Combining multiple expert annotations using semi-supervised learning and graph cuts for Crohn’s disease segmentation

Dwarikanath Mahapatra; Peter J. Schüffler; Jeroen A.W. Tielbeek; Carl A.J. Puylaert; Jesica C. Makanyanga; Alex Menys; Rado Andriantsimiavona; Jaap Stoker; Stuart A. Taylor; Franciscus M. Vos; Joachim M. Buhmann

doi:10.1007/978-3-319-13692-9_13

Combining multiple expert annotations using semi-supervised learning and graph cuts for Crohn’s disease segmentation

Dwarikanath Mahapatra, Peter J. Schüffler, Jeroen A.W. Tielbeek, Carl A.J. Puylaert, Jesica C. Makanyanga, Alex Menys, Rado Andriantsimiavona, Jaap Stoker, Stuart A. Taylor, Franciscus M. Vos, Joachim M. Buhmann

Publikation: Beitrag in Buch/Bericht/Konferenzband › Konferenzbeitrag › Begutachtung

2 Zitate (Scopus)

Abstract

We propose a graph cut (GC) based approach for combining annotations from multiple experts and segmenting Crohns disease (CD) tissues in magnetic resonance (MR) images. Random forest (RF) based semi supervised learning (SSL) predicts missing expert labels while a novel self consistency (SC) score quantifies the reliability of each expert label and also serves as the penalty cost in a second order Markov random field (MRF) cost function. The final consensus label is obtained by GC optimization. Experimental results on synthetic images and real CD patient data show our final segmentation to be more accurate than those obtained by competing methods. It also highlights the effectiveness of SC score in quantifying expert reliability and accuracy of SSL in predicting missing labels.

Originalsprache	Englisch
Titel	Abdominal Imaging
Untertitel	Computational and Clinical Applications - 6th International Workshop, ABDI 2014 held in conjunction with MICCAI 2014
Redakteure/-innen	Hiroyuki Yoshida, Janne J. Näppi, Sanjay Saini
Herausgeber (Verlag)	Springer Verlag
Seiten	139-147
Seitenumfang	9
ISBN (elektronisch)	9783319136912
DOIs	https://doi.org/10.1007/978-3-319-13692-9_13
Publikationsstatus	Veröffentlicht - 2014
Extern publiziert	Ja
Veranstaltung	6th International Workshop on Abdominal Imaging: Computational and Clinical Applications, ABDI 2014 held in conjunction with 17th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2014 - Cambridge, USA/Vereinigte Staaten Dauer: 14 Sept. 2014 → 14 Sept. 2014

Publikationsreihe

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Band	8676
ISSN (Print)	0302-9743
ISSN (elektronisch)	1611-3349

Konferenz

Konferenz	6th International Workshop on Abdominal Imaging: Computational and Clinical Applications, ABDI 2014 held in conjunction with 17th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2014
Land/Gebiet	USA/Vereinigte Staaten
Ort	Cambridge
Zeitraum	14/09/14 → 14/09/14

Zugriff auf Dokument

10.1007/978-3-319-13692-9_13

Andere Dateien und Links

Verknüpfung zur Publikation in Scopus

Dieses zitieren

Mahapatra, D., Schüffler, P. J., Tielbeek, J. A. W., Puylaert, C. A. J., Makanyanga, J. C., Menys, A., Andriantsimiavona, R., Stoker, J., Taylor, S. A., Vos, F. M., & Buhmann, J. M. (2014). Combining multiple expert annotations using semi-supervised learning and graph cuts for Crohn’s disease segmentation. in H. Yoshida, J. J. Näppi, & S. Saini (Hrsg.), Abdominal Imaging: Computational and Clinical Applications - 6th International Workshop, ABDI 2014 held in conjunction with MICCAI 2014 (S. 139-147). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Band 8676). Springer Verlag. https://doi.org/10.1007/978-3-319-13692-9_13

Mahapatra, Dwarikanath ; Schüffler, Peter J. ; Tielbeek, Jeroen A.W. et al. / Combining multiple expert annotations using semi-supervised learning and graph cuts for Crohn’s disease segmentation. Abdominal Imaging: Computational and Clinical Applications - 6th International Workshop, ABDI 2014 held in conjunction with MICCAI 2014. Hrsg. / Hiroyuki Yoshida ; Janne J. Näppi ; Sanjay Saini. Springer Verlag, 2014. S. 139-147 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{99885b7061d34f2a87cf408dea2909cb,

title = "Combining multiple expert annotations using semi-supervised learning and graph cuts for Crohn{\textquoteright}s disease segmentation",

abstract = "We propose a graph cut (GC) based approach for combining annotations from multiple experts and segmenting Crohns disease (CD) tissues in magnetic resonance (MR) images. Random forest (RF) based semi supervised learning (SSL) predicts missing expert labels while a novel self consistency (SC) score quantifies the reliability of each expert label and also serves as the penalty cost in a second order Markov random field (MRF) cost function. The final consensus label is obtained by GC optimization. Experimental results on synthetic images and real CD patient data show our final segmentation to be more accurate than those obtained by competing methods. It also highlights the effectiveness of SC score in quantifying expert reliability and accuracy of SSL in predicting missing labels.",

keywords = "Crohn{\textquoteright}s disease, Graph cut, Magnetic resonance, Markov random field, Random forest",

author = "Dwarikanath Mahapatra and Sch{\"u}ffler, {Peter J.} and Tielbeek, {Jeroen A.W.} and Puylaert, {Carl A.J.} and Makanyanga, {Jesica C.} and Alex Menys and Rado Andriantsimiavona and Jaap Stoker and Taylor, {Stuart A.} and Vos, {Franciscus M.} and Buhmann, {Joachim M.}",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing Switzerland 2014.; 6th International Workshop on Abdominal Imaging: Computational and Clinical Applications, ABDI 2014 held in conjunction with 17th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2014 ; Conference date: 14-09-2014 Through 14-09-2014",

year = "2014",

doi = "10.1007/978-3-319-13692-9_13",

language = "English",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "139--147",

editor = "Hiroyuki Yoshida and N{\"a}ppi, {Janne J.} and Sanjay Saini",

booktitle = "Abdominal Imaging",

}

Mahapatra, D, Schüffler, PJ, Tielbeek, JAW, Puylaert, CAJ, Makanyanga, JC, Menys, A, Andriantsimiavona, R, Stoker, J, Taylor, SA, Vos, FM & Buhmann, JM 2014, Combining multiple expert annotations using semi-supervised learning and graph cuts for Crohn’s disease segmentation. in H Yoshida, JJ Näppi & S Saini (Hrsg.), Abdominal Imaging: Computational and Clinical Applications - 6th International Workshop, ABDI 2014 held in conjunction with MICCAI 2014. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Bd. 8676, Springer Verlag, S. 139-147, 6th International Workshop on Abdominal Imaging: Computational and Clinical Applications, ABDI 2014 held in conjunction with 17th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2014, Cambridge, USA/Vereinigte Staaten, 14/09/14. https://doi.org/10.1007/978-3-319-13692-9_13

Combining multiple expert annotations using semi-supervised learning and graph cuts for Crohn’s disease segmentation. / Mahapatra, Dwarikanath; Schüffler, Peter J.; Tielbeek, Jeroen A.W. et al.
Abdominal Imaging: Computational and Clinical Applications - 6th International Workshop, ABDI 2014 held in conjunction with MICCAI 2014. Hrsg. / Hiroyuki Yoshida; Janne J. Näppi; Sanjay Saini. Springer Verlag, 2014. S. 139-147 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Band 8676).

Publikation: Beitrag in Buch/Bericht/Konferenzband › Konferenzbeitrag › Begutachtung

TY - GEN

T1 - Combining multiple expert annotations using semi-supervised learning and graph cuts for Crohn’s disease segmentation

AU - Mahapatra, Dwarikanath

AU - Schüffler, Peter J.

AU - Tielbeek, Jeroen A.W.

AU - Puylaert, Carl A.J.

AU - Makanyanga, Jesica C.

AU - Menys, Alex

AU - Andriantsimiavona, Rado

AU - Stoker, Jaap

AU - Taylor, Stuart A.

AU - Vos, Franciscus M.

AU - Buhmann, Joachim M.

N1 - Publisher Copyright: © Springer International Publishing Switzerland 2014.

PY - 2014

Y1 - 2014

N2 - We propose a graph cut (GC) based approach for combining annotations from multiple experts and segmenting Crohns disease (CD) tissues in magnetic resonance (MR) images. Random forest (RF) based semi supervised learning (SSL) predicts missing expert labels while a novel self consistency (SC) score quantifies the reliability of each expert label and also serves as the penalty cost in a second order Markov random field (MRF) cost function. The final consensus label is obtained by GC optimization. Experimental results on synthetic images and real CD patient data show our final segmentation to be more accurate than those obtained by competing methods. It also highlights the effectiveness of SC score in quantifying expert reliability and accuracy of SSL in predicting missing labels.

AB - We propose a graph cut (GC) based approach for combining annotations from multiple experts and segmenting Crohns disease (CD) tissues in magnetic resonance (MR) images. Random forest (RF) based semi supervised learning (SSL) predicts missing expert labels while a novel self consistency (SC) score quantifies the reliability of each expert label and also serves as the penalty cost in a second order Markov random field (MRF) cost function. The final consensus label is obtained by GC optimization. Experimental results on synthetic images and real CD patient data show our final segmentation to be more accurate than those obtained by competing methods. It also highlights the effectiveness of SC score in quantifying expert reliability and accuracy of SSL in predicting missing labels.

KW - Crohn’s disease

KW - Graph cut

KW - Magnetic resonance

KW - Markov random field

KW - Random forest

UR - http://www.scopus.com/inward/record.url?scp=84927669781&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-13692-9_13

DO - 10.1007/978-3-319-13692-9_13

M3 - Conference contribution

AN - SCOPUS:84927669781

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

SP - 139

EP - 147

BT - Abdominal Imaging

A2 - Yoshida, Hiroyuki

A2 - Näppi, Janne J.

A2 - Saini, Sanjay

PB - Springer Verlag

T2 - 6th International Workshop on Abdominal Imaging: Computational and Clinical Applications, ABDI 2014 held in conjunction with 17th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2014

Y2 - 14 September 2014 through 14 September 2014

ER -

Mahapatra D, Schüffler PJ, Tielbeek JAW, Puylaert CAJ, Makanyanga JC, Menys A et al. Combining multiple expert annotations using semi-supervised learning and graph cuts for Crohn’s disease segmentation. in Yoshida H, Näppi JJ, Saini S, Hrsg., Abdominal Imaging: Computational and Clinical Applications - 6th International Workshop, ABDI 2014 held in conjunction with MICCAI 2014. Springer Verlag. 2014. S. 139-147. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-13692-9_13

Combining multiple expert annotations using semi-supervised learning and graph cuts for Crohn’s disease segmentation

Abstract

Publikationsreihe

Konferenz

Zugriff auf Dokument

Andere Dateien und Links

Fingerprint

Dieses zitieren