Automated integer programming based separation of arteries and veins from thoracic CT images

Christian Payer, Michael Pienn, Zoltán Bálint, Oleksandr Shekhovtsov, Emina Talakic, Eszter Nagy, Andrea Olschewski, Horst Olschewski, Martin Urschler

Research output: Contribution to journalArticleResearchpeer-review

Abstract

Automated computer-aided analysis of lung vessels has shown to yield promising results for non-invasive diagnosis of lung diseases. To detect vascular changes which affect pulmonary arteries and veins differently, both compartments need to be identified. We present a novel, fully automatic method that separates arteries and veins in thoracic computed tomography images, by combining local as well as global properties of pulmonary vessels. We split the problem into two parts: the extraction of multiple distinct vessel subtrees, and their subsequent labeling into arteries and veins. Subtree extraction is performed with an integer program (IP), based on local vessel geometry. As naively solving this IP is time-consuming, we show how to drastically reduce computational effort by reformulating it as a Markov Random Field. Afterwards, each subtree is labeled as either arterial or venous by a second IP, using two anatomical properties of pulmonary vessels: the uniform distribution of arteries and veins, and the parallel configuration and close proximity of arteries and bronchi. We evaluate algorithm performance by comparing the results with 25 voxel-based manual reference segmentations. On this dataset, we show good performance of the subtree extraction, consisting of very few non-vascular structures (median value: 0.9%) and merged subtrees (median value: 0.6%). The resulting separation of arteries and veins achieves a median voxel-based overlap of 96.3% with the manual reference segmentations, outperforming a state-of-the-art interactive method. In conclusion, our novel approach provides an opportunity to become an integral part of computer aided pulmonary diagnosis, where artery/vein separation is important.

Original languageEnglish
Pages (from-to)109–122
Number of pages14
JournalMedical image analysis
Volume34
DOIs
Publication statusPublished - Dec 2016

Fingerprint

Thoracic Arteries
Integer programming
Veins
Arteries
Computer aided analysis
Pulmonary diseases
Lung
Pulmonary Artery
Labeling
Tomography
Pulmonary Veins
Bronchi
Geometry
Lung Diseases
Blood Vessels
Thorax

Fields of Expertise

  • Information, Communication & Computing

Cooperations

  • BioTechMed-Graz

Cite this

Payer, C., Pienn, M., Bálint, Z., Shekhovtsov, O., Talakic, E., Nagy, E., ... Urschler, M. (2016). Automated integer programming based separation of arteries and veins from thoracic CT images. Medical image analysis, 34, 109–122. https://doi.org/10.1016/j.media.2016.05.002

Automated integer programming based separation of arteries and veins from thoracic CT images. / Payer, Christian; Pienn, Michael; Bálint, Zoltán; Shekhovtsov, Oleksandr; Talakic, Emina; Nagy, Eszter; Olschewski, Andrea; Olschewski, Horst; Urschler, Martin.

In: Medical image analysis, Vol. 34, 12.2016, p. 109–122.

Research output: Contribution to journalArticleResearchpeer-review

Payer, C, Pienn, M, Bálint, Z, Shekhovtsov, O, Talakic, E, Nagy, E, Olschewski, A, Olschewski, H & Urschler, M 2016, 'Automated integer programming based separation of arteries and veins from thoracic CT images' Medical image analysis, vol. 34, pp. 109–122. https://doi.org/10.1016/j.media.2016.05.002
Payer, Christian ; Pienn, Michael ; Bálint, Zoltán ; Shekhovtsov, Oleksandr ; Talakic, Emina ; Nagy, Eszter ; Olschewski, Andrea ; Olschewski, Horst ; Urschler, Martin. / Automated integer programming based separation of arteries and veins from thoracic CT images. In: Medical image analysis. 2016 ; Vol. 34. pp. 109–122.
@article{f3dd52cf657f4c948fa2d86d11c4da6d,
title = "Automated integer programming based separation of arteries and veins from thoracic CT images",
abstract = "Automated computer-aided analysis of lung vessels has shown to yield promising results for non-invasive diagnosis of lung diseases. To detect vascular changes which affect pulmonary arteries and veins differently, both compartments need to be identified. We present a novel, fully automatic method that separates arteries and veins in thoracic computed tomography images, by combining local as well as global properties of pulmonary vessels. We split the problem into two parts: the extraction of multiple distinct vessel subtrees, and their subsequent labeling into arteries and veins. Subtree extraction is performed with an integer program (IP), based on local vessel geometry. As naively solving this IP is time-consuming, we show how to drastically reduce computational effort by reformulating it as a Markov Random Field. Afterwards, each subtree is labeled as either arterial or venous by a second IP, using two anatomical properties of pulmonary vessels: the uniform distribution of arteries and veins, and the parallel configuration and close proximity of arteries and bronchi. We evaluate algorithm performance by comparing the results with 25 voxel-based manual reference segmentations. On this dataset, we show good performance of the subtree extraction, consisting of very few non-vascular structures (median value: 0.9{\%}) and merged subtrees (median value: 0.6{\%}). The resulting separation of arteries and veins achieves a median voxel-based overlap of 96.3{\%} with the manual reference segmentations, outperforming a state-of-the-art interactive method. In conclusion, our novel approach provides an opportunity to become an integral part of computer aided pulmonary diagnosis, where artery/vein separation is important.",
author = "Christian Payer and Michael Pienn and Zolt{\'a}n B{\'a}lint and Oleksandr Shekhovtsov and Emina Talakic and Eszter Nagy and Andrea Olschewski and Horst Olschewski and Martin Urschler",
note = "Copyright {\circledC} 2016 Elsevier B.V. All rights reserved.",
year = "2016",
month = "12",
doi = "10.1016/j.media.2016.05.002",
language = "English",
volume = "34",
pages = "109–122",
journal = "Medical image analysis",
issn = "1361-8415",
publisher = "Elsevier B.V.",

}

TY - JOUR

T1 - Automated integer programming based separation of arteries and veins from thoracic CT images

AU - Payer, Christian

AU - Pienn, Michael

AU - Bálint, Zoltán

AU - Shekhovtsov, Oleksandr

AU - Talakic, Emina

AU - Nagy, Eszter

AU - Olschewski, Andrea

AU - Olschewski, Horst

AU - Urschler, Martin

N1 - Copyright © 2016 Elsevier B.V. All rights reserved.

PY - 2016/12

Y1 - 2016/12

N2 - Automated computer-aided analysis of lung vessels has shown to yield promising results for non-invasive diagnosis of lung diseases. To detect vascular changes which affect pulmonary arteries and veins differently, both compartments need to be identified. We present a novel, fully automatic method that separates arteries and veins in thoracic computed tomography images, by combining local as well as global properties of pulmonary vessels. We split the problem into two parts: the extraction of multiple distinct vessel subtrees, and their subsequent labeling into arteries and veins. Subtree extraction is performed with an integer program (IP), based on local vessel geometry. As naively solving this IP is time-consuming, we show how to drastically reduce computational effort by reformulating it as a Markov Random Field. Afterwards, each subtree is labeled as either arterial or venous by a second IP, using two anatomical properties of pulmonary vessels: the uniform distribution of arteries and veins, and the parallel configuration and close proximity of arteries and bronchi. We evaluate algorithm performance by comparing the results with 25 voxel-based manual reference segmentations. On this dataset, we show good performance of the subtree extraction, consisting of very few non-vascular structures (median value: 0.9%) and merged subtrees (median value: 0.6%). The resulting separation of arteries and veins achieves a median voxel-based overlap of 96.3% with the manual reference segmentations, outperforming a state-of-the-art interactive method. In conclusion, our novel approach provides an opportunity to become an integral part of computer aided pulmonary diagnosis, where artery/vein separation is important.

AB - Automated computer-aided analysis of lung vessels has shown to yield promising results for non-invasive diagnosis of lung diseases. To detect vascular changes which affect pulmonary arteries and veins differently, both compartments need to be identified. We present a novel, fully automatic method that separates arteries and veins in thoracic computed tomography images, by combining local as well as global properties of pulmonary vessels. We split the problem into two parts: the extraction of multiple distinct vessel subtrees, and their subsequent labeling into arteries and veins. Subtree extraction is performed with an integer program (IP), based on local vessel geometry. As naively solving this IP is time-consuming, we show how to drastically reduce computational effort by reformulating it as a Markov Random Field. Afterwards, each subtree is labeled as either arterial or venous by a second IP, using two anatomical properties of pulmonary vessels: the uniform distribution of arteries and veins, and the parallel configuration and close proximity of arteries and bronchi. We evaluate algorithm performance by comparing the results with 25 voxel-based manual reference segmentations. On this dataset, we show good performance of the subtree extraction, consisting of very few non-vascular structures (median value: 0.9%) and merged subtrees (median value: 0.6%). The resulting separation of arteries and veins achieves a median voxel-based overlap of 96.3% with the manual reference segmentations, outperforming a state-of-the-art interactive method. In conclusion, our novel approach provides an opportunity to become an integral part of computer aided pulmonary diagnosis, where artery/vein separation is important.

U2 - 10.1016/j.media.2016.05.002

DO - 10.1016/j.media.2016.05.002

M3 - Article

VL - 34

SP - 109

EP - 122

JO - Medical image analysis

JF - Medical image analysis

SN - 1361-8415

ER -