Fast catheter tracking in echocardiographic sequences for cardiac catheterization interventions

Xianliang Wu; R. James Housden; Niharika Varma; Yingliang Ma; Kawal S. Rhode; Daniel Rueckert

doi:10.1007/978-3-642-54268-8_20

Fast catheter tracking in echocardiographic sequences for cardiac catheterization interventions

Xianliang Wu, R. James Housden, Niharika Varma, Yingliang Ma, Kawal S. Rhode, Daniel Rueckert

Imperial College London
Fédération d'anesthésiologie-réanimation et de médecine hyperbare, hôpital d'instruction des armées du Val-de-Grâce, 75005 Paris, France; Inserm U676, hôpital Robert-Debré, 48, boulevard Sérurier, 75019 Paris, France; Université Paris Diderot, Sorbonne Paris Cité, UMRS 676, 75013 Paris, France; Centre for the Developing Brain, Department of Perinatal Imaging and Health, King's College London, London, Royaume-Uni. Electronic address: [email protected].

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

1 Citation (Scopus)

Abstract

For most cardiac catheterization interventions, X-ray imaging is currently used as a standard imaging technique. However, lack of 3D soft tissue information and harmful radiation mean that X-ray imaging is not an ideal modality. In contrast, 3D echocardiography can overcome these disadvantages. In this paper, we propose a fast catheter tracking strategy for 3D ultrasound sequences. The main advantage of our strategy is low use of X-ray imaging, which significantly decreases the radiation exposure. In addition, 3D soft tissue imaging can be introduced by using ultrasound. To enable the tracking procedure, initialization is carried out on the first ultrasound frame. Given the location of the catheter in the previous frame, which is in the form of a set of ordered landmarks, 3D Speeded-Up Robust Feature (SURF) responses are calculated for candidate voxels in the surrounding region of each landmark on the next frame. One candidate is selected among all voxels for each landmark based on Fast Primal-Dual optimization (Fast-PD). As a result, a new set of ordered landmarks is extracted, corresponding to the potential location of the catheter on the next frame. In order to adapt the tracking to the changing length of the catheter in the view, landmarks which may not be located on the catheter are ruled out. Then a catheter growing strategy is performed to extend the tracked part of the catheter to the untracked part. Based on 10 ultrasound phantom sequences and two clinical sequences, comprising more than 1300 frames, our experimental results show that the tracking system can track catheters with an error of less than 2.5mm and a speed of more than 3 fps.

Original language	English
Title of host publication	Statistical Atlases and Computational Models of the Heart: Imaging and Modelling Challenges - 4th International Workshop, STACOM 2013, Held in Conjunction with MICCAI 2013, Revised Selected Papers
Publisher	Springer Verlag
Pages	171-179
Number of pages	9
Volume	8330 LNCS
ISBN (Print)	9783642542671
DOIs	https://doi.org/10.1007/978-3-642-54268-8_20
Publication status	Published - 2014
Event	4th International Workshop on Statistical Atlases and Computational Models of the Heart: Imaging and Modelling Challenges, STACOM 2013, Held in Conjunction with MICCAI 2013 - Nagoya, Japan Duration: 26 Sept 2013 → 26 Sept 2013

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	8330 LNCS
ISSN (Print)	03029743
ISSN (Electronic)	16113349

Conference

Conference	4th International Workshop on Statistical Atlases and Computational Models of the Heart: Imaging and Modelling Challenges, STACOM 2013, Held in Conjunction with MICCAI 2013
Country/Territory	Japan
City	Nagoya
Period	26/09/2013 → 26/09/2013

Keywords

Cardiac catheterization
Catheter tracking
Ultrasound

Access to Document

10.1007/978-3-642-54268-8_20

Cite this

Wu, X., Housden, R. J., Varma, N., Ma, Y., Rhode, K. S., & Rueckert, D. (2014). Fast catheter tracking in echocardiographic sequences for cardiac catheterization interventions. In Statistical Atlases and Computational Models of the Heart: Imaging and Modelling Challenges - 4th International Workshop, STACOM 2013, Held in Conjunction with MICCAI 2013, Revised Selected Papers (Vol. 8330 LNCS, pp. 171-179). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8330 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-642-54268-8_20

Wu, Xianliang ; Housden, R. James ; Varma, Niharika et al. / Fast catheter tracking in echocardiographic sequences for cardiac catheterization interventions. Statistical Atlases and Computational Models of the Heart: Imaging and Modelling Challenges - 4th International Workshop, STACOM 2013, Held in Conjunction with MICCAI 2013, Revised Selected Papers. Vol. 8330 LNCS Springer Verlag, 2014. pp. 171-179 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inbook{245e0e22352d48bc93c9aaa01c9170e4,

title = "Fast catheter tracking in echocardiographic sequences for cardiac catheterization interventions",

abstract = "For most cardiac catheterization interventions, X-ray imaging is currently used as a standard imaging technique. However, lack of 3D soft tissue information and harmful radiation mean that X-ray imaging is not an ideal modality. In contrast, 3D echocardiography can overcome these disadvantages. In this paper, we propose a fast catheter tracking strategy for 3D ultrasound sequences. The main advantage of our strategy is low use of X-ray imaging, which significantly decreases the radiation exposure. In addition, 3D soft tissue imaging can be introduced by using ultrasound. To enable the tracking procedure, initialization is carried out on the first ultrasound frame. Given the location of the catheter in the previous frame, which is in the form of a set of ordered landmarks, 3D Speeded-Up Robust Feature (SURF) responses are calculated for candidate voxels in the surrounding region of each landmark on the next frame. One candidate is selected among all voxels for each landmark based on Fast Primal-Dual optimization (Fast-PD). As a result, a new set of ordered landmarks is extracted, corresponding to the potential location of the catheter on the next frame. In order to adapt the tracking to the changing length of the catheter in the view, landmarks which may not be located on the catheter are ruled out. Then a catheter growing strategy is performed to extend the tracked part of the catheter to the untracked part. Based on 10 ultrasound phantom sequences and two clinical sequences, comprising more than 1300 frames, our experimental results show that the tracking system can track catheters with an error of less than 2.5mm and a speed of more than 3 fps.",

keywords = "Cardiac catheterization, Catheter tracking, Ultrasound",

author = "Xianliang Wu and Housden, {R. James} and Niharika Varma and Yingliang Ma and Rhode, {Kawal S.} and Daniel Rueckert",

year = "2014",

doi = "10.1007/978-3-642-54268-8_20",

language = "English",

isbn = "9783642542671",

volume = "8330 LNCS",

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

publisher = "Springer Verlag",

pages = "171--179",

booktitle = "Statistical Atlases and Computational Models of the Heart: Imaging and Modelling Challenges - 4th International Workshop, STACOM 2013, Held in Conjunction with MICCAI 2013, Revised Selected Papers",

address = "Germany",

note = "4th International Workshop on Statistical Atlases and Computational Models of the Heart: Imaging and Modelling Challenges, STACOM 2013, Held in Conjunction with MICCAI 2013 ; Conference date: 26-09-2013 Through 26-09-2013",

}

Wu, X, Housden, RJ, Varma, N, Ma, Y, Rhode, KS & Rueckert, D 2014, Fast catheter tracking in echocardiographic sequences for cardiac catheterization interventions. in Statistical Atlases and Computational Models of the Heart: Imaging and Modelling Challenges - 4th International Workshop, STACOM 2013, Held in Conjunction with MICCAI 2013, Revised Selected Papers. vol. 8330 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 8330 LNCS, Springer Verlag, pp. 171-179, 4th International Workshop on Statistical Atlases and Computational Models of the Heart: Imaging and Modelling Challenges, STACOM 2013, Held in Conjunction with MICCAI 2013, Nagoya, Japan, 26/09/2013. https://doi.org/10.1007/978-3-642-54268-8_20

Fast catheter tracking in echocardiographic sequences for cardiac catheterization interventions. / Wu, Xianliang; Housden, R. James; Varma, Niharika et al.
Statistical Atlases and Computational Models of the Heart: Imaging and Modelling Challenges - 4th International Workshop, STACOM 2013, Held in Conjunction with MICCAI 2013, Revised Selected Papers. Vol. 8330 LNCS Springer Verlag, 2014. p. 171-179 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8330 LNCS).

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

TY - CHAP

T1 - Fast catheter tracking in echocardiographic sequences for cardiac catheterization interventions

AU - Wu, Xianliang

AU - Housden, R. James

AU - Varma, Niharika

AU - Ma, Yingliang

AU - Rhode, Kawal S.

AU - Rueckert, Daniel

PY - 2014

Y1 - 2014

N2 - For most cardiac catheterization interventions, X-ray imaging is currently used as a standard imaging technique. However, lack of 3D soft tissue information and harmful radiation mean that X-ray imaging is not an ideal modality. In contrast, 3D echocardiography can overcome these disadvantages. In this paper, we propose a fast catheter tracking strategy for 3D ultrasound sequences. The main advantage of our strategy is low use of X-ray imaging, which significantly decreases the radiation exposure. In addition, 3D soft tissue imaging can be introduced by using ultrasound. To enable the tracking procedure, initialization is carried out on the first ultrasound frame. Given the location of the catheter in the previous frame, which is in the form of a set of ordered landmarks, 3D Speeded-Up Robust Feature (SURF) responses are calculated for candidate voxels in the surrounding region of each landmark on the next frame. One candidate is selected among all voxels for each landmark based on Fast Primal-Dual optimization (Fast-PD). As a result, a new set of ordered landmarks is extracted, corresponding to the potential location of the catheter on the next frame. In order to adapt the tracking to the changing length of the catheter in the view, landmarks which may not be located on the catheter are ruled out. Then a catheter growing strategy is performed to extend the tracked part of the catheter to the untracked part. Based on 10 ultrasound phantom sequences and two clinical sequences, comprising more than 1300 frames, our experimental results show that the tracking system can track catheters with an error of less than 2.5mm and a speed of more than 3 fps.

AB - For most cardiac catheterization interventions, X-ray imaging is currently used as a standard imaging technique. However, lack of 3D soft tissue information and harmful radiation mean that X-ray imaging is not an ideal modality. In contrast, 3D echocardiography can overcome these disadvantages. In this paper, we propose a fast catheter tracking strategy for 3D ultrasound sequences. The main advantage of our strategy is low use of X-ray imaging, which significantly decreases the radiation exposure. In addition, 3D soft tissue imaging can be introduced by using ultrasound. To enable the tracking procedure, initialization is carried out on the first ultrasound frame. Given the location of the catheter in the previous frame, which is in the form of a set of ordered landmarks, 3D Speeded-Up Robust Feature (SURF) responses are calculated for candidate voxels in the surrounding region of each landmark on the next frame. One candidate is selected among all voxels for each landmark based on Fast Primal-Dual optimization (Fast-PD). As a result, a new set of ordered landmarks is extracted, corresponding to the potential location of the catheter on the next frame. In order to adapt the tracking to the changing length of the catheter in the view, landmarks which may not be located on the catheter are ruled out. Then a catheter growing strategy is performed to extend the tracked part of the catheter to the untracked part. Based on 10 ultrasound phantom sequences and two clinical sequences, comprising more than 1300 frames, our experimental results show that the tracking system can track catheters with an error of less than 2.5mm and a speed of more than 3 fps.

KW - Cardiac catheterization

KW - Catheter tracking

KW - Ultrasound

UR - http://www.scopus.com/inward/record.url?scp=84898900306&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-54268-8_20

DO - 10.1007/978-3-642-54268-8_20

M3 - Other chapter contribution

AN - SCOPUS:84898900306

SN - 9783642542671

VL - 8330 LNCS

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

SP - 171

EP - 179

BT - Statistical Atlases and Computational Models of the Heart: Imaging and Modelling Challenges - 4th International Workshop, STACOM 2013, Held in Conjunction with MICCAI 2013, Revised Selected Papers

PB - Springer Verlag

T2 - 4th International Workshop on Statistical Atlases and Computational Models of the Heart: Imaging and Modelling Challenges, STACOM 2013, Held in Conjunction with MICCAI 2013

Y2 - 26 September 2013 through 26 September 2013

ER -

Wu X, Housden RJ, Varma N, Ma Y, Rhode KS, Rueckert D. Fast catheter tracking in echocardiographic sequences for cardiac catheterization interventions. In Statistical Atlases and Computational Models of the Heart: Imaging and Modelling Challenges - 4th International Workshop, STACOM 2013, Held in Conjunction with MICCAI 2013, Revised Selected Papers. Vol. 8330 LNCS. Springer Verlag. 2014. p. 171-179. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-642-54268-8_20

Fast catheter tracking in echocardiographic sequences for cardiac catheterization interventions

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this