TY - JOUR
T1 - Bioinformatics image based decision support system for bone cancer detection
AU - Vallaboju, Sravanthi
AU - Prasad, P. W. C.
AU - Alsadoon, Abeer
AU - Paul, Manoranjan
AU - Elchouemi, Amr
N1 - Includes bibliographical references.
PY - 2018/7
Y1 - 2018/7
N2 - Bone cancer which may occur inside or on the bone can be life threatening similar to the other types of cancer. The aim of this paper is to improve the accuracy of the detection process. Currently, the detection process is carried out utilising data mining techniques and image processing methods as part of a medical image analysis process, using a non-automated framework which includes image acquisition, image filtering, image segmentation, the area of interest (intensity of the background or the segmented slices) and classification methods to evaluate the decision. Although these methods are effective to some extent, the existing methods have some limitations through false detection values, an increase in the processing time and accuracy. The result indicates that by using eigenvalues and eigenvectors, the processing time can be decreased by implementing normalization, while improving detection accuracy. This paper investigates the viability of using texture based magnetic resonance imaging (MRI) to locate different clusters and classify areas for determining bone cancer. This segmentation and classification processes are carried out by using eigenvalues and eigenvectors.
AB - Bone cancer which may occur inside or on the bone can be life threatening similar to the other types of cancer. The aim of this paper is to improve the accuracy of the detection process. Currently, the detection process is carried out utilising data mining techniques and image processing methods as part of a medical image analysis process, using a non-automated framework which includes image acquisition, image filtering, image segmentation, the area of interest (intensity of the background or the segmented slices) and classification methods to evaluate the decision. Although these methods are effective to some extent, the existing methods have some limitations through false detection values, an increase in the processing time and accuracy. The result indicates that by using eigenvalues and eigenvectors, the processing time can be decreased by implementing normalization, while improving detection accuracy. This paper investigates the viability of using texture based magnetic resonance imaging (MRI) to locate different clusters and classify areas for determining bone cancer. This segmentation and classification processes are carried out by using eigenvalues and eigenvectors.
KW - Eigenvectors
KW - Affinity matrix
KW - Clusters
KW - Feature extraction
KW - Normalized eigenvectors
KW - Bone tumour
UR - http://www.jcomputers.us/
U2 - 10.17706/jcp.13.7.771-783
DO - 10.17706/jcp.13.7.771-783
M3 - Article
SN - 1796-203X
VL - 13
SP - 771
EP - 783
JO - Journal of Computers
JF - Journal of Computers
IS - 7
ER -