Preview

Informatics

Advanced search

A RULE-BASED METHOD FOR TUMOR RECOGNITION IN LIVER ULTRASOUND IMAGES

Abstract

A method is suggested for detecting malignant tumors in liver ultrasonic images. The method is based on decision rules which utilize quantitative features of liver image texture such as the degree of texture anisotropy, distribution of intensity gradients, the area of connected components with certain image properties and some other. On a set of 1280 regions of interests sampled from 26 test images it is demonstrated that the method provides an acceptable level of tumor recognition accuracy in 96,8 % of cases.

For citations:


Kovalev V.А. A RULE-BASED METHOD FOR TUMOR RECOGNITION IN LIVER ULTRASOUND IMAGES. Informatics. 2016;(2):59-70. (In Russ.)

Views: 895


Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.


ISSN 1816-0301 (Print)
ISSN 2617-6963 (Online)