Coherent needle detection in ultrasound volumes using 3D conditional random fields

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

2 Downloads (Pure)

Abstract

3D ultrasound (US) transducers will improve the quality of image-guided medical interventions if an automated detection of the needle becomes possible. Image-based detection of the needle is challenging due to the presence of other echogenic structures in the acquired data, inconsistent visibility of needle parts and the low quality in US imaging. As the currently applied approaches for needle detection classify each voxel individually, they do not consider the global relations between the voxels. In this work, we introduce coherent needle labeling by using dense conditional random fields over a volume, along with 3D space-frequency features. The proposal includes long-distance dependencies in voxel pairs according to their similarities in the feature space and their spatial distance. This post-processing stage leads to better label assignment of volume voxels and a more compact and coherent segmented region. Our ex-vivo experiments based on measuring the F-1, F-2 and IoU scores show that the performance improves a significant 10-20 % compared with only using the linear SVM as a baseline for voxel classification.

Original languageEnglish
Title of host publicationMedical Imaging 2018
Subtitle of host publicationImage-Guided Procedures, Robotic Interventions, and Modeling
EditorsB. Fei, R.J. Webster
Place of PublicationBellingham
PublisherSPIE
Pages1-8
ISBN (Electronic)9781510616417
DOIs
Publication statusPublished - 13 Mar 2018
EventMedical Imaging 2018: Image-Guided Procedures, Robotic Interventions, and Modeling - Houston, United States
Duration: 12 Feb 201815 Feb 2018

Publication series

NameProceedings of SPIE
Volume10576

Conference

ConferenceMedical Imaging 2018: Image-Guided Procedures, Robotic Interventions, and Modeling
CountryUnited States
CityHouston
Period12/02/1815/02/18

Fingerprint

needles
Needles
Ultrasonics
Transducers
visibility
Visibility
Labeling
marking
proposals
Labels
Ultrasonography
transducers
Imaging techniques
Processing
Experiments

Keywords

  • 3D ultrasound
  • Conditional Random Fields
  • Needle detection

Cite this

Ghazvinian Zanjani, F., Pourtaherian, A., Tang, X., Zinger, S., Mihajlovic, N., Ng, G. C., ... de With, P. H. N. (2018). Coherent needle detection in ultrasound volumes using 3D conditional random fields. In B. Fei, & R. J. Webster (Eds.), Medical Imaging 2018: Image-Guided Procedures, Robotic Interventions, and Modeling (pp. 1-8). [105760W] (Proceedings of SPIE; Vol. 10576). Bellingham: SPIE. https://doi.org/10.1117/12.2293079
Ghazvinian Zanjani, Farhad ; Pourtaherian, Arash ; Tang, Xikai ; Zinger, Svitlana ; Mihajlovic, Nenad ; Ng, Gary C. ; Korsten, Hendrikus H.M. ; de With, Peter H.N. / Coherent needle detection in ultrasound volumes using 3D conditional random fields. Medical Imaging 2018: Image-Guided Procedures, Robotic Interventions, and Modeling. editor / B. Fei ; R.J. Webster. Bellingham : SPIE, 2018. pp. 1-8 (Proceedings of SPIE).
@inproceedings{5093093041eb4417be7b94a30da7ded3,
title = "Coherent needle detection in ultrasound volumes using 3D conditional random fields",
abstract = "3D ultrasound (US) transducers will improve the quality of image-guided medical interventions if an automated detection of the needle becomes possible. Image-based detection of the needle is challenging due to the presence of other echogenic structures in the acquired data, inconsistent visibility of needle parts and the low quality in US imaging. As the currently applied approaches for needle detection classify each voxel individually, they do not consider the global relations between the voxels. In this work, we introduce coherent needle labeling by using dense conditional random fields over a volume, along with 3D space-frequency features. The proposal includes long-distance dependencies in voxel pairs according to their similarities in the feature space and their spatial distance. This post-processing stage leads to better label assignment of volume voxels and a more compact and coherent segmented region. Our ex-vivo experiments based on measuring the F-1, F-2 and IoU scores show that the performance improves a significant 10-20 {\%} compared with only using the linear SVM as a baseline for voxel classification.",
keywords = "3D ultrasound, Conditional Random Fields, Needle detection",
author = "{Ghazvinian Zanjani}, Farhad and Arash Pourtaherian and Xikai Tang and Svitlana Zinger and Nenad Mihajlovic and Ng, {Gary C.} and Korsten, {Hendrikus H.M.} and {de With}, {Peter H.N.}",
year = "2018",
month = "3",
day = "13",
doi = "10.1117/12.2293079",
language = "English",
series = "Proceedings of SPIE",
publisher = "SPIE",
pages = "1--8",
editor = "B. Fei and R.J. Webster",
booktitle = "Medical Imaging 2018",
address = "United States",

}

Ghazvinian Zanjani, F, Pourtaherian, A, Tang, X, Zinger, S, Mihajlovic, N, Ng, GC, Korsten, HHM & de With, PHN 2018, Coherent needle detection in ultrasound volumes using 3D conditional random fields. in B Fei & RJ Webster (eds), Medical Imaging 2018: Image-Guided Procedures, Robotic Interventions, and Modeling., 105760W, Proceedings of SPIE, vol. 10576, SPIE, Bellingham, pp. 1-8, Medical Imaging 2018: Image-Guided Procedures, Robotic Interventions, and Modeling, Houston, United States, 12/02/18. https://doi.org/10.1117/12.2293079

Coherent needle detection in ultrasound volumes using 3D conditional random fields. / Ghazvinian Zanjani, Farhad; Pourtaherian, Arash; Tang, Xikai; Zinger, Svitlana; Mihajlovic, Nenad; Ng, Gary C.; Korsten, Hendrikus H.M.; de With, Peter H.N.

Medical Imaging 2018: Image-Guided Procedures, Robotic Interventions, and Modeling. ed. / B. Fei; R.J. Webster. Bellingham : SPIE, 2018. p. 1-8 105760W (Proceedings of SPIE; Vol. 10576).

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

TY - GEN

T1 - Coherent needle detection in ultrasound volumes using 3D conditional random fields

AU - Ghazvinian Zanjani, Farhad

AU - Pourtaherian, Arash

AU - Tang, Xikai

AU - Zinger, Svitlana

AU - Mihajlovic, Nenad

AU - Ng, Gary C.

AU - Korsten, Hendrikus H.M.

AU - de With, Peter H.N.

PY - 2018/3/13

Y1 - 2018/3/13

N2 - 3D ultrasound (US) transducers will improve the quality of image-guided medical interventions if an automated detection of the needle becomes possible. Image-based detection of the needle is challenging due to the presence of other echogenic structures in the acquired data, inconsistent visibility of needle parts and the low quality in US imaging. As the currently applied approaches for needle detection classify each voxel individually, they do not consider the global relations between the voxels. In this work, we introduce coherent needle labeling by using dense conditional random fields over a volume, along with 3D space-frequency features. The proposal includes long-distance dependencies in voxel pairs according to their similarities in the feature space and their spatial distance. This post-processing stage leads to better label assignment of volume voxels and a more compact and coherent segmented region. Our ex-vivo experiments based on measuring the F-1, F-2 and IoU scores show that the performance improves a significant 10-20 % compared with only using the linear SVM as a baseline for voxel classification.

AB - 3D ultrasound (US) transducers will improve the quality of image-guided medical interventions if an automated detection of the needle becomes possible. Image-based detection of the needle is challenging due to the presence of other echogenic structures in the acquired data, inconsistent visibility of needle parts and the low quality in US imaging. As the currently applied approaches for needle detection classify each voxel individually, they do not consider the global relations between the voxels. In this work, we introduce coherent needle labeling by using dense conditional random fields over a volume, along with 3D space-frequency features. The proposal includes long-distance dependencies in voxel pairs according to their similarities in the feature space and their spatial distance. This post-processing stage leads to better label assignment of volume voxels and a more compact and coherent segmented region. Our ex-vivo experiments based on measuring the F-1, F-2 and IoU scores show that the performance improves a significant 10-20 % compared with only using the linear SVM as a baseline for voxel classification.

KW - 3D ultrasound

KW - Conditional Random Fields

KW - Needle detection

UR - http://www.scopus.com/inward/record.url?scp=85050677490&partnerID=8YFLogxK

U2 - 10.1117/12.2293079

DO - 10.1117/12.2293079

M3 - Conference contribution

AN - SCOPUS:85050677490

T3 - Proceedings of SPIE

SP - 1

EP - 8

BT - Medical Imaging 2018

A2 - Fei, B.

A2 - Webster, R.J.

PB - SPIE

CY - Bellingham

ER -

Ghazvinian Zanjani F, Pourtaherian A, Tang X, Zinger S, Mihajlovic N, Ng GC et al. Coherent needle detection in ultrasound volumes using 3D conditional random fields. In Fei B, Webster RJ, editors, Medical Imaging 2018: Image-Guided Procedures, Robotic Interventions, and Modeling. Bellingham: SPIE. 2018. p. 1-8. 105760W. (Proceedings of SPIE). https://doi.org/10.1117/12.2293079