J Korean Soc Med Ultrasound.  1999 Jun;18(2):81-86.

Differentiation of Diffuse Liver Disease with Computer-Aided Tissue Echo Quantification

• Cha JH
• Choi BI
• Yun EJ
• Ko YH
• Song CS
• Kim SH
• Han JK
• Kim TK
• Lee DH
• Kim JH

Abstract

PURPOSE: The purpose of this study was to evaluate the efficacy of computer-aided tissue echo quantification technique in the differentiation of diffuse liver diseases.
MATERIALS and METHODS
Sixty-five patients with chronic liver disease including chronic hepatitis in 21 patients, fatty liver in 11, and liver cirrhosis in 33, and 55 normal volunteers were included in this study. The sonographic images by the Sono-PACS (MARO, Seoul) was recalled and the analysis was done for the hepatic parenchymal coarseness by the program using Visual C++. Difference histogram variation (DHV), edge density (ED) & inertia of cooccurrence matrix (ICM) were used as the coarseness index.
RESULTS
These three indexes were statistically significant (p<0.05) in the differentiation of these four groups. The accuracy of the differentiation between any two of four groups was 83.0%. The accuracy of the differentiation of these four groups was 70.8% at the same time.
CONCLUSION
The computer-aided tissue echo quantification technique is a complementary study for the differentiation of diffuse liver diseases.

MeSH Terms

Fatty Liver
Healthy Volunteers
Hepatitis, Chronic
Humans
Liver Cirrhosis
Liver Diseases*
Liver*
Ultrasonography