Catheter localization in 3D ultrasound using Voxel-of-interest-based ConvNets for cardiac intervention

Hongxu Yang, Caifeng Shan, Alexander F. Kolen, Peter de With

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

Abstract

Purpose
Efficient image-based catheter localization in 3D US during cardiac interventions is highly desired, since it facilitates the operation procedure, reduces the patient risk and improves the outcome. Current image-based catheter localization methods are not efficient or accurate enough for real clinical use.

Methods
We propose a catheter localization method for 3D cardiac ultrasound (US). The catheter candidate voxels are first pre-selected by the Frangi vesselness filter with adaptive thresholding, after which a triplanar-based ConvNet is applied to classify the remaining voxels as catheter or not. We propose a Share-ConvNet for 3D US, which reduces the computation complexity by sharing a single ConvNet for all orthogonal slices. To boost the performance of ConvNet, we also employ two-stage training with weighted cross-entropy. Using the classified voxels, the catheter is localized by a model fitting algorithm.

Results
To validate our method, we have collected challenging ex vivo datasets. Extensive experiments show that the proposed method outperforms state-of-the-art methods and can localize the catheter with an average error of 2.1 mm in around 10 s per volume.

Conclusion
Our method can automatically localize the cardiac catheter in challenging 3D cardiac US images. The efficiency and accuracy localization of the proposed method are considered promising for catheter detection and localization during clinical interventions.

Keywords
LanguageEnglish
Title of host publicationThe 10th International Conference on Information Processing in Computer-Assisted Interventions
StatePublished - 2019
EventThe 10th International Conference on Information Processing in Computer-Assisted Interventions, (IPCAI2019) - Rennes, France
Duration: 18 Jun 201919 Jun 2019
http://www.ipcai2019.org/

Conference

ConferenceThe 10th International Conference on Information Processing in Computer-Assisted Interventions, (IPCAI2019)
Abbreviated titleIPCAI2019
CountryFrance
CityRennes
Period18/06/1919/06/19
Internet address

Fingerprint

Catheters
Ultrasonics
Entropy

Bibliographical note

IPCAI 2019 Special Issue: Conference Information Processing for Computer-Assisted Interventions, 10th International Conference 2019 – Part 1

Cite this

Yang, H., Shan, C., Kolen, A. F., & de With, P. (2019). Catheter localization in 3D ultrasound using Voxel-of-interest-based ConvNets for cardiac intervention. In The 10th International Conference on Information Processing in Computer-Assisted Interventions
Yang, Hongxu ; Shan, Caifeng ; Kolen, Alexander F. ; de With, Peter. / Catheter localization in 3D ultrasound using Voxel-of-interest-based ConvNets for cardiac intervention. The 10th International Conference on Information Processing in Computer-Assisted Interventions. 2019.
@inproceedings{e3a55a15cd664374961ecdd9c4820715,
title = "Catheter localization in 3D ultrasound using Voxel-of-interest-based ConvNets for cardiac intervention",
abstract = "PurposeEfficient image-based catheter localization in 3D US during cardiac interventions is highly desired, since it facilitates the operation procedure, reduces the patient risk and improves the outcome. Current image-based catheter localization methods are not efficient or accurate enough for real clinical use.MethodsWe propose a catheter localization method for 3D cardiac ultrasound (US). The catheter candidate voxels are first pre-selected by the Frangi vesselness filter with adaptive thresholding, after which a triplanar-based ConvNet is applied to classify the remaining voxels as catheter or not. We propose a Share-ConvNet for 3D US, which reduces the computation complexity by sharing a single ConvNet for all orthogonal slices. To boost the performance of ConvNet, we also employ two-stage training with weighted cross-entropy. Using the classified voxels, the catheter is localized by a model fitting algorithm.ResultsTo validate our method, we have collected challenging ex vivo datasets. Extensive experiments show that the proposed method outperforms state-of-the-art methods and can localize the catheter with an average error of 2.1 mm in around 10 s per volume.ConclusionOur method can automatically localize the cardiac catheter in challenging 3D cardiac US images. The efficiency and accuracy localization of the proposed method are considered promising for catheter detection and localization during clinical interventions.Keywords",
author = "Hongxu Yang and Caifeng Shan and Kolen, {Alexander F.} and {de With}, Peter",
note = "IPCAI 2019 Special Issue: Conference Information Processing for Computer-Assisted Interventions, 10th International Conference 2019 – Part 1",
year = "2019",
language = "English",
booktitle = "The 10th International Conference on Information Processing in Computer-Assisted Interventions",

}

Yang, H, Shan, C, Kolen, AF & de With, P 2019, Catheter localization in 3D ultrasound using Voxel-of-interest-based ConvNets for cardiac intervention. in The 10th International Conference on Information Processing in Computer-Assisted Interventions. The 10th International Conference on Information Processing in Computer-Assisted Interventions, (IPCAI2019), Rennes, France, 18/06/19.

Catheter localization in 3D ultrasound using Voxel-of-interest-based ConvNets for cardiac intervention. / Yang, Hongxu; Shan, Caifeng; Kolen, Alexander F.; de With, Peter.

The 10th International Conference on Information Processing in Computer-Assisted Interventions. 2019.

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

TY - GEN

T1 - Catheter localization in 3D ultrasound using Voxel-of-interest-based ConvNets for cardiac intervention

AU - Yang,Hongxu

AU - Shan,Caifeng

AU - Kolen,Alexander F.

AU - de With,Peter

N1 - IPCAI 2019 Special Issue: Conference Information Processing for Computer-Assisted Interventions, 10th International Conference 2019 – Part 1

PY - 2019

Y1 - 2019

N2 - PurposeEfficient image-based catheter localization in 3D US during cardiac interventions is highly desired, since it facilitates the operation procedure, reduces the patient risk and improves the outcome. Current image-based catheter localization methods are not efficient or accurate enough for real clinical use.MethodsWe propose a catheter localization method for 3D cardiac ultrasound (US). The catheter candidate voxels are first pre-selected by the Frangi vesselness filter with adaptive thresholding, after which a triplanar-based ConvNet is applied to classify the remaining voxels as catheter or not. We propose a Share-ConvNet for 3D US, which reduces the computation complexity by sharing a single ConvNet for all orthogonal slices. To boost the performance of ConvNet, we also employ two-stage training with weighted cross-entropy. Using the classified voxels, the catheter is localized by a model fitting algorithm.ResultsTo validate our method, we have collected challenging ex vivo datasets. Extensive experiments show that the proposed method outperforms state-of-the-art methods and can localize the catheter with an average error of 2.1 mm in around 10 s per volume.ConclusionOur method can automatically localize the cardiac catheter in challenging 3D cardiac US images. The efficiency and accuracy localization of the proposed method are considered promising for catheter detection and localization during clinical interventions.Keywords

AB - PurposeEfficient image-based catheter localization in 3D US during cardiac interventions is highly desired, since it facilitates the operation procedure, reduces the patient risk and improves the outcome. Current image-based catheter localization methods are not efficient or accurate enough for real clinical use.MethodsWe propose a catheter localization method for 3D cardiac ultrasound (US). The catheter candidate voxels are first pre-selected by the Frangi vesselness filter with adaptive thresholding, after which a triplanar-based ConvNet is applied to classify the remaining voxels as catheter or not. We propose a Share-ConvNet for 3D US, which reduces the computation complexity by sharing a single ConvNet for all orthogonal slices. To boost the performance of ConvNet, we also employ two-stage training with weighted cross-entropy. Using the classified voxels, the catheter is localized by a model fitting algorithm.ResultsTo validate our method, we have collected challenging ex vivo datasets. Extensive experiments show that the proposed method outperforms state-of-the-art methods and can localize the catheter with an average error of 2.1 mm in around 10 s per volume.ConclusionOur method can automatically localize the cardiac catheter in challenging 3D cardiac US images. The efficiency and accuracy localization of the proposed method are considered promising for catheter detection and localization during clinical interventions.Keywords

M3 - Conference contribution

BT - The 10th International Conference on Information Processing in Computer-Assisted Interventions

ER -

Yang H, Shan C, Kolen AF, de With P. Catheter localization in 3D ultrasound using Voxel-of-interest-based ConvNets for cardiac intervention. In The 10th International Conference on Information Processing in Computer-Assisted Interventions. 2019.