Two-layer competitive based Hopfield neural network for medical image edge detection

Chuan Yu Chang, Pau-Choo Chung

Research output: Contribution to journalArticle

18 Citations (Scopus)

Abstract

In medical applications, the detection and outlining of boundaries of organs and tumors in computed tomography (CT) and magnetic resonance imaging (MRI) images are prerequisite. A two-layer Hopfield neural network called the competitive Hopfield edge-finding neural network (CHEFNN) is presented for finding the edges of CT and MRI images. Different from conventional 2-D Hopfield neural networks, the CHEFNN extends the one-layer 2-D Hopfield network at the original image plane a two-layer 3-D Hopfield network with edge detection to be implemented on its third dimension. With the extended 3-D architecture, the network is capable of incorporating a pixel's contextual information into a pixel-labeling procedure. As a result, the effect of tiny details or noises will be effectively removed by the CHEFNN and the drawback of disconnected fractions can be overcome. Furthermore, by making use of the competitive learning rule to update the neuron states, the problem of satisfying strong constraints can be alleviated and results in a fast convergence. Our experimental results show that the CHEFNN can obtain more appropriate, more continued edge points than the Laplacian-based, Marr-Hildreth, Canny, and wavelet-based methods.

Original languageEnglish
Pages (from-to)695-703
Number of pages9
JournalOptical Engineering
Volume39
Issue number3
DOIs
Publication statusPublished - 2000 Mar 1

Fingerprint

Hopfield neural networks
edge detection
Edge detection
Neural networks
Magnetic resonance
Tomography
Pixels
Imaging techniques
Medical applications
Labeling
Neurons
Tumors
magnetic resonance
tomography
pixels
neurons
organs
learning
marking
tumors

All Science Journal Classification (ASJC) codes

  • Atomic and Molecular Physics, and Optics
  • Engineering(all)

Cite this

@article{cad6ce31bd6d4f8a9bbf5e781eb7b00e,
title = "Two-layer competitive based Hopfield neural network for medical image edge detection",
abstract = "In medical applications, the detection and outlining of boundaries of organs and tumors in computed tomography (CT) and magnetic resonance imaging (MRI) images are prerequisite. A two-layer Hopfield neural network called the competitive Hopfield edge-finding neural network (CHEFNN) is presented for finding the edges of CT and MRI images. Different from conventional 2-D Hopfield neural networks, the CHEFNN extends the one-layer 2-D Hopfield network at the original image plane a two-layer 3-D Hopfield network with edge detection to be implemented on its third dimension. With the extended 3-D architecture, the network is capable of incorporating a pixel's contextual information into a pixel-labeling procedure. As a result, the effect of tiny details or noises will be effectively removed by the CHEFNN and the drawback of disconnected fractions can be overcome. Furthermore, by making use of the competitive learning rule to update the neuron states, the problem of satisfying strong constraints can be alleviated and results in a fast convergence. Our experimental results show that the CHEFNN can obtain more appropriate, more continued edge points than the Laplacian-based, Marr-Hildreth, Canny, and wavelet-based methods.",
author = "Chang, {Chuan Yu} and Pau-Choo Chung",
year = "2000",
month = "3",
day = "1",
doi = "10.1117/1.602416",
language = "English",
volume = "39",
pages = "695--703",
journal = "Optical Engineering",
issn = "0091-3286",
publisher = "SPIE",
number = "3",

}

Two-layer competitive based Hopfield neural network for medical image edge detection. / Chang, Chuan Yu; Chung, Pau-Choo.

In: Optical Engineering, Vol. 39, No. 3, 01.03.2000, p. 695-703.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Two-layer competitive based Hopfield neural network for medical image edge detection

AU - Chang, Chuan Yu

AU - Chung, Pau-Choo

PY - 2000/3/1

Y1 - 2000/3/1

N2 - In medical applications, the detection and outlining of boundaries of organs and tumors in computed tomography (CT) and magnetic resonance imaging (MRI) images are prerequisite. A two-layer Hopfield neural network called the competitive Hopfield edge-finding neural network (CHEFNN) is presented for finding the edges of CT and MRI images. Different from conventional 2-D Hopfield neural networks, the CHEFNN extends the one-layer 2-D Hopfield network at the original image plane a two-layer 3-D Hopfield network with edge detection to be implemented on its third dimension. With the extended 3-D architecture, the network is capable of incorporating a pixel's contextual information into a pixel-labeling procedure. As a result, the effect of tiny details or noises will be effectively removed by the CHEFNN and the drawback of disconnected fractions can be overcome. Furthermore, by making use of the competitive learning rule to update the neuron states, the problem of satisfying strong constraints can be alleviated and results in a fast convergence. Our experimental results show that the CHEFNN can obtain more appropriate, more continued edge points than the Laplacian-based, Marr-Hildreth, Canny, and wavelet-based methods.

AB - In medical applications, the detection and outlining of boundaries of organs and tumors in computed tomography (CT) and magnetic resonance imaging (MRI) images are prerequisite. A two-layer Hopfield neural network called the competitive Hopfield edge-finding neural network (CHEFNN) is presented for finding the edges of CT and MRI images. Different from conventional 2-D Hopfield neural networks, the CHEFNN extends the one-layer 2-D Hopfield network at the original image plane a two-layer 3-D Hopfield network with edge detection to be implemented on its third dimension. With the extended 3-D architecture, the network is capable of incorporating a pixel's contextual information into a pixel-labeling procedure. As a result, the effect of tiny details or noises will be effectively removed by the CHEFNN and the drawback of disconnected fractions can be overcome. Furthermore, by making use of the competitive learning rule to update the neuron states, the problem of satisfying strong constraints can be alleviated and results in a fast convergence. Our experimental results show that the CHEFNN can obtain more appropriate, more continued edge points than the Laplacian-based, Marr-Hildreth, Canny, and wavelet-based methods.

UR - http://www.scopus.com/inward/record.url?scp=0033734154&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0033734154&partnerID=8YFLogxK

U2 - 10.1117/1.602416

DO - 10.1117/1.602416

M3 - Article

VL - 39

SP - 695

EP - 703

JO - Optical Engineering

JF - Optical Engineering

SN - 0091-3286

IS - 3

ER -