King's College London

Research portal

Automatic segmentation method of pelvic floor levator hiatus in ultrasound using a self-normalizing neural network

Research output: Contribution to journalArticle

Standard

Automatic segmentation method of pelvic floor levator hiatus in ultrasound using a self-normalizing neural network. / Bonmati, Ester; Hu, Yipeng; Sindhwani, Nikhil; Dietz, Hans Peter; D'hooge, Jan; Barratt, Dean; Deprest, Jan; Vercauteren, Tom.

In: Journal of Medical Imaging, Vol. 5, No. 2, 021206, 01.04.2018, p. 021206.

Research output: Contribution to journalArticle

Harvard

Bonmati, E, Hu, Y, Sindhwani, N, Dietz, HP, D'hooge, J, Barratt, D, Deprest, J & Vercauteren, T 2018, 'Automatic segmentation method of pelvic floor levator hiatus in ultrasound using a self-normalizing neural network', Journal of Medical Imaging, vol. 5, no. 2, 021206, pp. 021206. https://doi.org/10.1117/1.JMI.5.2.021206

APA

Bonmati, E., Hu, Y., Sindhwani, N., Dietz, H. P., D'hooge, J., Barratt, D., ... Vercauteren, T. (2018). Automatic segmentation method of pelvic floor levator hiatus in ultrasound using a self-normalizing neural network. Journal of Medical Imaging, 5(2), 021206. [021206]. https://doi.org/10.1117/1.JMI.5.2.021206

Vancouver

Bonmati E, Hu Y, Sindhwani N, Dietz HP, D'hooge J, Barratt D et al. Automatic segmentation method of pelvic floor levator hiatus in ultrasound using a self-normalizing neural network. Journal of Medical Imaging. 2018 Apr 1;5(2):021206. 021206. https://doi.org/10.1117/1.JMI.5.2.021206

Author

Bonmati, Ester ; Hu, Yipeng ; Sindhwani, Nikhil ; Dietz, Hans Peter ; D'hooge, Jan ; Barratt, Dean ; Deprest, Jan ; Vercauteren, Tom. / Automatic segmentation method of pelvic floor levator hiatus in ultrasound using a self-normalizing neural network. In: Journal of Medical Imaging. 2018 ; Vol. 5, No. 2. pp. 021206.

Bibtex Download

@article{19d3b93b5b1740798b1e49f932ab794f,
title = "Automatic segmentation method of pelvic floor levator hiatus in ultrasound using a self-normalizing neural network",
abstract = "Segmentation of the levator hiatus in ultrasound allows the extraction of biometrics, which are of importance for pelvic floor disorder assessment. We present a fully automatic method using a convolutional neural network (CNN) to outline the levator hiatus in a two-dimensional image extracted from a three-dimensional ultrasound volume. In particular, our method uses a recently developed scaled exponential linear unit (SELU) as a nonlinear self-normalizing activation function, which for the first time has been applied in medical imaging with CNN. SELU has important advantages such as being parameter-free and mini-batch independent, which may help to overcome memory constraints during training. A dataset with 91 images from 35 patients during Valsalva, contraction, and rest, all labeled by three operators, is used for training and evaluation in a leave-one-patient-out cross validation. Results show a median Dice similarity coefficient of 0.90 with an interquartile range of 0.08, with equivalent performance to the three operators (with a Williams' index of 1.03), and outperforming a U-Net architecture without the need for batch normalization. We conclude that the proposed fully automatic method achieved equivalent accuracy in segmenting the pelvic floor levator hiatus compared to a previous semiautomatic approach.",
keywords = "automatic segmentation, convolutional neural network, levator hiatus, self-normalizing neural network, ultrasound",
author = "Ester Bonmati and Yipeng Hu and Nikhil Sindhwani and Dietz, {Hans Peter} and Jan D'hooge and Dean Barratt and Jan Deprest and Tom Vercauteren",
year = "2018",
month = "4",
day = "1",
doi = "10.1117/1.JMI.5.2.021206",
language = "English",
volume = "5",
pages = "021206",
journal = "Journal of medical imaging (Bellingham, Wash.)",
issn = "2329-4302",
publisher = "SPIE",
number = "2",

}

RIS (suitable for import to EndNote) Download

TY - JOUR

T1 - Automatic segmentation method of pelvic floor levator hiatus in ultrasound using a self-normalizing neural network

AU - Bonmati, Ester

AU - Hu, Yipeng

AU - Sindhwani, Nikhil

AU - Dietz, Hans Peter

AU - D'hooge, Jan

AU - Barratt, Dean

AU - Deprest, Jan

AU - Vercauteren, Tom

PY - 2018/4/1

Y1 - 2018/4/1

N2 - Segmentation of the levator hiatus in ultrasound allows the extraction of biometrics, which are of importance for pelvic floor disorder assessment. We present a fully automatic method using a convolutional neural network (CNN) to outline the levator hiatus in a two-dimensional image extracted from a three-dimensional ultrasound volume. In particular, our method uses a recently developed scaled exponential linear unit (SELU) as a nonlinear self-normalizing activation function, which for the first time has been applied in medical imaging with CNN. SELU has important advantages such as being parameter-free and mini-batch independent, which may help to overcome memory constraints during training. A dataset with 91 images from 35 patients during Valsalva, contraction, and rest, all labeled by three operators, is used for training and evaluation in a leave-one-patient-out cross validation. Results show a median Dice similarity coefficient of 0.90 with an interquartile range of 0.08, with equivalent performance to the three operators (with a Williams' index of 1.03), and outperforming a U-Net architecture without the need for batch normalization. We conclude that the proposed fully automatic method achieved equivalent accuracy in segmenting the pelvic floor levator hiatus compared to a previous semiautomatic approach.

AB - Segmentation of the levator hiatus in ultrasound allows the extraction of biometrics, which are of importance for pelvic floor disorder assessment. We present a fully automatic method using a convolutional neural network (CNN) to outline the levator hiatus in a two-dimensional image extracted from a three-dimensional ultrasound volume. In particular, our method uses a recently developed scaled exponential linear unit (SELU) as a nonlinear self-normalizing activation function, which for the first time has been applied in medical imaging with CNN. SELU has important advantages such as being parameter-free and mini-batch independent, which may help to overcome memory constraints during training. A dataset with 91 images from 35 patients during Valsalva, contraction, and rest, all labeled by three operators, is used for training and evaluation in a leave-one-patient-out cross validation. Results show a median Dice similarity coefficient of 0.90 with an interquartile range of 0.08, with equivalent performance to the three operators (with a Williams' index of 1.03), and outperforming a U-Net architecture without the need for batch normalization. We conclude that the proposed fully automatic method achieved equivalent accuracy in segmenting the pelvic floor levator hiatus compared to a previous semiautomatic approach.

KW - automatic segmentation

KW - convolutional neural network

KW - levator hiatus

KW - self-normalizing neural network

KW - ultrasound

UR - http://www.scopus.com/inward/record.url?scp=85040461956&partnerID=8YFLogxK

U2 - 10.1117/1.JMI.5.2.021206

DO - 10.1117/1.JMI.5.2.021206

M3 - Article

C2 - 29340289

AN - SCOPUS:85040461956

VL - 5

SP - 021206

JO - Journal of medical imaging (Bellingham, Wash.)

JF - Journal of medical imaging (Bellingham, Wash.)

SN - 2329-4302

IS - 2

M1 - 021206

ER -

View graph of relations

© 2018 King's College London | Strand | London WC2R 2LS | England | United Kingdom | Tel +44 (0)20 7836 5454