Confidence-based training for clinical data uncertainty in image-based prediction of cardiac ablation targets

Rocío Cabrera-Lozoya; Jan Margeta; Loïc Le Folgoc; Yuki Komatsu; Benjamin Berte; Jatin Relan; Hubert Cochet; Michel Haïssaguerre; Pierre Jaïs; Nicholas Ayache; Maxime Sermesant

doi:10.1007/978-3-319-13972-2_14

Confidence-based training for clinical data uncertainty in image-based prediction of cardiac ablation targets

Rocío Cabrera-Lozoya, Jan Margeta, Loïc Le Folgoc, Yuki Komatsu, Benjamin Berte, Jatin Relan, Hubert Cochet, Michel Haïssaguerre, Pierre Jaïs, Nicholas Ayache, Maxime Sermesant

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

Ventricular radio-frequency ablation (RFA) can have a critical impact on preventing sudden cardiac arrest but is challenging due to a highly complex arrhythmogenic substrate. This work aims to identify local image characteristics capable of predicting the presence of local abnormal ventricular activities (LAVA). This can allow, pre-operatively and non-invasively, to improve and accelerate the procedure. To achieve this, intensity and texture-based local image features are computed and random forests are used for classification. However using machinelearning approaches on such complex multimodal data can prove difficult due to the inherent errors in the training set. In this manuscript we present a detailed analysis of these error sources due in particular to catheter motion and the data fusion process. We derived a principled analysis of confidence impact on classification. Moreover, we demonstrate how formal integration of these uncertainties in the training process improves the algorithm’s performance, opening up possibilities for noninvasive image-based prediction of RFA targets.

Original language	English
Title of host publication	Medical Computer Vision
Subtitle of host publication	Algorithms for Big Data - International Workshop, MCV 2014 held in Conjunction with MICCAI 2014, Revised Selected Papers
Editors	Henning Müller, Bjoern Menze, Shaoting Zhang, Weidong (Tom) Cai, Bjoern Menze, Georg Langs, Dimitris Metaxas, Georg Langs, Henning Müller, Michael Kelm, Albert Montillo, Weidong (Tom) Cai
Publisher	Springer Verlag
Pages	148-159
Number of pages	12
ISBN (Electronic)	9783319139715
DOIs	https://doi.org/10.1007/978-3-319-13972-2_14
Publication status	Published - 1 Jan 2014
Externally published	Yes
Event	International Workshop on Medical Computer Vision: Algorithms for Big Data was held in conjunction with 17th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI-bigMCV 2014 - Cambridge, United States Duration: 18 Sept 2014 → 18 Sept 2014

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	8848
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	International Workshop on Medical Computer Vision: Algorithms for Big Data was held in conjunction with 17th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI-bigMCV 2014
Country/Territory	United States
City	Cambridge
Period	18/09/14 → 18/09/14

Access to Document

10.1007/978-3-319-13972-2_14

Cite this

Cabrera-Lozoya, R., Margeta, J., Le Folgoc, L., Komatsu, Y., Berte, B., Relan, J., Cochet, H., Haïssaguerre, M., Jaïs, P., Ayache, N., & Sermesant, M. (2014). Confidence-based training for clinical data uncertainty in image-based prediction of cardiac ablation targets. In H. Müller, B. Menze, S. Zhang, W. Cai, B. Menze, G. Langs, D. Metaxas, G. Langs, H. Müller, M. Kelm, A. Montillo, & W. Cai (Eds.), Medical Computer Vision: Algorithms for Big Data - International Workshop, MCV 2014 held in Conjunction with MICCAI 2014, Revised Selected Papers (pp. 148-159). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8848). Springer Verlag. https://doi.org/10.1007/978-3-319-13972-2_14

Cabrera-Lozoya, Rocío ; Margeta, Jan ; Le Folgoc, Loïc et al. / Confidence-based training for clinical data uncertainty in image-based prediction of cardiac ablation targets. Medical Computer Vision: Algorithms for Big Data - International Workshop, MCV 2014 held in Conjunction with MICCAI 2014, Revised Selected Papers. editor / Henning Müller ; Bjoern Menze ; Shaoting Zhang ; Weidong (Tom) Cai ; Bjoern Menze ; Georg Langs ; Dimitris Metaxas ; Georg Langs ; Henning Müller ; Michael Kelm ; Albert Montillo ; Weidong (Tom) Cai. Springer Verlag, 2014. pp. 148-159 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{cb6fd5d42fb743738407c6899d5a4e0c,

title = "Confidence-based training for clinical data uncertainty in image-based prediction of cardiac ablation targets",

abstract = "Ventricular radio-frequency ablation (RFA) can have a critical impact on preventing sudden cardiac arrest but is challenging due to a highly complex arrhythmogenic substrate. This work aims to identify local image characteristics capable of predicting the presence of local abnormal ventricular activities (LAVA). This can allow, pre-operatively and non-invasively, to improve and accelerate the procedure. To achieve this, intensity and texture-based local image features are computed and random forests are used for classification. However using machinelearning approaches on such complex multimodal data can prove difficult due to the inherent errors in the training set. In this manuscript we present a detailed analysis of these error sources due in particular to catheter motion and the data fusion process. We derived a principled analysis of confidence impact on classification. Moreover, we demonstrate how formal integration of these uncertainties in the training process improves the algorithm{\textquoteright}s performance, opening up possibilities for noninvasive image-based prediction of RFA targets.",

author = "Roc{\'i}o Cabrera-Lozoya and Jan Margeta and \{Le Folgoc\}, Lo{\"i}c and Yuki Komatsu and Benjamin Berte and Jatin Relan and Hubert Cochet and Michel Ha{\"i}ssaguerre and Pierre Ja{\"i}s and Nicholas Ayache and Maxime Sermesant",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing Switzerland 2014.; International Workshop on Medical Computer Vision: Algorithms for Big Data was held in conjunction with 17th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI-bigMCV 2014 ; Conference date: 18-09-2014 Through 18-09-2014",

year = "2014",

month = jan,

day = "1",

doi = "10.1007/978-3-319-13972-2\_14",

language = "English",

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

publisher = "Springer Verlag",

pages = "148--159",

editor = "Henning M{\"u}ller and Bjoern Menze and Shaoting Zhang and Cai, \{Weidong (Tom)\} and Bjoern Menze and Georg Langs and Dimitris Metaxas and Georg Langs and Henning M{\"u}ller and Michael Kelm and Albert Montillo and Cai, \{Weidong (Tom)\}",

booktitle = "Medical Computer Vision",

}

Cabrera-Lozoya, R, Margeta, J, Le Folgoc, L, Komatsu, Y, Berte, B, Relan, J, Cochet, H, Haïssaguerre, M, Jaïs, P, Ayache, N & Sermesant, M 2014, Confidence-based training for clinical data uncertainty in image-based prediction of cardiac ablation targets. in H Müller, B Menze, S Zhang, W Cai, B Menze, G Langs, D Metaxas, G Langs, H Müller, M Kelm, A Montillo & W Cai (eds), Medical Computer Vision: Algorithms for Big Data - International Workshop, MCV 2014 held in Conjunction with MICCAI 2014, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 8848, Springer Verlag, pp. 148-159, International Workshop on Medical Computer Vision: Algorithms for Big Data was held in conjunction with 17th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI-bigMCV 2014, Cambridge, United States, 18/09/14. https://doi.org/10.1007/978-3-319-13972-2_14

Confidence-based training for clinical data uncertainty in image-based prediction of cardiac ablation targets. / Cabrera-Lozoya, Rocío; Margeta, Jan; Le Folgoc, Loïc et al.
Medical Computer Vision: Algorithms for Big Data - International Workshop, MCV 2014 held in Conjunction with MICCAI 2014, Revised Selected Papers. ed. / Henning Müller; Bjoern Menze; Shaoting Zhang; Weidong (Tom) Cai; Bjoern Menze; Georg Langs; Dimitris Metaxas; Georg Langs; Henning Müller; Michael Kelm; Albert Montillo; Weidong (Tom) Cai. Springer Verlag, 2014. p. 148-159 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8848).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Confidence-based training for clinical data uncertainty in image-based prediction of cardiac ablation targets

AU - Cabrera-Lozoya, Rocío

AU - Margeta, Jan

AU - Le Folgoc, Loïc

AU - Komatsu, Yuki

AU - Berte, Benjamin

AU - Relan, Jatin

AU - Cochet, Hubert

AU - Haïssaguerre, Michel

AU - Jaïs, Pierre

AU - Ayache, Nicholas

AU - Sermesant, Maxime

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

PY - 2014/1/1

Y1 - 2014/1/1

N2 - Ventricular radio-frequency ablation (RFA) can have a critical impact on preventing sudden cardiac arrest but is challenging due to a highly complex arrhythmogenic substrate. This work aims to identify local image characteristics capable of predicting the presence of local abnormal ventricular activities (LAVA). This can allow, pre-operatively and non-invasively, to improve and accelerate the procedure. To achieve this, intensity and texture-based local image features are computed and random forests are used for classification. However using machinelearning approaches on such complex multimodal data can prove difficult due to the inherent errors in the training set. In this manuscript we present a detailed analysis of these error sources due in particular to catheter motion and the data fusion process. We derived a principled analysis of confidence impact on classification. Moreover, we demonstrate how formal integration of these uncertainties in the training process improves the algorithm’s performance, opening up possibilities for noninvasive image-based prediction of RFA targets.

AB - Ventricular radio-frequency ablation (RFA) can have a critical impact on preventing sudden cardiac arrest but is challenging due to a highly complex arrhythmogenic substrate. This work aims to identify local image characteristics capable of predicting the presence of local abnormal ventricular activities (LAVA). This can allow, pre-operatively and non-invasively, to improve and accelerate the procedure. To achieve this, intensity and texture-based local image features are computed and random forests are used for classification. However using machinelearning approaches on such complex multimodal data can prove difficult due to the inherent errors in the training set. In this manuscript we present a detailed analysis of these error sources due in particular to catheter motion and the data fusion process. We derived a principled analysis of confidence impact on classification. Moreover, we demonstrate how formal integration of these uncertainties in the training process improves the algorithm’s performance, opening up possibilities for noninvasive image-based prediction of RFA targets.

U2 - 10.1007/978-3-319-13972-2_14

DO - 10.1007/978-3-319-13972-2_14

M3 - Conference contribution

AN - SCOPUS:84917706467

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

SP - 148

EP - 159

BT - Medical Computer Vision

A2 - Müller, Henning

A2 - Menze, Bjoern

A2 - Zhang, Shaoting

A2 - Cai, Weidong (Tom)

A2 - Menze, Bjoern

A2 - Langs, Georg

A2 - Metaxas, Dimitris

A2 - Langs, Georg

A2 - Müller, Henning

A2 - Kelm, Michael

A2 - Montillo, Albert

A2 - Cai, Weidong (Tom)

PB - Springer Verlag

T2 - International Workshop on Medical Computer Vision: Algorithms for Big Data was held in conjunction with 17th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI-bigMCV 2014

Y2 - 18 September 2014 through 18 September 2014

ER -

Cabrera-Lozoya R, Margeta J, Le Folgoc L, Komatsu Y, Berte B, Relan J et al. Confidence-based training for clinical data uncertainty in image-based prediction of cardiac ablation targets. In Müller H, Menze B, Zhang S, Cai W, Menze B, Langs G, Metaxas D, Langs G, Müller H, Kelm M, Montillo A, Cai W, editors, Medical Computer Vision: Algorithms for Big Data - International Workshop, MCV 2014 held in Conjunction with MICCAI 2014, Revised Selected Papers. Springer Verlag. 2014. p. 148-159. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-13972-2_14

Confidence-based training for clinical data uncertainty in image-based prediction of cardiac ablation targets

Abstract

Publication series

Conference

Access to Document

Fingerprint

Cite this