Japanese journal of medical electronics and biological engineering
Online ISSN : 2185-5498
Print ISSN : 0021-3292
ISSN-L : 0021-3292
Tissue Characterization of Gastric Cancer Echo Images by Using Texture Analysis
Keiichi SAITOHiroshi HASHIMOTOHiroshi OBATAAkihiko UCHIYAMA
Author information
JOURNAL FREE ACCESS

1991 Volume 29 Issue 2 Pages 109-114

Details
Abstract
We studied the feasibility of quantitative texture analysis on two dimensional echoic images in 9 cases of resected gastric cancers in order to differentiate medullary type from intermediate type. 4 cases of medullary type and 5 cases of intermediate type were studied. The echo system we used had a linear arrey probe, and the central frequency of this was 7.5MHz. On examination, the parameters of the system (gain control, etc.) were usually on the same level, and we scanned at about 2.5cm distance from the regions. Regions of interest were set on echo images of cancer tissue. The texture analysis we used was the grey level difference method. As a result, the grey level difference distribution of medullary type was seen from that of interemedite type and the error rate of discriminant analysis was found to be 10%. We would expect the usefulness of this method to be evident in endoscopic ultrasonography.
Content from these authors
© Japanese Society for Medical and Biological Engineering
Previous article Next article
feedback
Top