Comparison of Several ANN Architectures on the Thyroid Diseases Grades Diagnosis

Author

Rouhani, Modjtaba ; Mansouri, Kamran

Author_Institution

Islamic Azad Univ., Gonabad, Iran

fYear

2009

fDate

17-20 April 2009

Firstpage

526

Lastpage

528

Abstract

Nowadays, with advancement of technology and science and expansion of computer usage in high-tech calculations, especially in the field of medicine, intelligence systems and in particular ANN are becoming of significant importance in automatic diagnosis and prognoses of different diseases. In this article, we have used several ANN architectures (namely RBF, PNN, LVQ) and SVMs, diagnosing thyroid diseases. As the degree of disease development is a critical parameter in medical treatment, we design those networks to classify the grade of diseases, too. The performance of each of them has studied and the best method is selected for each of classification tasks. The overall accuracy of diagnosis system is near 99%.

Keywords

biological organs; diseases; medical diagnostic computing; patient diagnosis; probability; radial basis function networks; regression analysis; support vector machines; LVQ network; automatic diagnosis; disease grades; generalized regression neural network; hepatitis; medical treatment; probabilistic neural networks; radial basis functions; support vector machines; thyroid diseases; Artificial neural networks; Computer architecture; Diseases; Neural networks; Neurons; Pathology; Springs; Support vector machine classification; Support vector machines; Testing; GRNN; LVQ; PNN; RBF; SVM; Thyroids diseases;

fLanguage

English

Publisher

ieee

Conference_Titel

Computer Science and Information Technology - Spring Conference, 2009. IACSITSC '09. International Association of

Conference_Location

Singapore

Print_ISBN

978-0-7695-3653-8

Type

conf

DOI

10.1109/IACSIT-SC.2009.24

Filename

5169408