Automated bed detection and removal from abdominal CT images for automatic segmentation applications

Ashrani Aizzuddin Abd Rahni, Muhamad Fazwan Mohamed Fuzaie, Omar Ibrahim Al Irr

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Image segmentation is an important component of automated image based diagnosis, since these application need information in form of the location of the anatomy of the patient. In order to realise a fully automated system, the exact location of the patient in the image should not affect an automated image processing algorithm. An initial step in automatic localisation of the patient is to identify where the bed is located in the image. For this work we focused on locating the bed in 3D from a volumetric abdominal CT dataset intended for liver cancer diagnosis. Our method is based on low level image processing namely multiple thresholding to automatically identify the field of view of the image, and identify the patient’s body separately from the bed, if the bed is within the field of view. The bed is then removed so as to not affect further image processing steps. The method was chosen to be fast and generic, so that it is applicable for a wide range of abdominal CT images, regardless of the CT scanner used, use of contrast, image resolution, field of view (FOV) and location of patient and bed. We tested our method on the publicly available Liver Tumor Segmentation (LiTS) Challenge dataset which comprises of anonymised abdominal CT volumes for over 100 patients. Our proposed method is able to correctly identify and remove the bed from all CT volumes we evaluated with.

Original languageEnglish
Title of host publication2018 IEEE EMBS Conference on Biomedical Engineering and Sciences, IECBES 2018 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages677-679
Number of pages3
ISBN (Electronic)9781538624715
DOIs
Publication statusPublished - 24 Jan 2019
Event2018 IEEE EMBS Conference on Biomedical Engineering and Sciences, IECBES 2018 - Kuching, Malaysia
Duration: 3 Dec 20186 Dec 2018

Publication series

Name2018 IEEE EMBS Conference on Biomedical Engineering and Sciences, IECBES 2018 - Proceedings

Conference

Conference2018 IEEE EMBS Conference on Biomedical Engineering and Sciences, IECBES 2018
CountryMalaysia
CityKuching
Period3/12/186/12/18

Fingerprint

beds
Image processing
Liver
Cone-Beam Computed Tomography
field of view
image processing
Image resolution
Image segmentation
Tumors
liver
Liver Neoplasms
anatomy
image resolution
Anatomy
scanners
tumors
cancer
Neoplasms
Datasets

Keywords

  • Abdominal CT
  • Bed removal
  • Volumetric segmentation

ASJC Scopus subject areas

  • Biomedical Engineering
  • Medicine (miscellaneous)
  • Health Informatics
  • Instrumentation

Cite this

Abd Rahni, A. A., Mohamed Fuzaie, M. F., & Al Irr, O. I. (2019). Automated bed detection and removal from abdominal CT images for automatic segmentation applications. In 2018 IEEE EMBS Conference on Biomedical Engineering and Sciences, IECBES 2018 - Proceedings (pp. 677-679). [8626638] (2018 IEEE EMBS Conference on Biomedical Engineering and Sciences, IECBES 2018 - Proceedings). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/IECBES.2018.8626638

Automated bed detection and removal from abdominal CT images for automatic segmentation applications. / Abd Rahni, Ashrani Aizzuddin; Mohamed Fuzaie, Muhamad Fazwan; Al Irr, Omar Ibrahim.

2018 IEEE EMBS Conference on Biomedical Engineering and Sciences, IECBES 2018 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2019. p. 677-679 8626638 (2018 IEEE EMBS Conference on Biomedical Engineering and Sciences, IECBES 2018 - Proceedings).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abd Rahni, AA, Mohamed Fuzaie, MF & Al Irr, OI 2019, Automated bed detection and removal from abdominal CT images for automatic segmentation applications. in 2018 IEEE EMBS Conference on Biomedical Engineering and Sciences, IECBES 2018 - Proceedings., 8626638, 2018 IEEE EMBS Conference on Biomedical Engineering and Sciences, IECBES 2018 - Proceedings, Institute of Electrical and Electronics Engineers Inc., pp. 677-679, 2018 IEEE EMBS Conference on Biomedical Engineering and Sciences, IECBES 2018, Kuching, Malaysia, 3/12/18. https://doi.org/10.1109/IECBES.2018.8626638
Abd Rahni AA, Mohamed Fuzaie MF, Al Irr OI. Automated bed detection and removal from abdominal CT images for automatic segmentation applications. In 2018 IEEE EMBS Conference on Biomedical Engineering and Sciences, IECBES 2018 - Proceedings. Institute of Electrical and Electronics Engineers Inc. 2019. p. 677-679. 8626638. (2018 IEEE EMBS Conference on Biomedical Engineering and Sciences, IECBES 2018 - Proceedings). https://doi.org/10.1109/IECBES.2018.8626638
Abd Rahni, Ashrani Aizzuddin ; Mohamed Fuzaie, Muhamad Fazwan ; Al Irr, Omar Ibrahim. / Automated bed detection and removal from abdominal CT images for automatic segmentation applications. 2018 IEEE EMBS Conference on Biomedical Engineering and Sciences, IECBES 2018 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2019. pp. 677-679 (2018 IEEE EMBS Conference on Biomedical Engineering and Sciences, IECBES 2018 - Proceedings).
@inproceedings{995b5e2ae58f41c5b403a3e1816d1a27,
title = "Automated bed detection and removal from abdominal CT images for automatic segmentation applications",
abstract = "Image segmentation is an important component of automated image based diagnosis, since these application need information in form of the location of the anatomy of the patient. In order to realise a fully automated system, the exact location of the patient in the image should not affect an automated image processing algorithm. An initial step in automatic localisation of the patient is to identify where the bed is located in the image. For this work we focused on locating the bed in 3D from a volumetric abdominal CT dataset intended for liver cancer diagnosis. Our method is based on low level image processing namely multiple thresholding to automatically identify the field of view of the image, and identify the patient’s body separately from the bed, if the bed is within the field of view. The bed is then removed so as to not affect further image processing steps. The method was chosen to be fast and generic, so that it is applicable for a wide range of abdominal CT images, regardless of the CT scanner used, use of contrast, image resolution, field of view (FOV) and location of patient and bed. We tested our method on the publicly available Liver Tumor Segmentation (LiTS) Challenge dataset which comprises of anonymised abdominal CT volumes for over 100 patients. Our proposed method is able to correctly identify and remove the bed from all CT volumes we evaluated with.",
keywords = "Abdominal CT, Bed removal, Volumetric segmentation",
author = "{Abd Rahni}, {Ashrani Aizzuddin} and {Mohamed Fuzaie}, {Muhamad Fazwan} and {Al Irr}, {Omar Ibrahim}",
year = "2019",
month = "1",
day = "24",
doi = "10.1109/IECBES.2018.8626638",
language = "English",
series = "2018 IEEE EMBS Conference on Biomedical Engineering and Sciences, IECBES 2018 - Proceedings",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "677--679",
booktitle = "2018 IEEE EMBS Conference on Biomedical Engineering and Sciences, IECBES 2018 - Proceedings",
address = "United States",

}

TY - GEN

T1 - Automated bed detection and removal from abdominal CT images for automatic segmentation applications

AU - Abd Rahni, Ashrani Aizzuddin

AU - Mohamed Fuzaie, Muhamad Fazwan

AU - Al Irr, Omar Ibrahim

PY - 2019/1/24

Y1 - 2019/1/24

N2 - Image segmentation is an important component of automated image based diagnosis, since these application need information in form of the location of the anatomy of the patient. In order to realise a fully automated system, the exact location of the patient in the image should not affect an automated image processing algorithm. An initial step in automatic localisation of the patient is to identify where the bed is located in the image. For this work we focused on locating the bed in 3D from a volumetric abdominal CT dataset intended for liver cancer diagnosis. Our method is based on low level image processing namely multiple thresholding to automatically identify the field of view of the image, and identify the patient’s body separately from the bed, if the bed is within the field of view. The bed is then removed so as to not affect further image processing steps. The method was chosen to be fast and generic, so that it is applicable for a wide range of abdominal CT images, regardless of the CT scanner used, use of contrast, image resolution, field of view (FOV) and location of patient and bed. We tested our method on the publicly available Liver Tumor Segmentation (LiTS) Challenge dataset which comprises of anonymised abdominal CT volumes for over 100 patients. Our proposed method is able to correctly identify and remove the bed from all CT volumes we evaluated with.

AB - Image segmentation is an important component of automated image based diagnosis, since these application need information in form of the location of the anatomy of the patient. In order to realise a fully automated system, the exact location of the patient in the image should not affect an automated image processing algorithm. An initial step in automatic localisation of the patient is to identify where the bed is located in the image. For this work we focused on locating the bed in 3D from a volumetric abdominal CT dataset intended for liver cancer diagnosis. Our method is based on low level image processing namely multiple thresholding to automatically identify the field of view of the image, and identify the patient’s body separately from the bed, if the bed is within the field of view. The bed is then removed so as to not affect further image processing steps. The method was chosen to be fast and generic, so that it is applicable for a wide range of abdominal CT images, regardless of the CT scanner used, use of contrast, image resolution, field of view (FOV) and location of patient and bed. We tested our method on the publicly available Liver Tumor Segmentation (LiTS) Challenge dataset which comprises of anonymised abdominal CT volumes for over 100 patients. Our proposed method is able to correctly identify and remove the bed from all CT volumes we evaluated with.

KW - Abdominal CT

KW - Bed removal

KW - Volumetric segmentation

UR - http://www.scopus.com/inward/record.url?scp=85062791934&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85062791934&partnerID=8YFLogxK

U2 - 10.1109/IECBES.2018.8626638

DO - 10.1109/IECBES.2018.8626638

M3 - Conference contribution

AN - SCOPUS:85062791934

T3 - 2018 IEEE EMBS Conference on Biomedical Engineering and Sciences, IECBES 2018 - Proceedings

SP - 677

EP - 679

BT - 2018 IEEE EMBS Conference on Biomedical Engineering and Sciences, IECBES 2018 - Proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

ER -