Catheter detection in 3D ultrasound by CNN

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

Research output: Contribution to conferenceOtherAcademic

Abstract

In this paper, we propose a catheter detection method based on convolutional neural networks (CNNs) in 3D US. Voxels in US images are classified as catheter (or not) using triplanarbased CNNs. Our proposed CNN employs two-stage training with weighted loss function, which can cope with highly imbalanced training data and improves classification accuracy. Based on classified volumes, the catheters are localized with an average position error of smaller than 3 voxels in the examined datasets, indicating that catheters are always detected
in noisy and low-resolution images.

Conference

ConferenceThe Netherlands Conference on Computer Vision 2018
Abbreviated titleNCCV18
Period26/09/1827/09/18

Fingerprint

Catheters
Ultrasonics
Neural networks
Image resolution

Cite this

Yang, H., Shan, C., Kolen, A. F., & de With, P. (2018). Catheter detection in 3D ultrasound by CNN. The Netherlands Conference on Computer Vision 2018, .
Yang, Hongxu ; Shan, Caifeng ; Kolen, Alexander F. ; de With, Peter. / Catheter detection in 3D ultrasound by CNN. The Netherlands Conference on Computer Vision 2018, .2 p.
@conference{239736ceb4d84905b3663ae1b8e9db86,
title = "Catheter detection in 3D ultrasound by CNN",
abstract = "In this paper, we propose a catheter detection method based on convolutional neural networks (CNNs) in 3D US. Voxels in US images are classified as catheter (or not) using triplanarbased CNNs. Our proposed CNN employs two-stage training with weighted loss function, which can cope with highly imbalanced training data and improves classification accuracy. Based on classified volumes, the catheters are localized with an average position error of smaller than 3 voxels in the examined datasets, indicating that catheters are always detectedin noisy and low-resolution images.",
author = "Hongxu Yang and Caifeng Shan and Kolen, {Alexander F.} and {de With}, Peter",
year = "2018",
language = "English",
note = "The Netherlands Conference on Computer Vision 2018, NCCV18 ; Conference date: 26-09-2018 Through 27-09-2018",

}

Yang, H, Shan, C, Kolen, AF & de With, P 2018, 'Catheter detection in 3D ultrasound by CNN' The Netherlands Conference on Computer Vision 2018, 26/09/18 - 27/09/18, .

Catheter detection in 3D ultrasound by CNN. / Yang, Hongxu; Shan, Caifeng; Kolen, Alexander F.; de With, Peter.

2018. The Netherlands Conference on Computer Vision 2018, .

Research output: Contribution to conferenceOtherAcademic

TY - CONF

T1 - Catheter detection in 3D ultrasound by CNN

AU - Yang,Hongxu

AU - Shan,Caifeng

AU - Kolen,Alexander F.

AU - de With,Peter

PY - 2018

Y1 - 2018

N2 - In this paper, we propose a catheter detection method based on convolutional neural networks (CNNs) in 3D US. Voxels in US images are classified as catheter (or not) using triplanarbased CNNs. Our proposed CNN employs two-stage training with weighted loss function, which can cope with highly imbalanced training data and improves classification accuracy. Based on classified volumes, the catheters are localized with an average position error of smaller than 3 voxels in the examined datasets, indicating that catheters are always detectedin noisy and low-resolution images.

AB - In this paper, we propose a catheter detection method based on convolutional neural networks (CNNs) in 3D US. Voxels in US images are classified as catheter (or not) using triplanarbased CNNs. Our proposed CNN employs two-stage training with weighted loss function, which can cope with highly imbalanced training data and improves classification accuracy. Based on classified volumes, the catheters are localized with an average position error of smaller than 3 voxels in the examined datasets, indicating that catheters are always detectedin noisy and low-resolution images.

M3 - Other

ER -

Yang H, Shan C, Kolen AF, de With P. Catheter detection in 3D ultrasound by CNN. 2018. The Netherlands Conference on Computer Vision 2018, .