Deep semantic lung segmentation for tracking potential pulmonary perfusion biomarkers in chronic obstructive pulmonary disease (COPD): The multi-ethnic study of atherosclerosis COPD study

Publication type:
Journal article
Metadata:
Autoren
  • Hinrich B Winther
  • Marcel Gutberlet
  • Christian Hundt
  • Till F Kaireit
  • Tawfik Moher Alsady
  • Bertil Schmidt
  • Frank Wacker
  • Yanping Sun
  • Sabine Dettmer
  • Sabine K Maschke
  • Jan B Hinrichs
  • Sachin Jambawalikar
  • Martin R Prince
  • R Graham Barr
  • Jens Vogel-Claussen
Autoren-URL
https://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=fis-test-1&SrcAuth=WosAPI&KeyUT=WOS:000506450300023&DestLinkType=FullRecord&DestApp=WOS_CPL
DOI
10.1002/jmri.26853
eISSN
1522-2586
Externe Identifier
  • Clarivate Analytics Document Solution ID: KB4EG
  • PubMed Identifier: 31276264
ISSN
1053-1807
Ausgabe der Veröffentlichung
2
Zeitschrift
JOURNAL OF MAGNETIC RESONANCE IMAGING
Schlüsselwörter
  • lung perfusion
  • 4D DCE
  • COPD
  • deep semantic segmentation
  • nonparametric deconvolution
Paginierung
571 - 579
Datum der Veröffentlichung
2020
Status
Published
Titel
Deep semantic lung segmentation for tracking potential pulmonary perfusion biomarkers in chronic obstructive pulmonary disease (COPD): The multi-ethnic study of atherosclerosis COPD study
Sub types
  • Article
Ausgabe der Zeitschrift
51

Data source: Web of Science (Lite)

Other metadata sources:
Abstract
<jats:sec><jats:title>Background</jats:title><jats:p>Chronic obstructive pulmonary disease (COPD) is associated with high morbidity and mortality. Identification of imaging biomarkers for phenotyping is necessary for future treatment and therapy monitoring. However, translation of visual analytic pipelines into clinics or their use in large‐scale studies is significantly slowed by time‐consuming postprocessing steps.</jats:p></jats:sec><jats:sec><jats:title>Purpose</jats:title><jats:p>To implement an automated tool chain for regional quantification of pulmonary microvascular blood flow in order to reduce analysis time and user variability.</jats:p></jats:sec><jats:sec><jats:title>Study Type</jats:title><jats:p>Prospective.</jats:p></jats:sec><jats:sec><jats:title>Population</jats:title><jats:p>In all, 90 MRI scans of 63 patients, of which 31 had a COPD with a mean Global Initiative for Chronic Obstructive Lung Disease status of 1.9 ± 0.64 (μ ± σ).</jats:p></jats:sec><jats:sec><jats:title>Field Strength/Sequence</jats:title><jats:p>1.5T dynamic gadolinium‐enhanced MRI measurement using 4D dynamic contrast material‐enhanced (DCE) time‐resolved angiography acquired in a single breath‐hold in inspiration. [Correction added on August 20, 2019, after first online publication: The field strength in the preceding sentence was corrected.]</jats:p></jats:sec><jats:sec><jats:title>Assessment</jats:title><jats:p>We built a 3D convolutional neural network for semantic segmentation using 29 manually segmented perfusion maps. All five lobes of the lung are denoted, including the middle lobe. Evaluation was performed on 61 independent cases from two sites of the Multi‐Ethnic Study of Arteriosclerosis (MESA)‐COPD study. We publish our implementation of a model‐free deconvolution filter according to Sourbron et al for 4D DCE MRI scans as open source.</jats:p></jats:sec><jats:sec><jats:title>Statistical Test</jats:title><jats:p>Cross‐validation 29/61 (# training / # testing), intraclass correlation coefficient (ICC), Spearman ρ, Pearson <jats:italic>r</jats:italic>, Sørensen–Dice coefficient, and overlap.</jats:p></jats:sec><jats:sec><jats:title>Results</jats:title><jats:p>Segmentations and derived clinical parameters were processed in ~90 seconds per case on a Xeon E5‐2637v4 workstation with Tesla P40 GPUs. Clinical parameters and predicted segmentations exhibit high concordance with the ground truth regarding median perfusion for all lobes with an ICC of 0.99 and a Sørensen–Dice coefficient of 93.4 ± 2.8 (μ ± σ).</jats:p></jats:sec><jats:sec><jats:title>Data Conclusion</jats:title><jats:p>We present a robust end‐to‐end pipeline that allows for the extraction of perfusion‐based biomarkers for all lung lobes in 4D DCE MRI scans by combining model‐free deconvolution with deep learning.</jats:p><jats:p><jats:bold>Level of Evidence:</jats:bold> 3</jats:p><jats:p><jats:bold>Technical Efficacy:</jats:bold> Stage 2</jats:p><jats:p>J. Magn. Reson. Imaging 2020;51:571–579.</jats:p></jats:sec>
Autoren
  • Hinrich B Winther
  • Marcel Gutberlet
  • Christian Hundt
  • Till F Kaireit
  • Tawfik Moher Alsady
  • Bertil Schmidt
  • Frank Wacker
  • Yanping Sun
  • Sabine Dettmer
  • Sabine K Maschke
  • Jan B Hinrichs
  • Sachin Jambawalikar
  • Martin R Prince
  • R Graham Barr
  • Jens Vogel‐Claussen
DOI
10.1002/jmri.26853
eISSN
1522-2586
ISSN
1053-1807
Ausgabe der Veröffentlichung
2
Zeitschrift
Journal of Magnetic Resonance Imaging
Sprache
en
Online publication date
2019
Paginierung
571 - 579
Datum der Veröffentlichung
2020
Status
Published
Herausgeber
Wiley
Herausgeber URL
http://dx.doi.org/10.1002/jmri.26853
Datum der Datenerfassung
2023
Titel
Deep semantic lung segmentation for tracking potential pulmonary perfusion biomarkers in chronic obstructive pulmonary disease (COPD): The multi‐ethnic study of atherosclerosis COPD study
Ausgabe der Zeitschrift
51

Data source: Crossref

Abstract
<h4>Background</h4>Chronic obstructive pulmonary disease (COPD) is associated with high morbidity and mortality. Identification of imaging biomarkers for phenotyping is necessary for future treatment and therapy monitoring. However, translation of visual analytic pipelines into clinics or their use in large-scale studies is significantly slowed by time-consuming postprocessing steps.<h4>Purpose</h4>To implement an automated tool chain for regional quantification of pulmonary microvascular blood flow in order to reduce analysis time and user variability.<h4>Study type</h4>Prospective.<h4>Population</h4>In all, 90 MRI scans of 63 patients, of which 31 had a COPD with a mean Global Initiative for Chronic Obstructive Lung Disease status of 1.9 ± 0.64 (μ ± σ).<h4>Field strength/sequence</h4>1.5T dynamic gadolinium-enhanced MRI measurement using 4D dynamic contrast material-enhanced (DCE) time-resolved angiography acquired in a single breath-hold in inspiration. [Correction added on August 20, 2019, after first online publication: The field strength in the preceding sentence was corrected.] ASSESSMENT: We built a 3D convolutional neural network for semantic segmentation using 29 manually segmented perfusion maps. All five lobes of the lung are denoted, including the middle lobe. Evaluation was performed on 61 independent cases from two sites of the Multi-Ethnic Study of Arteriosclerosis (MESA)-COPD study. We publish our implementation of a model-free deconvolution filter according to Sourbron et al for 4D DCE MRI scans as open source.<h4>Statistical test</h4>Cross-validation 29/61 (# training / # testing), intraclass correlation coefficient (ICC), Spearman ρ, Pearson r, Sørensen-Dice coefficient, and overlap.<h4>Results</h4>Segmentations and derived clinical parameters were processed in ~90 seconds per case on a Xeon E5-2637v4 workstation with Tesla P40 GPUs. Clinical parameters and predicted segmentations exhibit high concordance with the ground truth regarding median perfusion for all lobes with an ICC of 0.99 and a Sørensen-Dice coefficient of 93.4 ± 2.8 (μ ± σ).<h4>Data conclusion</h4>We present a robust end-to-end pipeline that allows for the extraction of perfusion-based biomarkers for all lung lobes in 4D DCE MRI scans by combining model-free deconvolution with deep learning.<h4>Level of evidence</h4>3 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2020;51:571-579.
Addresses
  • Department of Diagnostic and Interventional Radiology, Hannover Medical School, Hannover, Germany.
Autoren
  • Hinrich B Winther
  • Marcel Gutberlet
  • Christian Hundt
  • Till F Kaireit
  • Tawfik Moher Alsady
  • Bertil Schmidt
  • Frank Wacker
  • Yanping Sun
  • Sabine Dettmer
  • Sabine K Maschke
  • Jan B Hinrichs
  • Sachin Jambawalikar
  • Martin R Prince
  • R Graham Barr
  • Jens Vogel-Claussen
DOI
10.1002/jmri.26853
eISSN
1522-2586
Externe Identifier
  • PubMed Identifier: 31276264
Funding acknowledgements
  • NHLBI NIH HHS: N01-HC-95160
  • NHLBI NIH HHS: N01-HC-95161
  • National Heart, Lung, and Blood Institute: N01‐HC‐95165
  • National Heart, Lung, and Blood Institute: N01‐HC‐95167
  • NHLBI NIH HHS: N01-HC-95168
  • NHLBI NIH HHS: N01-HC-95169
  • National Heart, Lung, and Blood Institute: N01‐HC‐95159
  • National Heart, Lung, and Blood Institute: N01‐HC‐95166
  • National Heart, Lung, and Blood Institute: N01‐HC‐95168
  • National Heart, Lung, and Blood Institute: HHSN268201500003I
  • National Heart, Lung, and Blood Institute: N01‐HC‐95162
  • National Heart, Lung, and Blood Institute: N01‐HC‐95163
  • NHLBI NIH HHS: R01-HL093081
  • National Heart, Lung, and Blood Institute: R01‐HL093081
  • National Center for Advancing Translational Sciences: UL1‐TR‐001420
  • NHLBI NIH HHS: N01-HC-95162
  • NHLBI NIH HHS: N01-HC-95166
  • National Center for Advancing Translational Sciences: UL1‐TR‐000040
  • NHLBI NIH HHS: N01-HC-95164
  • National Heart, Lung, and Blood Institute: N01‐HC‐95161
  • National Heart, Lung, and Blood Institute: R01‐HL077612
  • NCATS NIH HHS: UL1-TR-000040
  • NCATS NIH HHS: UL1-TR-001420
  • NHLBI NIH HHS: N01-HC-95163
  • NHLBI NIH HHS: N01-HC-95165
  • NCATS NIH HHS: UL1-TR-001079
  • NHLBI NIH HHS: HHSN268201500003I
  • NHLBI NIH HHS: N01-HC-95159
  • NHLBI NIH HHS: N01-HC-95167
  • National Heart, Lung, and Blood Institute: N01‐HC‐95160
  • National Heart, Lung, and Blood Institute: N01‐HC‐95164
  • National Heart, Lung, and Blood Institute: N01‐HC‐95169
  • NHLBI NIH HHS: R01-HL077612
  • National Center for Advancing Translational Sciences: UL1‐TR‐001079
  • Deutsche Forschungsgemeinschaft: WI 4739/1‐1
Open access
false
ISSN
1053-1807
Ausgabe der Veröffentlichung
2
Zeitschrift
Journal of magnetic resonance imaging : JMRI
Schlüsselwörter
  • Lung
  • Humans
  • Pulmonary Disease, Chronic Obstructive
  • Magnetic Resonance Imaging
  • Prospective Studies
  • Perfusion
  • Semantics
  • Atherosclerosis
  • Biomarkers
Sprache
eng
Medium
Print-Electronic
Online publication date
2019
Paginierung
571 - 579
Datum der Veröffentlichung
2020
Status
Published
Datum der Datenerfassung
2019
Titel
Deep semantic lung segmentation for tracking potential pulmonary perfusion biomarkers in chronic obstructive pulmonary disease (COPD): The multi-ethnic study of atherosclerosis COPD study.
Sub types
  • Research Support, Non-U.S. Gov't
  • Journal Article
  • Research Support, N.I.H., Extramural
Ausgabe der Zeitschrift
51

Data source: Europe PubMed Central

Abstract
BACKGROUND: Chronic obstructive pulmonary disease (COPD) is associated with high morbidity and mortality. Identification of imaging biomarkers for phenotyping is necessary for future treatment and therapy monitoring. However, translation of visual analytic pipelines into clinics or their use in large-scale studies is significantly slowed by time-consuming postprocessing steps. PURPOSE: To implement an automated tool chain for regional quantification of pulmonary microvascular blood flow in order to reduce analysis time and user variability. STUDY TYPE: Prospective. POPULATION: In all, 90 MRI scans of 63 patients, of which 31 had a COPD with a mean Global Initiative for Chronic Obstructive Lung Disease status of 1.9 ± 0.64 (μ ± σ). FIELD STRENGTH/SEQUENCE: 1.5T dynamic gadolinium-enhanced MRI measurement using 4D dynamic contrast material-enhanced (DCE) time-resolved angiography acquired in a single breath-hold in inspiration. [Correction added on August 20, 2019, after first online publication: The field strength in the preceding sentence was corrected.] ASSESSMENT: We built a 3D convolutional neural network for semantic segmentation using 29 manually segmented perfusion maps. All five lobes of the lung are denoted, including the middle lobe. Evaluation was performed on 61 independent cases from two sites of the Multi-Ethnic Study of Arteriosclerosis (MESA)-COPD study. We publish our implementation of a model-free deconvolution filter according to Sourbron et al for 4D DCE MRI scans as open source. STATISTICAL TEST: Cross-validation 29/61 (# training / # testing), intraclass correlation coefficient (ICC), Spearman ρ, Pearson r, Sørensen-Dice coefficient, and overlap. RESULTS: Segmentations and derived clinical parameters were processed in ~90 seconds per case on a Xeon E5-2637v4 workstation with Tesla P40 GPUs. Clinical parameters and predicted segmentations exhibit high concordance with the ground truth regarding median perfusion for all lobes with an ICC of 0.99 and a Sørensen-Dice coefficient of 93.4 ± 2.8 (μ ± σ). DATA CONCLUSION: We present a robust end-to-end pipeline that allows for the extraction of perfusion-based biomarkers for all lung lobes in 4D DCE MRI scans by combining model-free deconvolution with deep learning. LEVEL OF EVIDENCE: 3 Technical Efficacy: Stage 2 J. Magn. Reson. Imaging 2020;51:571-579.
Date of acceptance
2019
Autoren
  • Hinrich B Winther
  • Marcel Gutberlet
  • Christian Hundt
  • Till F Kaireit
  • Tawfik Moher Alsady
  • Bertil Schmidt
  • Frank Wacker
  • Yanping Sun
  • Sabine Dettmer
  • Sabine K Maschke
  • Jan B Hinrichs
  • Sachin Jambawalikar
  • Martin R Prince
  • R Graham Barr
  • Jens Vogel-Claussen
Autoren-URL
https://www.ncbi.nlm.nih.gov/pubmed/31276264
DOI
10.1002/jmri.26853
eISSN
1522-2586
Funding acknowledgements
  • NHLBI NIH HHS: N01-HC-95159
  • NHLBI NIH HHS: N01-HC-95166
  • NCATS NIH HHS: UL1-TR-001420
  • NHLBI NIH HHS: N01-HC-95160
  • NHLBI NIH HHS: N01-HC-95167
  • NCATS NIH HHS: UL1-TR-001079
  • NHLBI NIH HHS: N01-HC-95168
  • NHLBI NIH HHS: N01-HC-95162
  • NHLBI NIH HHS: N01-HC-95164
  • NHLBI NIH HHS: N01-HC-95165
  • NHLBI NIH HHS: R01-HL093081
  • NHLBI NIH HHS: R01-HL077612
  • NHLBI NIH HHS: HHSN268201500003I
  • NHLBI NIH HHS: N01-HC-95163
  • NCATS NIH HHS: UL1-TR-000040
  • NHLBI NIH HHS: N01-HC-95169
  • NHLBI NIH HHS: N01-HC-95161
Ausgabe der Veröffentlichung
2
Zeitschrift
J Magn Reson Imaging
Schlüsselwörter
  • 4D DCE
  • COPD
  • deep semantic segmentation
  • lung perfusion
  • nonparametric deconvolution
  • Atherosclerosis
  • Biomarkers
  • Humans
  • Lung
  • Magnetic Resonance Imaging
  • Perfusion
  • Prospective Studies
  • Pulmonary Disease, Chronic Obstructive
  • Semantics
Sprache
eng
Country
United States
Paginierung
571 - 579
Datum der Veröffentlichung
2020
Status
Published
Datum, an dem der Datensatz öffentlich gemacht wurde
2021
Titel
Deep semantic lung segmentation for tracking potential pulmonary perfusion biomarkers in chronic obstructive pulmonary disease (COPD): The multi-ethnic study of atherosclerosis COPD study.
Sub types
  • Journal Article
  • Research Support, N.I.H., Extramural
  • Research Support, Non-U.S. Gov't
Ausgabe der Zeitschrift
51

Data source: PubMed

Beziehungen:
Property of

Tools