SVM-based ultrasonic medicine image diagnosis

Jun Zhang; Xiaomao Liu; Jianguo Liu; Fuyuan Peng; Jinwen Tian; Ying Wang; Wenjun Zhang; Mingxing Xie

doi:10.1117/12.440251

18 September 2001 SVM-based ultrasonic medicine image diagnosis

Jun Zhang, Xiaomao Liu, Jianguo Liu, Fuyuan Peng, Jinwen Tian, Ying Wang, Wenjun Zhang, Mingxing Xie

Author Affiliations +

Proceedings Volume 4549, Medical Image Acquisition and Processing; (2001) https://doi.org/10.1117/12.440251
Event: Multispectral Image Processing and Pattern Recognition, 2001, Wuhan, China

Abstract

The technique of support vector machines (SVM's) has been used as a new method for solving classification, regression, time series prediction and function estimation problems with many successful applications. In this paper, we use SVM to solve the problem of classification for ultrasonic medicine image. We use statistical characteristics of the interesting part in an ultrasonic image to aid a doctor to give a correct diagnosis. The procedure is: first, to extract a number of small sampling regions in the interesting part; second, to calculate a series of moments about those sampling regions; third, to decide whether the interesting part of the organ is normal or abnormal according to the analyses of the series of moments based on SVM. SVM's structural risk minimization principle is the guarantee that the diagnosis has the minimum mistake probability. The diagnosis based on SVM is optimal from the viewpoint of structural risk minimization principle. It is hoped that the results presented here will be helpful to the diagnosis based on ultrasonic medicine image.

Citation Download Citation

Jun Zhang, Xiaomao Liu, Jianguo Liu, Fuyuan Peng, Jinwen Tian, Ying Wang, Wenjun Zhang, and Mingxing Xie "SVM-based ultrasonic medicine image diagnosis", Proc. SPIE 4549, Medical Image Acquisition and Processing, (18 September 2001); https://doi.org/10.1117/12.440251

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

;

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE