0.961 and PPV of 0.958 in distinguishing GOLD 1 from GOLD 4. Output probabilities from the severity determination model produced a correlation of 0.71 with percentage predicted FEV1.Conclusion The N-TidalTM device could be used alongside interpretable machine learning as an accurate, point-of-care diagnostic test for COPD, particularly in primary care as a rapid rule-in or rule-out test. N-TidalTM also could be effective in monitoring disease progression, providing a possible alternative to spirometry for disease monitoring.
format Article
id doaj-art-be0a1094073f45a89b833fd1e0528dff
institution Kabale University
issn 1541-2555
1541-2563
language English
publishDate 2024-12-01
publisher Taylor & Francis Group
record_format Article
series COPD
spelling doaj-art-be0a1094073f45a89b833fd1e0528dff2024-12-26T08:50:44ZengTaylor & Francis GroupCOPD1541-25551541-25632024-12-0121110.1080/15412555.2024.2321379Diagnosis and Severity Assessment of COPD Using a Novel Fast-Response Capnometer and Interpretable Machine LearningLeeran Talker0Cihan Dogan1Daniel Neville2Rui Hen Lim3Henry Broomfield4Gabriel Lambert5Ahmed Selim6Thomas Brown7Laura Wiffen8Julian Carter9Helen F. Ashdown10Gail Hayward11Elango Vijaykumar12Scott T. Weiss13Anoop Chauhan14Ameera X. Patel15Department of Machine Learning, TidalSense, Cambridge, UKDepartment of Machine Learning, TidalSense, Cambridge, UKRespiratory Department, Portsmouth Hospitals University NHS Foundation Trust, Portsmouth, UKDepartment of Machine Learning, TidalSense, Cambridge, UKDepartment of Machine Learning, TidalSense, Cambridge, UKDepartment of Clinical Operations, TidalSense, Cambridge, UKDepartment of Machine Learning, TidalSense, Cambridge, UKRespiratory Department, Portsmouth Hospitals University NHS Foundation Trust, Portsmouth, UKRespiratory Department, Portsmouth Hospitals University NHS Foundation Trust, Portsmouth, UKDepartment of Engineering, TidalSense, Cambridge, UKDepartment of Primary Care Health Sciences, NIHR Community Healthcare MedTech and IVD Cooperative, University of Oxford, Oxford, UKDepartment of Primary Care Health Sciences, NIHR Community Healthcare MedTech and IVD Cooperative, University of Oxford, Oxford, UKDepartment of Research, Modality GP Partnership, UKDepartment of Medicine, Channing Division of Network Medicine, Harvard Medical School, Boston, MA, USARespiratory Department, Portsmouth Hospitals University NHS Foundation Trust, Portsmouth, UKExecutive Department, TidalSense, Cambridge, UKIntroduction Spirometry is the gold standard for COPD diagnosis and severity determination, but is technique-dependent, nonspecific, and requires administration by a trained healthcare professional. There is a need for a fast, reliable, and precise alternative diagnostic test. This study’s aim was to use interpretable machine learning to diagnose COPD and assess severity using 75-second carbon dioxide (CO2) breath records captured with TidalSense’s N-TidalTM capnometer.Method For COPD diagnosis, machine learning algorithms were trained and evaluated on 294 COPD (including GOLD stages 1–4) and 705 non-COPD participants. A logistic regression model was also trained to distinguish GOLD 1 from GOLD 4 COPD with the output probability used as an index of severity.Results The best diagnostic model achieved an AUROC of 0.890, sensitivity of 0.771, specificity of 0.850 and positive predictive value (PPV) of 0.834. Evaluating performance on all test capnograms that were confidently ruled in or out yielded PPV of 0.930 and NPV of 0.890. The severity determination model yielded an AUROC of 0.980, sensitivity of 0.958, specificity of 0.961 and PPV of 0.958 in distinguishing GOLD 1 from GOLD 4. Output probabilities from the severity determination model produced a correlation of 0.71 with percentage predicted FEV1.Conclusion The N-TidalTM device could be used alongside interpretable machine learning as an accurate, point-of-care diagnostic test for COPD, particularly in primary care as a rapid rule-in or rule-out test. N-TidalTM also could be effective in monitoring disease progression, providing a possible alternative to spirometry for disease monitoring.https://www.tandfonline.com/doi/10.1080/15412555.2024.2321379Pulmonary Diseasechronic obstructivemachine learningdiagnosiscapnometryseverity assessment
spellingShingle Leeran Talker
Cihan Dogan
Daniel Neville
Rui Hen Lim
Henry Broomfield
Gabriel Lambert
Ahmed Selim
Thomas Brown
Laura Wiffen
Julian Carter
Helen F. Ashdown
Gail Hayward
Elango Vijaykumar
Scott T. Weiss
Anoop Chauhan
Ameera X. Patel
Diagnosis and Severity Assessment of COPD Using a Novel Fast-Response Capnometer and Interpretable Machine Learning
COPD
Pulmonary Disease
chronic obstructive
machine learning
diagnosis
capnometry
severity assessment
title Diagnosis and Severity Assessment of COPD Using a Novel Fast-Response Capnometer and Interpretable Machine Learning
title_full Diagnosis and Severity Assessment of COPD Using a Novel Fast-Response Capnometer and Interpretable Machine Learning
title_fullStr Diagnosis and Severity Assessment of COPD Using a Novel Fast-Response Capnometer and Interpretable Machine Learning
title_full_unstemmed Diagnosis and Severity Assessment of COPD Using a Novel Fast-Response Capnometer and Interpretable Machine Learning
title_short Diagnosis and Severity Assessment of COPD Using a Novel Fast-Response Capnometer and Interpretable Machine Learning
title_sort diagnosis and severity assessment of copd using a novel fast response capnometer and interpretable machine learning
topic Pulmonary Disease
chronic obstructive
machine learning
diagnosis
capnometry
severity assessment
url https://www.tandfonline.com/doi/10.1080/15412555.2024.2321379
work_keys_str_mv AT leerantalker diagnosisandseverityassessmentofcopdusinganovelfastresponsecapnometerandinterpretablemachinelearning
AT cihandogan diagnosisandseverityassessmentofcopdusinganovelfastresponsecapnometerandinterpretablemachinelearning
AT danielneville diagnosisandseverityassessmentofcopdusinganovelfastresponsecapnometerandinterpretablemachinelearning
AT ruihenlim diagnosisandseverityassessmentofcopdusinganovelfastresponsecapnometerandinterpretablemachinelearning
AT henrybroomfield diagnosisandseverityassessmentofcopdusinganovelfastresponsecapnometerandinterpretablemachinelearning
AT gabriellambert diagnosisandseverityassessmentofcopdusinganovelfastresponsecapnometerandinterpretablemachinelearning
AT ahmedselim diagnosisandseverityassessmentofcopdusinganovelfastresponsecapnometerandinterpretablemachinelearning
AT thomasbrown diagnosisandseverityassessmentofcopdusinganovelfastresponsecapnometerandinterpretablemachinelearning
AT laurawiffen diagnosisandseverityassessmentofcopdusinganovelfastresponsecapnometerandinterpretablemachinelearning
AT juliancarter diagnosisandseverityassessmentofcopdusinganovelfastresponsecapnometerandinterpretablemachinelearning
AT helenfashdown diagnosisandseverityassessmentofcopdusinganovelfastresponsecapnometerandinterpretablemachinelearning
AT gailhayward diagnosisandseverityassessmentofcopdusinganovelfastresponsecapnometerandinterpretablemachinelearning
AT elangovijaykumar diagnosisandseverityassessmentofcopdusinganovelfastresponsecapnometerandinterpretablemachinelearning
AT scotttweiss diagnosisandseverityassessmentofcopdusinganovelfastresponsecapnometerandinterpretablemachinelearning
AT anoopchauhan diagnosisandseverityassessmentofcopdusinganovelfastresponsecapnometerandinterpretablemachinelearning
AT ameeraxpatel diagnosisandseverityassessmentofcopdusinganovelfastresponsecapnometerandinterpretablemachinelearning