International Journal of Biometrics and Bioinformatics (IJBB)

---

INTERNATIONAL JOURNAL OF BIOMETRICS
AND BIOINFORMATICS (IJBB)







VOLUME 6, ISSUE 1, 2012

EDITED BY
DR. NABEEL TAHIR








ISSN (Online): 1985-2347
International Journal of Biometrics and Bioinformatics (IJBB) is published both in traditional paper
form and in Internet. This journal is published at the website http://www.cscjournals.org,
maintained by Computer Science Journals (CSC Journals), Malaysia.


IJBB Journal is a part of CSC Publishers
Computer Science Journals
http://www.cscjournals.org

---

INTERNATIONAL JOURNAL OF BIOMETRICS AND
BIOINFORMATICS (IJBB)

Book: Volume 6, Issue 1, February 2012
Publishing Date: 21-02-2012
ISSN (Online): 1985-2347

This work is subjected to copyright. All rights are reserved whether the whole or
part of the material is concerned, specifically the rights of translation, reprinting,
re-use of illusions, recitation, broadcasting, reproduction on microfilms or in any
other way, and storage in data banks. Duplication of this publication of parts
thereof is permitted only under the provision of the copyright law 1965, in its
current version, and permission of use must always be obtained from CSC
Publishers.



IJBB Journal is a part of CSC Publishers
http://www.cscjournals.org

(c) IJBB Journal
Published in Malaysia

Typesetting: Camera-ready by author, data conversation by CSC Publishing Services - CSC Journals,
Malaysia



CSC Publishers, 2012

---

EDITORIAL PREFACE

This is the first issue of volume six of International Journal of Biometric and Bioinformatics (IJBB).
The Journal is published bi-monthly, with papers being peer reviewed to high international
standards. The International Journal of Biometric and Bioinformatics is not limited to a specific
aspect of Biology but it is devoted to the publication of high quality papers on all division of Bio in
general. IJBB intends to disseminate knowledge in the various disciplines of the Biometric field
from theoretical, practical and analytical research to physical implications and theoretical or
quantitative discussion intended for academic and industrial progress. In order to position IJBB as
one of the good journal on Bio-sciences, a group of highly valuable scholars are serving on the
editorial board. The International Editorial Board ensures that significant developments in
Biometrics from around the world are reflected in the Journal. Some important topics covers by
journal are Bio-grid, biomedical image processing (fusion), Computational structural biology,
Molecular sequence analysis, Genetic algorithms etc.

The initial efforts helped to shape the editorial policy and to sharpen the focus of the journal.
Starting with volume 6, 2012, IJBB appears in more focused issues. Besides normal publications,
IJBB intend to organized special issues on more focused topics. Each special issue will have a
designated editor (editors) - either member of the editorial board or another recognized specialist
in the respective field.

The coverage of the journal includes all new theoretical and experimental findings in the fields of
Biometrics which enhance the knowledge of scientist, industrials, researchers and all those
persons who are coupled with Bioscience field. IJBB objective is to publish articles that are not
only technically proficient but also contains information and ideas of fresh interest for International
readership. IJBB aims to handle submissions courteously and promptly. IJBB objectives are to
promote and extend the use of all methods in the principal disciplines of Bioscience.

IJBB editors understand that how much it is important for authors and researchers to have their
work published with a minimum delay after submission of their papers. They also strongly believe
that the direct communication between the editors and authors are important for the welfare,
quality and wellbeing of the Journal and its readers. Therefore, all activities from paper
submission to paper publication are controlled through electronic systems that include electronic
submission, editorial panel and review system that ensures rapid decision with least delays in the
publication processes.

To build its international reputation, we are disseminating the publication information through
Google Books, Google Scholar, Directory of Open Access Journals (DOAJ), Open J Gate,
ScientificCommons, Docstoc and many more. Our International Editors are working on
establishing ISI listing and a good impact factor for IJBB. We would like to remind you that the
success of our journal depends directly on the number of quality articles submitted for review.
Accordingly, we would like to request your participation by submitting quality manuscripts for
review and encouraging your colleagues to submit quality manuscripts for review. One of the
great benefits we can provide to our prospective authors is the mentoring nature of our review
process. IJBB provides authors with high quality, helpful reviews that are shaped to assist authors
in improving their manuscripts.


Editorial Board Members
International Journal of Biometric and Bioinformatics (IJBB)

---

EDITORIAL BOARD

EDITOR-in-CHIEF (EiC)

Professor Joao Manuel R. S. Tavares
University of Porto (Portugal)


ASSOCIATE EDITORS (AEiCs)


Assistant Professor. Yongjie Jessica Zhang
Mellon University
United States of America

Professor. Jimmy Thomas Efird
University of North Carolina
United States of America

Professor. H. Fai Poon
Sigma-Aldrich Inc
United States of America

Professor. Fadiel Ahmed
Tennessee State University
United States of America

Professor. Yu Xue
Huazhong University of Science and Technology
China

Associate Professor Chang-Tsun Li
University of Warwick
United Kingdom

Professor. Calvin Yu-Chian Chen
China Medical university
Taiwan



EDITORIAL BOARD MEMBERS (EBMs)


Assistant Professor. M. Emre Celebi
Louisiana State University
United States of America


Dr. Ganesan Pugalenthi
Genome Institute of Singapore
Singapore

Dr. Vijayaraj Nagarajan
National Institutes of Health
United States of America

---

Dr. Wichian Sittiprapaporn
Mahasarakham University
Thailand

Dr. Paola Lecca
University of Trento
Italy

Associate Professor. Renato Natal Jorge
University of Porto
Portugal

Assistant Professor. Daniela Iacoviello
Sapienza University of Rome
Italy

Professor. Christos E. Constantinou
Stanford University School of Medicine
United States of America

Professor. Fiorella SGALLARI
University of Bologna
Italy

Professor. George Perry
University of Texas at San Antonio
United States of America

Assistant Professor. Giuseppe Placidi
Universita dell'Aquila
Italy

Assistant Professor. Sae Hwang
University of Illinois
United States of America

Associate Professor Quan Wen
University of Electronic Science and Technology
China

Dr. Paula Moreira
University of Coimbra
Portugal

Dr. Riadh Hammami
Laval University
Canada

Dr Antonio Marco
University of Manchester
United Kingdom

Dr Peng Jiang
University of Iowa
United States of America

---

Dr Shunzhou Yu
General Motors Global R&D Center
United States of America

---

TABLE OF CONTENTS




Volume 6, Issue 1, February 2012



Pages
1- 10
MR Image Segmentation of Patients' Brain Using Disease Specific a Priori Knowledge
Hassan Tavakkoli, Ali Sadeqi

11 - 23
3-D Face Recognition Using Improved 3D Mixed Transform

Hamid M. Hasan, Waleed A. AL.Jouhar , Majid A. Alwan

24 - 37
A Neural Network Based Diagnostic System for Classification of Industrial Carrying Jobs

With Respect of Low and High Musculoskeletal Injury Risk

Rohit Sharma, Ranjit Singh








International Journal of Biometrics and Bioinformatics (IJBB), Volume (6), Issue (1) : 2012

---

Hassan Tavakkoli & Ali Sadeqi
MR Image Segmentation of Patients' Brain Using Disease
Specific a Priori Knowledge


Hassan Tavakkoli





tavakoli@ibb.ut.ac.ir
Applied Neuroscience Research Center
Baqiyatallah University of Medical Sciences
Tehran, Iran

Ali Sadeqi





sadeqi_a@ymail.com
Applied Neuroscience Research Center
Baqiyatallah University of Medical Sciences
Tehran, Iran

Abstract

Segmentation of high quality brain MR images using a priori knowledge about brain structures
enables a more accurate and comprehensive interpretation. Benefits of applying a priori
knowledge about the brain structures may also be employed for image segmentation of specific
brain and neural patients. Such procedure may be performed to determine the disease stage or
monitor its gradual progression over time. However segmenting brain images of patients using
general a priori knowledge which corresponds to healthy subjects would result in inaccurate and
unreliable interpretation in the regions which are affected by the disease. In this paper, a
technique is proposed for extracting a priori knowledge about structural distribution of different
brain tissues affected by a specific disease to be applied for accurate segmentation of the
patients' brain images. For this purpose, extracted a priori knowledge is gradually represented as
disease specific probability maps throughout an iterative process, and then is utilized in a
statistical approach for segmentation of new patients' images. Experiments conducted on a large
set of images acquired from patients with a similar neurodegenerative disease implied success of
the proposed technique for representing meaningful a priori knowledge as disease specific
probability maps. Promising results obtained also indicated an accurate segmentation of brain
MR images of the new patients using the represented a priori knowledge, into three tissue
classes of gray matter, white matter, and cerebrospinal fluid. This enables an accurate estimation
of tissues' thickness and volumes and can be counted as a substantial forward step for more
reliable monitoring and interpretation of progression in specific brain and neural diseases.

Keywords: Brain MR Images, Segmentation, Tissue Classification, Image Registration, A Priori
Knowledge, Brain Tissue Probability Maps, Neurodegenerative Disease.


1. INTRODUCTION
Magnetic Resonance Imaging (MRI) [1] of brain is well known as a powerful technique for
diagnosis which is widely used by many clinicians to detect structural abnormalities that cause
neurological disorders. Until a few years ago, most of the neurologist just employed their personal
expertise to inspect a number of cross sections of the patient's brain MRI to diagnosis the
disorder or monitor the effects of a specific therapy. Swift progresses in medical image
processing [2] have introduced automatic and semi-automatic methods to enable more precise
and reliable diagnosis and treatments. Image segmentation [3] is an example for such methods
which is widely used in many biomedical applications. A popular application of image
segmentation in neurology is partitioning a specific brain structure into three major tissues of gray
matter, white matter, and cerebrospinal fluid. Such a segmentation process makes it possible to
estimate distributions of different tissue types in the brain in addition to calculate their volumes
accurately. This information is very beneficial to make an accurate diagnosis or monitor the
treatment progress precisely.
International Journal of Biometrics and Bioinformatics (IJBB), Volume (6) : Issue (1) : 2012
1

---

Hassan Tavakkoli & Ali Sadeqi
Either manual or automatic Interpretation of brain's high quality images can benefit from utilizing
atlases as well as a priori knowledge represented as comprehensive probabilistic maps [4, 5].
Brain atlases are three dimensional (3D) images produced by averaging on a large number of
brain images of healthy subjects. Probabilistic maps which are frequently represented as 3D
images are structures which determine the probabilities by which each voxel of a typical brain
image corresponds to any of the major tissue type. Developing advantageous atlases and/or a
priori knowledge in the forms of inclusive probabilistic maps requires collecting and processing a
large set of brain images [6] in addition to utilizing suitable strategies for image registration and
warping [7].

Great benefits of employing atlases as well as a priori knowledge hidden in probabilistic maps
can be extended for analyzing brain images of specific patients with a similar neurological
disorder such as neurodegenerative diseases. Neurodegenerative diseases including mild
cognitive impairment, Alzheimer [8], and major depression [9, 10], are disorders caused by
progressive deterioration of neurons that eventually leads to their death. Clinical symptoms of
neurodegenerative diseases include disabilities and dysfunctions resulting in dementia (memory
dysfunctions) and/or ataxia (movement disabilities) [11]. Progressive death of neurons in
neurodegenerative disease results in atrophy of the brain structures. Such gradual effects can be
monitored or even quantified using MRI [12]. The quantification is performed via MR image
segmentation and estimating distribution of different brain tissues. Segmentation is usually
performed through parametric methods [13] and its precision can be improved using a priori
knowledge of probabilistic maps registered on the image of patient's brain [14, 15].

Analyzing images of patients' brain using general a priori knowledge extracted from brain images
of healthy subjects would result in inaccurate and unreliable interpretation in the regions which
are affected by the disease [16]. For example in image segmentation of patients' brain, this may
result in misclassification of brain tissues and consequently inaccurate measurement of their
volumes. As such presentencing a systematic method for development of a priori knowledge
about structural distribution of brain tissues which is affected by a specific disease and maps is a
paramount necessity for more reliable studies on progression of the specific neurologic disorders,
and monitoring their gradual affects on different brain tissues.

In this paper a novel technique is proposed for systematically representation of a priori knowledge
about effects of a specific neurological disease on different tissue structures of the brain. The
initial application proposed for the developed knowledge in this paper is segmentation of patients'
brain images into its three major tissue types. Such customized segmentation process can also
accurately calculate the volumes of each tissue type accurately. Experiments conducted on a
large set of images acquired from patients with a similar neurodegenerative disease implied
success of the proposed technique for representing meaningful a priori knowledge as disease
specific probability maps. Promising results obtained also indicated an accurate segmentation of
brain MR images of the new patients using the represented a priori knowledge, into three tissue
classes of gray matter, white matter, and cerebrospinal fluid. This can pave the way for a more
precise disease diagnosis, mentoring its gradual growth, and following the treatment progress.

This paper is organized as follows. The required preliminaries are presented in the next section.
Section 3 introduces the iterative method used in this paper for development of the disease
specific a priori knowledge and the segmentation of the brain image of the patients. The
experiments conducted and the results obtained are presented in section 4. Finally section 5
discusses the obtained results and concludes this paper.

2. PRELIMINARIES

2.1
Medical Image Registration
In image registration, a one-to-one mapping or transformation is determined between the
coordinates of one image space to those in the other. In rigid registration the transformation is
International Journal of Biometrics and Bioinformatics (IJBB), Volume (6) : Issue (1) : 2012
2

---