B-line detection and localization by means of deep learning: preliminary in-vitro results

Ruud J.G. van Sloun, Libertario Demi

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

Abstract

Lung ultrasound imaging is nowadays receiving growing attention. In fact, the analysis of specific artefactual patterns reveals important diagnostic information. A- and B-line artifacts are particularly important. A-lines are generally considered a sign of a healthy lung, while B-line artifacts correlate with a large variety of pathological conditions. B-lines have been found to indicate an increase in extravascular lung water, the presence of interstitial lung diseases, non-cardiogenic lung edema, interstitial pneumonia and lung contusion. The capability to accurately and objectively detect and localize B-lines in a lung ultrasound video is therefore of great clinical interest. In this paper, we present a method aimed at supporting clinicians in the analysis of ultrasound videos by automatically detecting and localizing B-lines, in real-time. To this end, modern deep learning strategies have been used and a fully convolutional neural network has been trained to detect B-lines in B-mode images of dedicated ultrasound phantoms. Furthermore, neural attention maps have been calculated to visualize which components in the image triggered the network, thereby offering simultaneous weakly-supervised localization. An accuracy, sensitivity, specificity, negative and positive predictive value equal to 0.917, 0.915, 0.918, 0.950 and 0.864 were achieved in-vitro using data from dedicated lung-mimicking phantoms, respectively.

LanguageEnglish
Title of host publicationImage Analysis and Recognition - 16th International Conference, ICIAR 2019, Proceedings
EditorsFakhri Karray, Alfred Yu, Aurélio Campilho
Place of PublicationCham
PublisherSpringer
Pages418-424
Number of pages7
ISBN (Electronic)978-3-030-27202-9
ISBN (Print)978-3-030-27201-2
DOIs
StatePublished - 8 Aug 2019
Event16th International Conference on Image Analysis and Recognition, ICIAR 2019 - Waterloo, Canada
Duration: 27 Aug 201929 Aug 2019

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11662 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference16th International Conference on Image Analysis and Recognition, ICIAR 2019
CountryCanada
CityWaterloo
Period27/08/1929/08/19

Fingerprint

Line Detection
Lung
Ultrasonics
Ultrasound
Line
Pulmonary diseases
Phantom
Neural networks
Imaging techniques
Learning Strategies
Learning
Deep learning
Correlate
Specificity
Water
Diagnostics
Imaging
Neural Networks
Real-time

Keywords

  • B-lines
  • Deep learning
  • Image analysis
  • Lung ultrasound

Cite this

van Sloun, R. J. G., & Demi, L. (2019). B-line detection and localization by means of deep learning: preliminary in-vitro results. In F. Karray, A. Yu, & A. Campilho (Eds.), Image Analysis and Recognition - 16th International Conference, ICIAR 2019, Proceedings (pp. 418-424). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11662 LNCS). Cham: Springer. DOI: 10.1007/978-3-030-27202-9_38
van Sloun, Ruud J.G. ; Demi, Libertario. / B-line detection and localization by means of deep learning : preliminary in-vitro results. Image Analysis and Recognition - 16th International Conference, ICIAR 2019, Proceedings. editor / Fakhri Karray ; Alfred Yu ; Aurélio Campilho. Cham : Springer, 2019. pp. 418-424 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{6bf445a0b04e4f22abc320d9bc7ef4da,
title = "B-line detection and localization by means of deep learning: preliminary in-vitro results",
abstract = "Lung ultrasound imaging is nowadays receiving growing attention. In fact, the analysis of specific artefactual patterns reveals important diagnostic information. A- and B-line artifacts are particularly important. A-lines are generally considered a sign of a healthy lung, while B-line artifacts correlate with a large variety of pathological conditions. B-lines have been found to indicate an increase in extravascular lung water, the presence of interstitial lung diseases, non-cardiogenic lung edema, interstitial pneumonia and lung contusion. The capability to accurately and objectively detect and localize B-lines in a lung ultrasound video is therefore of great clinical interest. In this paper, we present a method aimed at supporting clinicians in the analysis of ultrasound videos by automatically detecting and localizing B-lines, in real-time. To this end, modern deep learning strategies have been used and a fully convolutional neural network has been trained to detect B-lines in B-mode images of dedicated ultrasound phantoms. Furthermore, neural attention maps have been calculated to visualize which components in the image triggered the network, thereby offering simultaneous weakly-supervised localization. An accuracy, sensitivity, specificity, negative and positive predictive value equal to 0.917, 0.915, 0.918, 0.950 and 0.864 were achieved in-vitro using data from dedicated lung-mimicking phantoms, respectively.",
keywords = "B-lines, Deep learning, Image analysis, Lung ultrasound",
author = "{van Sloun}, {Ruud J.G.} and Libertario Demi",
year = "2019",
month = "8",
day = "8",
doi = "10.1007/978-3-030-27202-9_38",
language = "English",
isbn = "978-3-030-27201-2",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer",
pages = "418--424",
editor = "Fakhri Karray and Alfred Yu and Aur{\'e}lio Campilho",
booktitle = "Image Analysis and Recognition - 16th International Conference, ICIAR 2019, Proceedings",
address = "Germany",

}

van Sloun, RJG & Demi, L 2019, B-line detection and localization by means of deep learning: preliminary in-vitro results. in F Karray, A Yu & A Campilho (eds), Image Analysis and Recognition - 16th International Conference, ICIAR 2019, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11662 LNCS, Springer, Cham, pp. 418-424, 16th International Conference on Image Analysis and Recognition, ICIAR 2019, Waterloo, Canada, 27/08/19. DOI: 10.1007/978-3-030-27202-9_38

B-line detection and localization by means of deep learning : preliminary in-vitro results. / van Sloun, Ruud J.G.; Demi, Libertario.

Image Analysis and Recognition - 16th International Conference, ICIAR 2019, Proceedings. ed. / Fakhri Karray; Alfred Yu; Aurélio Campilho. Cham : Springer, 2019. p. 418-424 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11662 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

TY - GEN

T1 - B-line detection and localization by means of deep learning

T2 - preliminary in-vitro results

AU - van Sloun,Ruud J.G.

AU - Demi,Libertario

PY - 2019/8/8

Y1 - 2019/8/8

N2 - Lung ultrasound imaging is nowadays receiving growing attention. In fact, the analysis of specific artefactual patterns reveals important diagnostic information. A- and B-line artifacts are particularly important. A-lines are generally considered a sign of a healthy lung, while B-line artifacts correlate with a large variety of pathological conditions. B-lines have been found to indicate an increase in extravascular lung water, the presence of interstitial lung diseases, non-cardiogenic lung edema, interstitial pneumonia and lung contusion. The capability to accurately and objectively detect and localize B-lines in a lung ultrasound video is therefore of great clinical interest. In this paper, we present a method aimed at supporting clinicians in the analysis of ultrasound videos by automatically detecting and localizing B-lines, in real-time. To this end, modern deep learning strategies have been used and a fully convolutional neural network has been trained to detect B-lines in B-mode images of dedicated ultrasound phantoms. Furthermore, neural attention maps have been calculated to visualize which components in the image triggered the network, thereby offering simultaneous weakly-supervised localization. An accuracy, sensitivity, specificity, negative and positive predictive value equal to 0.917, 0.915, 0.918, 0.950 and 0.864 were achieved in-vitro using data from dedicated lung-mimicking phantoms, respectively.

AB - Lung ultrasound imaging is nowadays receiving growing attention. In fact, the analysis of specific artefactual patterns reveals important diagnostic information. A- and B-line artifacts are particularly important. A-lines are generally considered a sign of a healthy lung, while B-line artifacts correlate with a large variety of pathological conditions. B-lines have been found to indicate an increase in extravascular lung water, the presence of interstitial lung diseases, non-cardiogenic lung edema, interstitial pneumonia and lung contusion. The capability to accurately and objectively detect and localize B-lines in a lung ultrasound video is therefore of great clinical interest. In this paper, we present a method aimed at supporting clinicians in the analysis of ultrasound videos by automatically detecting and localizing B-lines, in real-time. To this end, modern deep learning strategies have been used and a fully convolutional neural network has been trained to detect B-lines in B-mode images of dedicated ultrasound phantoms. Furthermore, neural attention maps have been calculated to visualize which components in the image triggered the network, thereby offering simultaneous weakly-supervised localization. An accuracy, sensitivity, specificity, negative and positive predictive value equal to 0.917, 0.915, 0.918, 0.950 and 0.864 were achieved in-vitro using data from dedicated lung-mimicking phantoms, respectively.

KW - B-lines

KW - Deep learning

KW - Image analysis

KW - Lung ultrasound

UR - http://www.scopus.com/inward/record.url?scp=85071492610&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-27202-9_38

DO - 10.1007/978-3-030-27202-9_38

M3 - Conference contribution

SN - 978-3-030-27201-2

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 418

EP - 424

BT - Image Analysis and Recognition - 16th International Conference, ICIAR 2019, Proceedings

PB - Springer

CY - Cham

ER -

van Sloun RJG, Demi L. B-line detection and localization by means of deep learning: preliminary in-vitro results. In Karray F, Yu A, Campilho A, editors, Image Analysis and Recognition - 16th International Conference, ICIAR 2019, Proceedings. Cham: Springer. 2019. p. 418-424. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). Available from, DOI: 10.1007/978-3-030-27202-9_38