Development and Evaluation of a Deep Learning–Based Pulmonary Hypertension Screening Algorithm Using a Digital Stethoscope

Background Despite the poor outcomes related to the presence of pulmonary hypertension, it often goes undiagnosed in part because of low suspicion and screening tools not being easily accessible such as echocardiography. A new readily available screening tool to identify elevated pulmonary artery sy...

Full description

Saved in:
Bibliographic Details
Main Authors: Ling Guo, Nivedita Khobragade, Spencer Kieu, Suleman Ilyas, Preston N. Nicely, Emmanuel K. Asiedu, Fabio V. Lima, Caroline Currie, Emileigh Lastowski, Gaurav Choudhary
Format: Article
Language:English
Published: Wiley 2025-02-01
Series:Journal of the American Heart Association: Cardiovascular and Cerebrovascular Disease
Subjects:
Online Access:https://www.ahajournals.org/doi/10.1161/JAHA.124.036882
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1832540980278984704
author Ling Guo
Nivedita Khobragade
Spencer Kieu
Suleman Ilyas
Preston N. Nicely
Emmanuel K. Asiedu
Fabio V. Lima
Caroline Currie
Emileigh Lastowski
Gaurav Choudhary
author_facet Ling Guo
Nivedita Khobragade
Spencer Kieu
Suleman Ilyas
Preston N. Nicely
Emmanuel K. Asiedu
Fabio V. Lima
Caroline Currie
Emileigh Lastowski
Gaurav Choudhary
author_sort Ling Guo
collection DOAJ
description Background Despite the poor outcomes related to the presence of pulmonary hypertension, it often goes undiagnosed in part because of low suspicion and screening tools not being easily accessible such as echocardiography. A new readily available screening tool to identify elevated pulmonary artery systolic pressures is needed to help with the prognosis and timely treatment of underlying causes such as heart failure or pulmonary vascular remodeling. We developed a deep learning–based method that uses phonocardiograms (PCGs) for the detection of elevated pulmonary artery systolic pressure, an indicator of pulmonary hypertension. Methods Approximately 6000 PCG recordings with the corresponding echocardiogram‐based estimated pulmonary artery systolic pressure values, as well as ≈169 000 PCG recordings without associated echocardiograms, were used for training a deep convolutional network to detect pulmonary artery systolic pressures ≥40 mm Hg in a semisupervised manner. Each 15‐second PCG, recorded using a digital stethoscope, was processed to generate 5‐second mel‐spectrograms. An additional labeled data set of 196 patients was used for testing. GradCAM++ was used to visualize high importance segments contributing to the network decision. Results An average area under the receiver operator characteristic curve of 0.79 was obtained across 5 cross‐validation folds. The testing data set gave a sensitivity of 0.71 and a specificity of 0.73, with pulmonic and left subclavicular locations having higher sensitivities. GradCAM++ technique highlighted physiologically meaningful PCG segments in example pulmonary hypertension recordings. Conclusions We demonstrated the feasibility of using digital stethoscopes in conjunction with deep learning algorithms as a low‐cost, noninvasive, and easily accessible screening tool for early detection of pulmonary hypertension.
format Article
id doaj-art-f7cdc8ac0d3f4af8bb444786bace1f2c
institution Kabale University
issn 2047-9980
language English
publishDate 2025-02-01
publisher Wiley
record_format Article
series Journal of the American Heart Association: Cardiovascular and Cerebrovascular Disease
spelling doaj-art-f7cdc8ac0d3f4af8bb444786bace1f2c2025-02-04T11:00:01ZengWileyJournal of the American Heart Association: Cardiovascular and Cerebrovascular Disease2047-99802025-02-0114310.1161/JAHA.124.036882Development and Evaluation of a Deep Learning–Based Pulmonary Hypertension Screening Algorithm Using a Digital StethoscopeLing Guo0Nivedita Khobragade1Spencer Kieu2Suleman Ilyas3Preston N. Nicely4Emmanuel K. Asiedu5Fabio V. Lima6Caroline Currie7Emileigh Lastowski8Gaurav Choudhary9Eko Health Inc. Emeryville CA USAEko Health Inc. Emeryville CA USAEko Health Inc. Emeryville CA USAWarren Alpert Medical School Brown University Providence RI USAWarren Alpert Medical School Brown University Providence RI USAWarren Alpert Medical School Brown University Providence RI USAWarren Alpert Medical School Brown University Providence RI USAEko Health Inc. Emeryville CA USAEko Health Inc. Emeryville CA USAWarren Alpert Medical School Brown University Providence RI USABackground Despite the poor outcomes related to the presence of pulmonary hypertension, it often goes undiagnosed in part because of low suspicion and screening tools not being easily accessible such as echocardiography. A new readily available screening tool to identify elevated pulmonary artery systolic pressures is needed to help with the prognosis and timely treatment of underlying causes such as heart failure or pulmonary vascular remodeling. We developed a deep learning–based method that uses phonocardiograms (PCGs) for the detection of elevated pulmonary artery systolic pressure, an indicator of pulmonary hypertension. Methods Approximately 6000 PCG recordings with the corresponding echocardiogram‐based estimated pulmonary artery systolic pressure values, as well as ≈169 000 PCG recordings without associated echocardiograms, were used for training a deep convolutional network to detect pulmonary artery systolic pressures ≥40 mm Hg in a semisupervised manner. Each 15‐second PCG, recorded using a digital stethoscope, was processed to generate 5‐second mel‐spectrograms. An additional labeled data set of 196 patients was used for testing. GradCAM++ was used to visualize high importance segments contributing to the network decision. Results An average area under the receiver operator characteristic curve of 0.79 was obtained across 5 cross‐validation folds. The testing data set gave a sensitivity of 0.71 and a specificity of 0.73, with pulmonic and left subclavicular locations having higher sensitivities. GradCAM++ technique highlighted physiologically meaningful PCG segments in example pulmonary hypertension recordings. Conclusions We demonstrated the feasibility of using digital stethoscopes in conjunction with deep learning algorithms as a low‐cost, noninvasive, and easily accessible screening tool for early detection of pulmonary hypertension.https://www.ahajournals.org/doi/10.1161/JAHA.124.036882deep learningdigital stethoscopespulmonary hypertension
spellingShingle Ling Guo
Nivedita Khobragade
Spencer Kieu
Suleman Ilyas
Preston N. Nicely
Emmanuel K. Asiedu
Fabio V. Lima
Caroline Currie
Emileigh Lastowski
Gaurav Choudhary
Development and Evaluation of a Deep Learning–Based Pulmonary Hypertension Screening Algorithm Using a Digital Stethoscope
Journal of the American Heart Association: Cardiovascular and Cerebrovascular Disease
deep learning
digital stethoscopes
pulmonary hypertension
title Development and Evaluation of a Deep Learning–Based Pulmonary Hypertension Screening Algorithm Using a Digital Stethoscope
title_full Development and Evaluation of a Deep Learning–Based Pulmonary Hypertension Screening Algorithm Using a Digital Stethoscope
title_fullStr Development and Evaluation of a Deep Learning–Based Pulmonary Hypertension Screening Algorithm Using a Digital Stethoscope
title_full_unstemmed Development and Evaluation of a Deep Learning–Based Pulmonary Hypertension Screening Algorithm Using a Digital Stethoscope
title_short Development and Evaluation of a Deep Learning–Based Pulmonary Hypertension Screening Algorithm Using a Digital Stethoscope
title_sort development and evaluation of a deep learning based pulmonary hypertension screening algorithm using a digital stethoscope
topic deep learning
digital stethoscopes
pulmonary hypertension
url https://www.ahajournals.org/doi/10.1161/JAHA.124.036882
work_keys_str_mv AT lingguo developmentandevaluationofadeeplearningbasedpulmonaryhypertensionscreeningalgorithmusingadigitalstethoscope
AT niveditakhobragade developmentandevaluationofadeeplearningbasedpulmonaryhypertensionscreeningalgorithmusingadigitalstethoscope
AT spencerkieu developmentandevaluationofadeeplearningbasedpulmonaryhypertensionscreeningalgorithmusingadigitalstethoscope
AT sulemanilyas developmentandevaluationofadeeplearningbasedpulmonaryhypertensionscreeningalgorithmusingadigitalstethoscope
AT prestonnnicely developmentandevaluationofadeeplearningbasedpulmonaryhypertensionscreeningalgorithmusingadigitalstethoscope
AT emmanuelkasiedu developmentandevaluationofadeeplearningbasedpulmonaryhypertensionscreeningalgorithmusingadigitalstethoscope
AT fabiovlima developmentandevaluationofadeeplearningbasedpulmonaryhypertensionscreeningalgorithmusingadigitalstethoscope
AT carolinecurrie developmentandevaluationofadeeplearningbasedpulmonaryhypertensionscreeningalgorithmusingadigitalstethoscope
AT emileighlastowski developmentandevaluationofadeeplearningbasedpulmonaryhypertensionscreeningalgorithmusingadigitalstethoscope
AT gauravchoudhary developmentandevaluationofadeeplearningbasedpulmonaryhypertensionscreeningalgorithmusingadigitalstethoscope