Accelerating 3D medical volume segmentation using GPUs

Mahmoud Al-Ayyoub; Shadi AlZu’bi; Yaser Jararweh; Mohammed A. Shehab; Brij B. Gupta

doi:10.1007/s11042-016-4218-0

Accelerating 3D medical volume segmentation using GPUs

Mahmoud Al-Ayyoub
, Shadi AlZu’bi
, Yaser Jararweh
, Mohammed A. Shehab
, Brij B. Gupta

Jordan University of Science and Technology
Al-Zaytoonah University of Jordan
National Institute of Technology Kurukshetra

Research output: Contribution to journal › Article › peer-review

73 Scopus citations

Abstract

Medical images have an undeniably integral role in the process of diagnosing and treating of a very large number of ailments. Processing such images (for different purposes) can significantly improve the efficiency and effectiveness of this process. The first step in many medical image processing applications is segmentation, which is used to extract the Region of Interest (ROI) from a given image. Due to its effectiveness, a very popular segmentation algorithm is the Fuzzy C-Means (FCM) algorithm. However, FCM takes a long processing time especially for 3D model. This problem can be solved by utilizing parallel programming using Graphics Processing Unit (GPU). In this paper, a hybrid parallel implementation of FCM for extracting volume object from medical DICOM files has been proposed. The proposed algorithm improves the performance 5× compared with the sequential version.

Original language	English
Pages (from-to)	4939-4958
Number of pages	20
Journal	Multimedia Tools and Applications
Volume	77
Issue number	4
DOIs	https://doi.org/10.1007/s11042-016-4218-0
State	Published - 1 Feb 2018
Externally published	Yes

Keywords

3D segmentation
Fuzzy C-Means (FCM) algorithm
Medical image processing
Parallel programming

Access to Document

10.1007/s11042-016-4218-0

Cite this

@article{086cea098dfd45b99a5d69b0f6e0adbe,

title = "Accelerating 3D medical volume segmentation using GPUs",

abstract = "Medical images have an undeniably integral role in the process of diagnosing and treating of a very large number of ailments. Processing such images (for different purposes) can significantly improve the efficiency and effectiveness of this process. The first step in many medical image processing applications is segmentation, which is used to extract the Region of Interest (ROI) from a given image. Due to its effectiveness, a very popular segmentation algorithm is the Fuzzy C-Means (FCM) algorithm. However, FCM takes a long processing time especially for 3D model. This problem can be solved by utilizing parallel programming using Graphics Processing Unit (GPU). In this paper, a hybrid parallel implementation of FCM for extracting volume object from medical DICOM files has been proposed. The proposed algorithm improves the performance 5× compared with the sequential version.",

keywords = "3D segmentation, Fuzzy C-Means (FCM) algorithm, Medical image processing, Parallel programming",

author = "Mahmoud Al-Ayyoub and Shadi AlZu{\textquoteright}bi and Yaser Jararweh and Shehab, \{Mohammed A.\} and Gupta, \{Brij B.\}",

note = "Publisher Copyright: {\textcopyright} 2016, Springer Science+Business Media New York.",

year = "2018",

month = feb,

day = "1",

doi = "10.1007/s11042-016-4218-0",

language = "English",

volume = "77",

pages = "4939--4958",

journal = "Multimedia Tools and Applications",

issn = "1380-7501",

publisher = "Springer",

number = "4",

}

TY - JOUR

T1 - Accelerating 3D medical volume segmentation using GPUs

AU - Al-Ayyoub, Mahmoud

AU - AlZu’bi, Shadi

AU - Jararweh, Yaser

AU - Shehab, Mohammed A.

AU - Gupta, Brij B.

PY - 2018/2/1

Y1 - 2018/2/1

N2 - Medical images have an undeniably integral role in the process of diagnosing and treating of a very large number of ailments. Processing such images (for different purposes) can significantly improve the efficiency and effectiveness of this process. The first step in many medical image processing applications is segmentation, which is used to extract the Region of Interest (ROI) from a given image. Due to its effectiveness, a very popular segmentation algorithm is the Fuzzy C-Means (FCM) algorithm. However, FCM takes a long processing time especially for 3D model. This problem can be solved by utilizing parallel programming using Graphics Processing Unit (GPU). In this paper, a hybrid parallel implementation of FCM for extracting volume object from medical DICOM files has been proposed. The proposed algorithm improves the performance 5× compared with the sequential version.

AB - Medical images have an undeniably integral role in the process of diagnosing and treating of a very large number of ailments. Processing such images (for different purposes) can significantly improve the efficiency and effectiveness of this process. The first step in many medical image processing applications is segmentation, which is used to extract the Region of Interest (ROI) from a given image. Due to its effectiveness, a very popular segmentation algorithm is the Fuzzy C-Means (FCM) algorithm. However, FCM takes a long processing time especially for 3D model. This problem can be solved by utilizing parallel programming using Graphics Processing Unit (GPU). In this paper, a hybrid parallel implementation of FCM for extracting volume object from medical DICOM files has been proposed. The proposed algorithm improves the performance 5× compared with the sequential version.

KW - 3D segmentation

KW - Fuzzy C-Means (FCM) algorithm

KW - Medical image processing

KW - Parallel programming

UR - https://www.scopus.com/pages/publications/85006489555

U2 - 10.1007/s11042-016-4218-0

DO - 10.1007/s11042-016-4218-0

M3 - Article

AN - SCOPUS:85006489555

SN - 1380-7501

VL - 77

SP - 4939

EP - 4958

JO - Multimedia Tools and Applications

JF - Multimedia Tools and Applications

IS - 4

ER -

Accelerating 3D medical volume segmentation using GPUs

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this