A deep learning approach for slice to volume biomedical image integration

Bassam Almogadwy, Kenneth McLeod, Albert Burger

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

Abstract

Biomedical atlas images obtained from multiple sources need to be aligned and transformed into a single coordinate system so as to be able to integrate and relate these different sets of data. Formally known as image registration, this process of image pre-processing has proven to be integral in a wide array of computer vision applications, most notably in the area of medical imaging. During the last decade slice-to-volume registration, a particular case of image registration problem, has received further attention in the medical imaging community due to the emergence of several medical applications of slice-to-volume mapping. This paper proposes a Convolutional Neural Network (CNN) based deep learning approach for registering a 2D image slice to the 3D volume of images in a Biomedical atlas. The proposed CNN model is trained to determine the distance and pitch values that are used to describe the position of the 2D slice in the atlas coordinate system. High-level features are automatically extracted from the training dataset of images, which addresses the limitation of shallow machine learning techniques for handcrafted features followed by the classification task. Then on the basis of predicted values of distance and pitch, the target image is registered to the 3D volume of images. Experimental results showing the effect on the similarity of images with variation in distance and the impact of varying the distances among the classes on the regression are presented. It was observed that using the successive images at a distance of 10 lead to the maximum accuracy. These results demonstrate the applicability of the proposed approach to slice-to-volume image registration.

Original languageEnglish
Title of host publicationICBBT'19 Proceedings of the 2019 11th International Conference on Bioinformatics and Biomedical Technology
PublisherACM
Pages62-68
Number of pages7
ISBN (Electronic)9781450362313
DOIs
Publication statusPublished - 29 May 2019
Event11th International Conference on Bioinformatics and Biomedical Technology 2019 - Stockholm, Sweden
Duration: 29 May 201931 May 2019

Conference

Conference11th International Conference on Bioinformatics and Biomedical Technology 2019
Abbreviated titleICBBT 2019
CountrySweden
CityStockholm
Period29/05/1931/05/19

Fingerprint

Image registration
Medical imaging
Neural networks
Medical applications
Computer vision
Learning systems
Processing
Deep learning

Keywords

  • Biomedical atlases
  • Deep learning
  • Image registration

ASJC Scopus subject areas

  • Software
  • Human-Computer Interaction
  • Computer Vision and Pattern Recognition
  • Computer Networks and Communications

Cite this

Almogadwy, B., McLeod, K., & Burger, A. (2019). A deep learning approach for slice to volume biomedical image integration. In ICBBT'19 Proceedings of the 2019 11th International Conference on Bioinformatics and Biomedical Technology (pp. 62-68). ACM. https://doi.org/10.1145/3340074.3340089
Almogadwy, Bassam ; McLeod, Kenneth ; Burger, Albert. / A deep learning approach for slice to volume biomedical image integration. ICBBT'19 Proceedings of the 2019 11th International Conference on Bioinformatics and Biomedical Technology. ACM, 2019. pp. 62-68
@inproceedings{5babdb4a7c9b484b951fc1f30f6689f0,
title = "A deep learning approach for slice to volume biomedical image integration",
abstract = "Biomedical atlas images obtained from multiple sources need to be aligned and transformed into a single coordinate system so as to be able to integrate and relate these different sets of data. Formally known as image registration, this process of image pre-processing has proven to be integral in a wide array of computer vision applications, most notably in the area of medical imaging. During the last decade slice-to-volume registration, a particular case of image registration problem, has received further attention in the medical imaging community due to the emergence of several medical applications of slice-to-volume mapping. This paper proposes a Convolutional Neural Network (CNN) based deep learning approach for registering a 2D image slice to the 3D volume of images in a Biomedical atlas. The proposed CNN model is trained to determine the distance and pitch values that are used to describe the position of the 2D slice in the atlas coordinate system. High-level features are automatically extracted from the training dataset of images, which addresses the limitation of shallow machine learning techniques for handcrafted features followed by the classification task. Then on the basis of predicted values of distance and pitch, the target image is registered to the 3D volume of images. Experimental results showing the effect on the similarity of images with variation in distance and the impact of varying the distances among the classes on the regression are presented. It was observed that using the successive images at a distance of 10 lead to the maximum accuracy. These results demonstrate the applicability of the proposed approach to slice-to-volume image registration.",
keywords = "Biomedical atlases, Deep learning, Image registration",
author = "Bassam Almogadwy and Kenneth McLeod and Albert Burger",
year = "2019",
month = "5",
day = "29",
doi = "10.1145/3340074.3340089",
language = "English",
pages = "62--68",
booktitle = "ICBBT'19 Proceedings of the 2019 11th International Conference on Bioinformatics and Biomedical Technology",
publisher = "ACM",

}

Almogadwy, B, McLeod, K & Burger, A 2019, A deep learning approach for slice to volume biomedical image integration. in ICBBT'19 Proceedings of the 2019 11th International Conference on Bioinformatics and Biomedical Technology. ACM, pp. 62-68, 11th International Conference on Bioinformatics and Biomedical Technology 2019, Stockholm, Sweden, 29/05/19. https://doi.org/10.1145/3340074.3340089

A deep learning approach for slice to volume biomedical image integration. / Almogadwy, Bassam; McLeod, Kenneth; Burger, Albert.

ICBBT'19 Proceedings of the 2019 11th International Conference on Bioinformatics and Biomedical Technology. ACM, 2019. p. 62-68.

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

TY - GEN

T1 - A deep learning approach for slice to volume biomedical image integration

AU - Almogadwy, Bassam

AU - McLeod, Kenneth

AU - Burger, Albert

PY - 2019/5/29

Y1 - 2019/5/29

N2 - Biomedical atlas images obtained from multiple sources need to be aligned and transformed into a single coordinate system so as to be able to integrate and relate these different sets of data. Formally known as image registration, this process of image pre-processing has proven to be integral in a wide array of computer vision applications, most notably in the area of medical imaging. During the last decade slice-to-volume registration, a particular case of image registration problem, has received further attention in the medical imaging community due to the emergence of several medical applications of slice-to-volume mapping. This paper proposes a Convolutional Neural Network (CNN) based deep learning approach for registering a 2D image slice to the 3D volume of images in a Biomedical atlas. The proposed CNN model is trained to determine the distance and pitch values that are used to describe the position of the 2D slice in the atlas coordinate system. High-level features are automatically extracted from the training dataset of images, which addresses the limitation of shallow machine learning techniques for handcrafted features followed by the classification task. Then on the basis of predicted values of distance and pitch, the target image is registered to the 3D volume of images. Experimental results showing the effect on the similarity of images with variation in distance and the impact of varying the distances among the classes on the regression are presented. It was observed that using the successive images at a distance of 10 lead to the maximum accuracy. These results demonstrate the applicability of the proposed approach to slice-to-volume image registration.

AB - Biomedical atlas images obtained from multiple sources need to be aligned and transformed into a single coordinate system so as to be able to integrate and relate these different sets of data. Formally known as image registration, this process of image pre-processing has proven to be integral in a wide array of computer vision applications, most notably in the area of medical imaging. During the last decade slice-to-volume registration, a particular case of image registration problem, has received further attention in the medical imaging community due to the emergence of several medical applications of slice-to-volume mapping. This paper proposes a Convolutional Neural Network (CNN) based deep learning approach for registering a 2D image slice to the 3D volume of images in a Biomedical atlas. The proposed CNN model is trained to determine the distance and pitch values that are used to describe the position of the 2D slice in the atlas coordinate system. High-level features are automatically extracted from the training dataset of images, which addresses the limitation of shallow machine learning techniques for handcrafted features followed by the classification task. Then on the basis of predicted values of distance and pitch, the target image is registered to the 3D volume of images. Experimental results showing the effect on the similarity of images with variation in distance and the impact of varying the distances among the classes on the regression are presented. It was observed that using the successive images at a distance of 10 lead to the maximum accuracy. These results demonstrate the applicability of the proposed approach to slice-to-volume image registration.

KW - Biomedical atlases

KW - Deep learning

KW - Image registration

UR - http://www.scopus.com/inward/record.url?scp=85070565458&partnerID=8YFLogxK

U2 - 10.1145/3340074.3340089

DO - 10.1145/3340074.3340089

M3 - Conference contribution

SP - 62

EP - 68

BT - ICBBT'19 Proceedings of the 2019 11th International Conference on Bioinformatics and Biomedical Technology

PB - ACM

ER -

Almogadwy B, McLeod K, Burger A. A deep learning approach for slice to volume biomedical image integration. In ICBBT'19 Proceedings of the 2019 11th International Conference on Bioinformatics and Biomedical Technology. ACM. 2019. p. 62-68 https://doi.org/10.1145/3340074.3340089