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

Hongxu Yang (Corresponding author), Caifeng Shan, Alexander F. Kolen, Peter de With

Research output: Contribution to journalArticleAcademicpeer-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
Catheter localization 3D ultrasound Frangi pre-filtering Convolutional neural network
LanguageEnglish
Pages1069-1077
Number of pages9
JournalInternational Journal of Computer Assisted Radiology and Surgery
Volume14
Issue number6
Early online date9 Apr 2019
DOIs
StatePublished - Jun 2019

Fingerprint

Catheters
Ultrasonics
Cardiac Catheters
Entropy
Neural networks

Cite this

@article{811830286c2b43fc8a7adabc8f468387,
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.KeywordsCatheter localization 3D ultrasound Frangi pre-filtering Convolutional neural network",
author = "Hongxu Yang and Caifeng Shan and Kolen, {Alexander F.} and {de With}, Peter",
year = "2019",
month = "6",
doi = "10.1007/s11548-019-01960-y",
language = "English",
volume = "14",
pages = "1069--1077",
journal = "International Journal of Computer Assisted Radiology and Surgery",
issn = "1861-6410",
publisher = "Springer",
number = "6",

}

TY - JOUR

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

PY - 2019/6

Y1 - 2019/6

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.KeywordsCatheter localization 3D ultrasound Frangi pre-filtering Convolutional neural network

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.KeywordsCatheter localization 3D ultrasound Frangi pre-filtering Convolutional neural network

U2 - 10.1007/s11548-019-01960-y

DO - 10.1007/s11548-019-01960-y

M3 - Article

VL - 14

SP - 1069

EP - 1077

JO - International Journal of Computer Assisted Radiology and Surgery

T2 - International Journal of Computer Assisted Radiology and Surgery

JF - International Journal of Computer Assisted Radiology and Surgery

SN - 1861-6410

IS - 6

ER -