Hepatic vessel segmentation using a reduced filter 3D U-Net in ultrasound imaging

Bart R. Thomson; Jasper Nijkamp; Oleksandra Ivashchenko; Ferdinand van der Heijden; Jasper N. Smit; Niels F.M. Kok; Koert F.D. Kuhlmann; Theo J.M. Ruers; Matteo Fusaglia

Hepatic vessel segmentation using a reduced filter 3D U-Net in ultrasound imaging

Bart R. Thomson, Jasper Nijkamp, Oleksandra Ivashchenko, Ferdinand van der Heijden, Jasper N. Smit, Niels F.M. Kok, Koert F.D. Kuhlmann, Theo J.M. Ruers, Matteo Fusaglia

Research output: Contribution to conference › Abstract › Academic

125 Downloads (Pure)

Abstract

Accurate hepatic vessel segmentation on ultrasound (US) images can be an important tool in the planning and execution of surgery, however proves to be a challenging task due to noise and speckle. Our method comprises a reduced filter 3D U-Net implementation to automatically detect hepatic vasculature in 3D US volumes. A comparison is made between volumes acquired with a 3D probe and stacked 2D US images based on electromagnetic tracking. Experiments are conducted on 67 scans, where 45 are used in training, 12 in validation and 10 in testing. This network architecture yields Dice scores of 0.740 and 0.781 for 3D and stacked 2D volumes respectively, comparing promising to literature and inter-observer performance (Dice = 0.879).

Original language	English
Number of pages	4
Publication status	Published - 28 Jul 2019
Event	2nd International Conference on Medical Imaging with Deep Learning, MIDL 2019 - London, United Kingdom Duration: 8 Jul 2019 → 10 Jul 2019 Conference number: 2

Conference

Conference	2nd International Conference on Medical Imaging with Deep Learning, MIDL 2019
Abbreviated title	MIDL
Country/Territory	United Kingdom
City	London
Period	8/07/19 → 10/07/19

Keywords

Deep learning
Segmentation
Liver
Ultrasound
U-Net
eess.IV

Access to Document

Thomson2019hepaticFinal published version, 525 KB

https://arxiv.org/abs/1907.08612

Cite this

Thomson, B. R., Nijkamp, J., Ivashchenko, O., Heijden, F. V. D., Smit, J. N., Kok, N. F. M., Kuhlmann, K. F. D., Ruers, T. J. M., & Fusaglia, M. (2019). Hepatic vessel segmentation using a reduced filter 3D U-Net in ultrasound imaging. Abstract from 2nd International Conference on Medical Imaging with Deep Learning, MIDL 2019, London, United Kingdom. https://arxiv.org/abs/1907.08612

@conference{e505be79356e4a9d857e1a793e2dc064,

title = "Hepatic vessel segmentation using a reduced filter 3D U-Net in ultrasound imaging",

abstract = " Accurate hepatic vessel segmentation on ultrasound (US) images can be an important tool in the planning and execution of surgery, however proves to be a challenging task due to noise and speckle. Our method comprises a reduced filter 3D U-Net implementation to automatically detect hepatic vasculature in 3D US volumes. A comparison is made between volumes acquired with a 3D probe and stacked 2D US images based on electromagnetic tracking. Experiments are conducted on 67 scans, where 45 are used in training, 12 in validation and 10 in testing. This network architecture yields Dice scores of 0.740 and 0.781 for 3D and stacked 2D volumes respectively, comparing promising to literature and inter-observer performance (Dice = 0.879). ",

keywords = "Deep learning, Segmentation, Liver, Ultrasound, U-Net, eess.IV",

author = "Thomson, {Bart R.} and Jasper Nijkamp and Oleksandra Ivashchenko and Heijden, {Ferdinand van der} and Smit, {Jasper N.} and Kok, {Niels F.M.} and Kuhlmann, {Koert F.D.} and Ruers, {Theo J.M.} and Matteo Fusaglia",

note = "Extended Abstract; 2nd International Conference on Medical Imaging with Deep Learning, MIDL 2019, MIDL ; Conference date: 08-07-2019 Through 10-07-2019",

year = "2019",

month = jul,

day = "28",

language = "English",

}

TY - CONF

T1 - Hepatic vessel segmentation using a reduced filter 3D U-Net in ultrasound imaging

AU - Thomson, Bart R.

AU - Nijkamp, Jasper

AU - Ivashchenko, Oleksandra

AU - Heijden, Ferdinand van der

AU - Smit, Jasper N.

AU - Kok, Niels F.M.

AU - Kuhlmann, Koert F.D.

AU - Ruers, Theo J.M.

AU - Fusaglia, Matteo

N1 - Conference code: 2

PY - 2019/7/28

Y1 - 2019/7/28

N2 - Accurate hepatic vessel segmentation on ultrasound (US) images can be an important tool in the planning and execution of surgery, however proves to be a challenging task due to noise and speckle. Our method comprises a reduced filter 3D U-Net implementation to automatically detect hepatic vasculature in 3D US volumes. A comparison is made between volumes acquired with a 3D probe and stacked 2D US images based on electromagnetic tracking. Experiments are conducted on 67 scans, where 45 are used in training, 12 in validation and 10 in testing. This network architecture yields Dice scores of 0.740 and 0.781 for 3D and stacked 2D volumes respectively, comparing promising to literature and inter-observer performance (Dice = 0.879).

AB - Accurate hepatic vessel segmentation on ultrasound (US) images can be an important tool in the planning and execution of surgery, however proves to be a challenging task due to noise and speckle. Our method comprises a reduced filter 3D U-Net implementation to automatically detect hepatic vasculature in 3D US volumes. A comparison is made between volumes acquired with a 3D probe and stacked 2D US images based on electromagnetic tracking. Experiments are conducted on 67 scans, where 45 are used in training, 12 in validation and 10 in testing. This network architecture yields Dice scores of 0.740 and 0.781 for 3D and stacked 2D volumes respectively, comparing promising to literature and inter-observer performance (Dice = 0.879).

KW - Deep learning

KW - Segmentation

KW - Liver

KW - Ultrasound

KW - U-Net

KW - eess.IV

M3 - Abstract

T2 - 2nd International Conference on Medical Imaging with Deep Learning, MIDL 2019

Y2 - 8 July 2019 through 10 July 2019

ER -

Hepatic vessel segmentation using a reduced filter 3D U-Net in ultrasound imaging

Abstract

Conference

Keywords

Access to Document

Fingerprint

Cite this