Extraction of brain tissue from CT head images using fully convolutional neural networks

Zeynettin Akkus, Petro M. Kostandy, Kenneth A. Philbrick, Bradley J Erickson

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

4 Citations (Scopus)

Abstract

Removing non-brain tissues such as the skull, scalp and face from head computed tomography (CT) images is an important field of study in brain image processing applications. It is a prerequisite step in numerous quantitative imaging analyses of neurological diseases as it improves the computational speed and accuracy of quantitative analyses and image coregistration. In this study, we present an accurate method based on fully convolutional neural networks (fCNN) to remove non-brain tissues from head CT images in a time-efficient manner. The method includes an encoding part which has sequential convolutional filters that produce activation maps of the input image in low dimensional space; and it has a decoding part consisting of convolutional filters that reconstruct the input image from the reduced representation. We trained the fCNN on 122 volumetric head CT images and tested on 22 unseen volumetric CT head images based on an expert's manual brain segmentation masks. The performance of our method on the test set is: Dice Coefficient= 0.998±0.001 (mean ± standard deviation), recall=0.998±0.001, precision=0.998±0.001, and accuracy=0.9995±0.0001. Our method extracts complete volumetric brain from head CT images in 2s which is much faster than previous methods. To the best of our knowledge, this is the first study using fCNN to perform skull stripping from CT images. Our approach based on fCNN provides accurate extraction of brain tissue from head CT images in a time-efficient manner.

Original languageEnglish (US)
Title of host publicationMedical Imaging 2018
Subtitle of host publicationImage Processing
PublisherSPIE
Volume10574
ISBN (Electronic)9781510616370
DOIs
StatePublished - Jan 1 2018
EventMedical Imaging 2018: Image Processing - Houston, United States
Duration: Feb 11 2018Feb 13 2018

Other

OtherMedical Imaging 2018: Image Processing
CountryUnited States
CityHouston
Period2/11/182/13/18

Fingerprint

brain
Tomography
Brain
tomography
Head
Tissue
Neural networks
Cone-Beam Computed Tomography
Skull
skull
Masks
Scalp
filters
Decoding
Image processing
Chemical activation
decoding
stripping
Imaging techniques
image processing

Keywords

  • brain extraction
  • computed tomography
  • convolutional neural network
  • skull stripping

ASJC Scopus subject areas

  • Electronic, Optical and Magnetic Materials
  • Biomaterials
  • Atomic and Molecular Physics, and Optics
  • Radiology Nuclear Medicine and imaging

Cite this

Akkus, Z., Kostandy, P. M., Philbrick, K. A., & Erickson, B. J. (2018). Extraction of brain tissue from CT head images using fully convolutional neural networks. In Medical Imaging 2018: Image Processing (Vol. 10574). [1057420] SPIE. https://doi.org/10.1117/12.2293423

Extraction of brain tissue from CT head images using fully convolutional neural networks. / Akkus, Zeynettin; Kostandy, Petro M.; Philbrick, Kenneth A.; Erickson, Bradley J.

Medical Imaging 2018: Image Processing. Vol. 10574 SPIE, 2018. 1057420.

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

Akkus, Z, Kostandy, PM, Philbrick, KA & Erickson, BJ 2018, Extraction of brain tissue from CT head images using fully convolutional neural networks. in Medical Imaging 2018: Image Processing. vol. 10574, 1057420, SPIE, Medical Imaging 2018: Image Processing, Houston, United States, 2/11/18. https://doi.org/10.1117/12.2293423
Akkus Z, Kostandy PM, Philbrick KA, Erickson BJ. Extraction of brain tissue from CT head images using fully convolutional neural networks. In Medical Imaging 2018: Image Processing. Vol. 10574. SPIE. 2018. 1057420 https://doi.org/10.1117/12.2293423
Akkus, Zeynettin ; Kostandy, Petro M. ; Philbrick, Kenneth A. ; Erickson, Bradley J. / Extraction of brain tissue from CT head images using fully convolutional neural networks. Medical Imaging 2018: Image Processing. Vol. 10574 SPIE, 2018.
@inproceedings{79af0b7d606a44ef822f115393a21afb,
title = "Extraction of brain tissue from CT head images using fully convolutional neural networks",
abstract = "Removing non-brain tissues such as the skull, scalp and face from head computed tomography (CT) images is an important field of study in brain image processing applications. It is a prerequisite step in numerous quantitative imaging analyses of neurological diseases as it improves the computational speed and accuracy of quantitative analyses and image coregistration. In this study, we present an accurate method based on fully convolutional neural networks (fCNN) to remove non-brain tissues from head CT images in a time-efficient manner. The method includes an encoding part which has sequential convolutional filters that produce activation maps of the input image in low dimensional space; and it has a decoding part consisting of convolutional filters that reconstruct the input image from the reduced representation. We trained the fCNN on 122 volumetric head CT images and tested on 22 unseen volumetric CT head images based on an expert's manual brain segmentation masks. The performance of our method on the test set is: Dice Coefficient= 0.998±0.001 (mean ± standard deviation), recall=0.998±0.001, precision=0.998±0.001, and accuracy=0.9995±0.0001. Our method extracts complete volumetric brain from head CT images in 2s which is much faster than previous methods. To the best of our knowledge, this is the first study using fCNN to perform skull stripping from CT images. Our approach based on fCNN provides accurate extraction of brain tissue from head CT images in a time-efficient manner.",
keywords = "brain extraction, computed tomography, convolutional neural network, skull stripping",
author = "Zeynettin Akkus and Kostandy, {Petro M.} and Philbrick, {Kenneth A.} and Erickson, {Bradley J}",
year = "2018",
month = "1",
day = "1",
doi = "10.1117/12.2293423",
language = "English (US)",
volume = "10574",
booktitle = "Medical Imaging 2018",
publisher = "SPIE",

}

TY - GEN

T1 - Extraction of brain tissue from CT head images using fully convolutional neural networks

AU - Akkus, Zeynettin

AU - Kostandy, Petro M.

AU - Philbrick, Kenneth A.

AU - Erickson, Bradley J

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Removing non-brain tissues such as the skull, scalp and face from head computed tomography (CT) images is an important field of study in brain image processing applications. It is a prerequisite step in numerous quantitative imaging analyses of neurological diseases as it improves the computational speed and accuracy of quantitative analyses and image coregistration. In this study, we present an accurate method based on fully convolutional neural networks (fCNN) to remove non-brain tissues from head CT images in a time-efficient manner. The method includes an encoding part which has sequential convolutional filters that produce activation maps of the input image in low dimensional space; and it has a decoding part consisting of convolutional filters that reconstruct the input image from the reduced representation. We trained the fCNN on 122 volumetric head CT images and tested on 22 unseen volumetric CT head images based on an expert's manual brain segmentation masks. The performance of our method on the test set is: Dice Coefficient= 0.998±0.001 (mean ± standard deviation), recall=0.998±0.001, precision=0.998±0.001, and accuracy=0.9995±0.0001. Our method extracts complete volumetric brain from head CT images in 2s which is much faster than previous methods. To the best of our knowledge, this is the first study using fCNN to perform skull stripping from CT images. Our approach based on fCNN provides accurate extraction of brain tissue from head CT images in a time-efficient manner.

AB - Removing non-brain tissues such as the skull, scalp and face from head computed tomography (CT) images is an important field of study in brain image processing applications. It is a prerequisite step in numerous quantitative imaging analyses of neurological diseases as it improves the computational speed and accuracy of quantitative analyses and image coregistration. In this study, we present an accurate method based on fully convolutional neural networks (fCNN) to remove non-brain tissues from head CT images in a time-efficient manner. The method includes an encoding part which has sequential convolutional filters that produce activation maps of the input image in low dimensional space; and it has a decoding part consisting of convolutional filters that reconstruct the input image from the reduced representation. We trained the fCNN on 122 volumetric head CT images and tested on 22 unseen volumetric CT head images based on an expert's manual brain segmentation masks. The performance of our method on the test set is: Dice Coefficient= 0.998±0.001 (mean ± standard deviation), recall=0.998±0.001, precision=0.998±0.001, and accuracy=0.9995±0.0001. Our method extracts complete volumetric brain from head CT images in 2s which is much faster than previous methods. To the best of our knowledge, this is the first study using fCNN to perform skull stripping from CT images. Our approach based on fCNN provides accurate extraction of brain tissue from head CT images in a time-efficient manner.

KW - brain extraction

KW - computed tomography

KW - convolutional neural network

KW - skull stripping

UR - http://www.scopus.com/inward/record.url?scp=85047328084&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85047328084&partnerID=8YFLogxK

U2 - 10.1117/12.2293423

DO - 10.1117/12.2293423

M3 - Conference contribution

AN - SCOPUS:85047328084

VL - 10574

BT - Medical Imaging 2018

PB - SPIE

ER -