Mean pulmonary artery pressure prediction with explainable multi-view cardiovascular magnetic resonance cine series deep learning model

ABSTRACT: Background: Pulmonary hypertension (PH) is a heterogeneous condition and regardless of etiology impacts negatively on survival. Diagnosis of PH is based on hemodynamic parameters measured invasively at right heart catheterization (RHC); however, a non-invasive alternative would be clinica...

Full description

Saved in:
Bibliographic Details
Main Authors: Li-Hsin Cheng, Xiaowu Sun, Charlie Elliot, Robin Condliffe, David G. Kiely, Samer Alabed, Andrew J. Swift, Rob J. van der Geest, David G Kiely, Lisa Watson, Iain Armstrong, Catherine Billings, Athanasios Charalampopoulos, Abdul Hameed, Neil Hamilton, Judith Hurdman, Allan Lawrie, Robert A Lewis, Smitha Rajaram, Alex Rothman, Andy J. Swift, Steven Wood, AA Roger Thompson, Jim Wild
Format: Article
Language:English
Published: Elsevier 2025-01-01
Series:Journal of Cardiovascular Magnetic Resonance
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S1097664724011608
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849333102004928512
author Li-Hsin Cheng
Xiaowu Sun
Charlie Elliot
Robin Condliffe
David G. Kiely
Samer Alabed
Andrew J. Swift
Rob J. van der Geest
David G Kiely
Lisa Watson
Iain Armstrong
Catherine Billings
Athanasios Charalampopoulos
Robin Condliffe
Charlie Elliot
Abdul Hameed
Neil Hamilton
Judith Hurdman
Allan Lawrie
Robert A Lewis
Smitha Rajaram
Alex Rothman
Andy J. Swift
Steven Wood
AA Roger Thompson
Jim Wild
author_facet Li-Hsin Cheng
Xiaowu Sun
Charlie Elliot
Robin Condliffe
David G. Kiely
Samer Alabed
Andrew J. Swift
Rob J. van der Geest
David G Kiely
Lisa Watson
Iain Armstrong
Catherine Billings
Athanasios Charalampopoulos
Robin Condliffe
Charlie Elliot
Abdul Hameed
Neil Hamilton
Judith Hurdman
Allan Lawrie
Robert A Lewis
Smitha Rajaram
Alex Rothman
Andy J. Swift
Steven Wood
AA Roger Thompson
Jim Wild
author_sort Li-Hsin Cheng
collection DOAJ
description ABSTRACT: Background: Pulmonary hypertension (PH) is a heterogeneous condition and regardless of etiology impacts negatively on survival. Diagnosis of PH is based on hemodynamic parameters measured invasively at right heart catheterization (RHC); however, a non-invasive alternative would be clinically valuable. Our aim was to estimate RHC parameters non-invasively from cardiac magnetic resonance (MR) data using deep learning models and to identify key contributing imaging features. Methods: We constructed an explainable convolutional neural network (CNN) taking cardiac MR cine series from four different views as input to predict mean pulmonary artery pressure (mPAP). The model was trained and evaluated on 1646 examinations. The model’s attention weight and predictive performance associated with each frame, view, or phase were used to judge its importance. Additionally, the importance of each cardiac chamber was inferred by perturbing part of the input pixels. Results: The model achieved a Pearson correlation coefficient of 0.80 and R2 of 0.64 in predicting mPAP and identified the right ventricle region on short-axis view to be especially informative. Conclusion: Hemodynamic parameters can be estimated non-invasively with a CNN, using MR cine series from four views, revealing key contributing features at the same time.
format Article
id doaj-art-199c341d9c4a481eac0c849a7681bfc3
institution Kabale University
issn 1097-6647
language English
publishDate 2025-01-01
publisher Elsevier
record_format Article
series Journal of Cardiovascular Magnetic Resonance
spelling doaj-art-199c341d9c4a481eac0c849a7681bfc32025-08-20T03:45:59ZengElsevierJournal of Cardiovascular Magnetic Resonance1097-66472025-01-0127110113310.1016/j.jocmr.2024.101133Mean pulmonary artery pressure prediction with explainable multi-view cardiovascular magnetic resonance cine series deep learning modelLi-Hsin Cheng0Xiaowu Sun1Charlie Elliot2Robin Condliffe3David G. Kiely4Samer Alabed5Andrew J. Swift6Rob J. van der Geest7David G Kiely8Lisa Watson9Iain Armstrong10Catherine Billings11Athanasios Charalampopoulos12Robin Condliffe13Charlie Elliot14Abdul Hameed15Neil Hamilton16Judith Hurdman17Allan Lawrie18Robert A Lewis19Smitha Rajaram20Alex Rothman21Andy J. Swift22Steven Wood23AA Roger Thompson24Jim Wild25Division of Image Processing (LKEB), Department of Radiology, Leiden University Medical, Center, the NetherlandsDivision of Image Processing (LKEB), Department of Radiology, Leiden University Medical, Center, the NetherlandsSheffield Pulmonary Vascular Disease Unit, Sheffield Teaching Hospitals NHS Foundation, Trust, UKSheffield Pulmonary Vascular Disease Unit, Sheffield Teaching Hospitals NHS Foundation, Trust, UKSheffield Pulmonary Vascular Disease Unit, Sheffield Teaching Hospitals NHS Foundation, Trust, UK; National Institute for Health and Care Research (NIHR) Sheffield Biomedical Research Centre, Sheffield, UKDepartment of Infection, Immunity and Cardiovascular Disease, University of Sheffield, UKDepartment of Infection, Immunity and Cardiovascular Disease, University of Sheffield, UKDivision of Image Processing (LKEB), Department of Radiology, Leiden University Medical, Center, the Netherlands; Corresponding author.Sheffield Pulmonary Vascular Disease Unit, Sheffield Teaching Hospitals NHS Foundation Trust, Royal Hallamshire Hospital, Sheffield UK; National Institute for Health and Care Research (NIHR) Sheffield Biomedical Research Centre (NIHR203321)Sheffield Pulmonary Vascular Disease Unit, Sheffield Teaching Hospitals NHS Foundation Trust, Royal Hallamshire Hospital, Sheffield UK; National Institute for Health and Care Research (NIHR) Sheffield Biomedical Research Centre (NIHR203321)Sheffield Pulmonary Vascular Disease Unit, Sheffield Teaching Hospitals NHS Foundation Trust, Royal Hallamshire Hospital, Sheffield UK; National Institute for Health and Care Research (NIHR) Sheffield Biomedical Research Centre (NIHR203321)Sheffield Pulmonary Vascular Disease Unit, Sheffield Teaching Hospitals NHS Foundation Trust, Royal Hallamshire Hospital, Sheffield UK; National Institute for Health and Care Research (NIHR) Sheffield Biomedical Research Centre (NIHR203321)Sheffield Pulmonary Vascular Disease Unit, Sheffield Teaching Hospitals NHS Foundation Trust, Royal Hallamshire Hospital, Sheffield UK; National Institute for Health and Care Research (NIHR) Sheffield Biomedical Research Centre (NIHR203321)Sheffield Pulmonary Vascular Disease Unit, Sheffield Teaching Hospitals NHS Foundation Trust, Royal Hallamshire Hospital, Sheffield UK; National Institute for Health and Care Research (NIHR) Sheffield Biomedical Research Centre (NIHR203321)Sheffield Pulmonary Vascular Disease Unit, Sheffield Teaching Hospitals NHS Foundation Trust, Royal Hallamshire Hospital, Sheffield UK; National Institute for Health and Care Research (NIHR) Sheffield Biomedical Research Centre (NIHR203321)Sheffield Pulmonary Vascular Disease Unit, Sheffield Teaching Hospitals NHS Foundation Trust, Royal Hallamshire Hospital, Sheffield UK; National Institute for Health and Care Research (NIHR) Sheffield Biomedical Research Centre (NIHR203321)Sheffield Pulmonary Vascular Disease Unit, Sheffield Teaching Hospitals NHS Foundation Trust, Royal Hallamshire Hospital, Sheffield UK; National Institute for Health and Care Research (NIHR) Sheffield Biomedical Research Centre (NIHR203321)Sheffield Pulmonary Vascular Disease Unit, Sheffield Teaching Hospitals NHS Foundation Trust, Royal Hallamshire Hospital, Sheffield UK; National Institute for Health and Care Research (NIHR) Sheffield Biomedical Research Centre (NIHR203321)National Institute for Health and Care Research (NIHR) Sheffield Biomedical Research Centre (NIHR203321); University of Sheffield, Sheffield, UKSheffield Pulmonary Vascular Disease Unit, Sheffield Teaching Hospitals NHS Foundation Trust, Royal Hallamshire Hospital, Sheffield UK; National Institute for Health and Care Research (NIHR) Sheffield Biomedical Research Centre (NIHR203321)Sheffield Pulmonary Vascular Disease Unit, Sheffield Teaching Hospitals NHS Foundation Trust, Royal Hallamshire Hospital, Sheffield UK; National Institute for Health and Care Research (NIHR) Sheffield Biomedical Research Centre (NIHR203321)National Institute for Health and Care Research (NIHR) Sheffield Biomedical Research Centre (NIHR203321); University of Sheffield, Sheffield, UKNational Institute for Health and Care Research (NIHR) Sheffield Biomedical Research Centre (NIHR203321); University of Sheffield, Sheffield, UKSheffield Pulmonary Vascular Disease Unit, Sheffield Teaching Hospitals NHS Foundation Trust, Royal Hallamshire Hospital, Sheffield UK; National Institute for Health and Care Research (NIHR) Sheffield Biomedical Research Centre (NIHR203321)National Institute for Health and Care Research (NIHR) Sheffield Biomedical Research Centre (NIHR203321); University of Sheffield, Sheffield, UKNational Institute for Health and Care Research (NIHR) Sheffield Biomedical Research Centre (NIHR203321); University of Sheffield, Sheffield, UKABSTRACT: Background: Pulmonary hypertension (PH) is a heterogeneous condition and regardless of etiology impacts negatively on survival. Diagnosis of PH is based on hemodynamic parameters measured invasively at right heart catheterization (RHC); however, a non-invasive alternative would be clinically valuable. Our aim was to estimate RHC parameters non-invasively from cardiac magnetic resonance (MR) data using deep learning models and to identify key contributing imaging features. Methods: We constructed an explainable convolutional neural network (CNN) taking cardiac MR cine series from four different views as input to predict mean pulmonary artery pressure (mPAP). The model was trained and evaluated on 1646 examinations. The model’s attention weight and predictive performance associated with each frame, view, or phase were used to judge its importance. Additionally, the importance of each cardiac chamber was inferred by perturbing part of the input pixels. Results: The model achieved a Pearson correlation coefficient of 0.80 and R2 of 0.64 in predicting mPAP and identified the right ventricle region on short-axis view to be especially informative. Conclusion: Hemodynamic parameters can be estimated non-invasively with a CNN, using MR cine series from four views, revealing key contributing features at the same time.http://www.sciencedirect.com/science/article/pii/S1097664724011608Mean pulmonary artery pressurePulmonary hypertensionMulti-view cardiac MRDeep learningExplainable AI
spellingShingle Li-Hsin Cheng
Xiaowu Sun
Charlie Elliot
Robin Condliffe
David G. Kiely
Samer Alabed
Andrew J. Swift
Rob J. van der Geest
David G Kiely
Lisa Watson
Iain Armstrong
Catherine Billings
Athanasios Charalampopoulos
Robin Condliffe
Charlie Elliot
Abdul Hameed
Neil Hamilton
Judith Hurdman
Allan Lawrie
Robert A Lewis
Smitha Rajaram
Alex Rothman
Andy J. Swift
Steven Wood
AA Roger Thompson
Jim Wild
Mean pulmonary artery pressure prediction with explainable multi-view cardiovascular magnetic resonance cine series deep learning model
Journal of Cardiovascular Magnetic Resonance
Mean pulmonary artery pressure
Pulmonary hypertension
Multi-view cardiac MR
Deep learning
Explainable AI
title Mean pulmonary artery pressure prediction with explainable multi-view cardiovascular magnetic resonance cine series deep learning model
title_full Mean pulmonary artery pressure prediction with explainable multi-view cardiovascular magnetic resonance cine series deep learning model
title_fullStr Mean pulmonary artery pressure prediction with explainable multi-view cardiovascular magnetic resonance cine series deep learning model
title_full_unstemmed Mean pulmonary artery pressure prediction with explainable multi-view cardiovascular magnetic resonance cine series deep learning model
title_short Mean pulmonary artery pressure prediction with explainable multi-view cardiovascular magnetic resonance cine series deep learning model
title_sort mean pulmonary artery pressure prediction with explainable multi view cardiovascular magnetic resonance cine series deep learning model
topic Mean pulmonary artery pressure
Pulmonary hypertension
Multi-view cardiac MR
Deep learning
Explainable AI
url http://www.sciencedirect.com/science/article/pii/S1097664724011608
work_keys_str_mv AT lihsincheng meanpulmonaryarterypressurepredictionwithexplainablemultiviewcardiovascularmagneticresonancecineseriesdeeplearningmodel
AT xiaowusun meanpulmonaryarterypressurepredictionwithexplainablemultiviewcardiovascularmagneticresonancecineseriesdeeplearningmodel
AT charlieelliot meanpulmonaryarterypressurepredictionwithexplainablemultiviewcardiovascularmagneticresonancecineseriesdeeplearningmodel
AT robincondliffe meanpulmonaryarterypressurepredictionwithexplainablemultiviewcardiovascularmagneticresonancecineseriesdeeplearningmodel
AT davidgkiely meanpulmonaryarterypressurepredictionwithexplainablemultiviewcardiovascularmagneticresonancecineseriesdeeplearningmodel
AT sameralabed meanpulmonaryarterypressurepredictionwithexplainablemultiviewcardiovascularmagneticresonancecineseriesdeeplearningmodel
AT andrewjswift meanpulmonaryarterypressurepredictionwithexplainablemultiviewcardiovascularmagneticresonancecineseriesdeeplearningmodel
AT robjvandergeest meanpulmonaryarterypressurepredictionwithexplainablemultiviewcardiovascularmagneticresonancecineseriesdeeplearningmodel
AT davidgkiely meanpulmonaryarterypressurepredictionwithexplainablemultiviewcardiovascularmagneticresonancecineseriesdeeplearningmodel
AT lisawatson meanpulmonaryarterypressurepredictionwithexplainablemultiviewcardiovascularmagneticresonancecineseriesdeeplearningmodel
AT iainarmstrong meanpulmonaryarterypressurepredictionwithexplainablemultiviewcardiovascularmagneticresonancecineseriesdeeplearningmodel
AT catherinebillings meanpulmonaryarterypressurepredictionwithexplainablemultiviewcardiovascularmagneticresonancecineseriesdeeplearningmodel
AT athanasioscharalampopoulos meanpulmonaryarterypressurepredictionwithexplainablemultiviewcardiovascularmagneticresonancecineseriesdeeplearningmodel
AT robincondliffe meanpulmonaryarterypressurepredictionwithexplainablemultiviewcardiovascularmagneticresonancecineseriesdeeplearningmodel
AT charlieelliot meanpulmonaryarterypressurepredictionwithexplainablemultiviewcardiovascularmagneticresonancecineseriesdeeplearningmodel
AT abdulhameed meanpulmonaryarterypressurepredictionwithexplainablemultiviewcardiovascularmagneticresonancecineseriesdeeplearningmodel
AT neilhamilton meanpulmonaryarterypressurepredictionwithexplainablemultiviewcardiovascularmagneticresonancecineseriesdeeplearningmodel
AT judithhurdman meanpulmonaryarterypressurepredictionwithexplainablemultiviewcardiovascularmagneticresonancecineseriesdeeplearningmodel
AT allanlawrie meanpulmonaryarterypressurepredictionwithexplainablemultiviewcardiovascularmagneticresonancecineseriesdeeplearningmodel
AT robertalewis meanpulmonaryarterypressurepredictionwithexplainablemultiviewcardiovascularmagneticresonancecineseriesdeeplearningmodel
AT smitharajaram meanpulmonaryarterypressurepredictionwithexplainablemultiviewcardiovascularmagneticresonancecineseriesdeeplearningmodel
AT alexrothman meanpulmonaryarterypressurepredictionwithexplainablemultiviewcardiovascularmagneticresonancecineseriesdeeplearningmodel
AT andyjswift meanpulmonaryarterypressurepredictionwithexplainablemultiviewcardiovascularmagneticresonancecineseriesdeeplearningmodel
AT stevenwood meanpulmonaryarterypressurepredictionwithexplainablemultiviewcardiovascularmagneticresonancecineseriesdeeplearningmodel
AT aarogerthompson meanpulmonaryarterypressurepredictionwithexplainablemultiviewcardiovascularmagneticresonancecineseriesdeeplearningmodel
AT jimwild meanpulmonaryarterypressurepredictionwithexplainablemultiviewcardiovascularmagneticresonancecineseriesdeeplearningmodel