Diagnostic accuracy of computerized B‐scan texture analysis and conventional ultrasonography in diffuse parenchymal and malignant liver disease

Abstract

The use of a diagnostic system for ultrasonic liver tissue characterization based on computerized B-mode image analysis is clinically tested and compared with the results of conventional realtime and static grey scale liver ultrasound as independently assessed by three experienced observers. The diagnostic classes, normal, diffuse parenchymal and malignant disease, are clearly differentiated by computerized image analysis which is superior to subjective evaluation of liver echograms. Computerized analysis also renders a reliable and clinically useful diagnostic subclassification of diffuse parenchymal disease into echopattern changes prevalent in chronic hepatitis, cirrhosis/fibrosis, fatty infiltration and a mixed state of cirrhosis/fibrosis with fatty infiltration which cannot be achieved by conventional liver ultrasound.

Real-time ultrasound (RTU) imaging methods for quality control of meats

“…On the whole, results obtained with mathematical procedures were superior to the subjective RTU image evaluation, even when this task was conducted by an experienced individual (Raeth et al ., 1985;Couto et al ., 2011). Since the first attempt was made to predict IMF using RTU image analysis, significant advances have been possible as a result of developments in equipment and software.…”

Quantitative analysis of ultrasound B-mode images of carotid atherosclerotic plaque: correlation with visual classification and histological examination

“…This basis for calculating second-order features has previously been used in attempts to classify tissue from ultrasound images [4], [5], [24], [29], [36], [39], [40]. The GLCM is a measure of the co-occurrences in the image and is calculated for a given displacement vector, .…”

“…Quantitative image analysis has previously been applied to echocardiographic images, [7], [39] ultrasound images of the breast [14], [15] and to ultrasound images of liver [4], [24], [29], [36], [40]. In the latter investigations, a large number of features has been calculated from regions located inside liver tissue on B-mode images.…”

Analysis of Follicle Wall of Normal and Polycystic Ovaries

“…Methods based on textural feature analysis are used in identifying diffuse diseases of the liver [8]. Raeth [9] showed the progress of atherosclerosis using the texture method. Burr [10] used texture method to detect breast cancer.…”