Tools
Tools
Wu, Xianliang, Housden, R. James, Varma, Niharika, Ma, YingLiang 
ORCID: https://orcid.org/0000-0001-5770-5843, Rhode, Kawal S. and Rueckert, Daniel
(2014)
Fast catheter tracking in echocardiographic sequences for cardiac catheterization interventions.
In:
Statistical Atlases and Computational Models of the Heart. Imaging and Modelling Challenges.
Lecture Notes in Computer Science
.
Springer, 171–179.
ISBN 978-3-642-54267-1
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.
| Item Type: | Book Section |
|---|---|
| Faculty \ School: | Faculty of Science > School of Computing Sciences |
| UEA Research Groups: | Faculty of Science > Research Groups > Norwich Epidemiology Centre Faculty of Medicine and Health Sciences > Research Groups > Norwich Epidemiology Centre Faculty of Science > Research Groups > Data Science and AI |
| Related URLs: | |
| Depositing User: | LivePure Connector |
| Date Deposited: | 05 Jan 2023 11:30 |
| Last Modified: | 10 Dec 2024 01:13 |
| URI: | https://ueaeprints.uea.ac.uk/id/eprint/90407 |
| DOI: | 10.1007/978-3-642-54268-8_20 |
Actions (login required)
 |
View Item |