Ultrasonic Tissue Characterization with Common B-mode Device: GLHW

Abstract

Medical ultrasound is superb in the real-time B-mode, transvaginal scan, 3-4D ultrasound, for the morphological examination and diagnosis. Only the Doppler ultrasound is unique in the functional study of blood flow and fetal movement. Although the advanced ultrasound detected the nature of tissue structure in the brightness, back scatter, texture and so on, the technique needed advance computer and programs, so that most ultrasound user are out of the benefit of tissue characterization, for example, ovarian malignancy is still diagnosed by the B-mode patterns. Maeda searched for stable technique for the B-mode brightness and found the histogram of B-mode brightness, which was easily found in every B-mode machine. However, the mean gray level value changed when the sensitivity was controlled, the contrast of B-mode image changed, and when the subject depth was changed, in the studies with an ultrasound phantom in around 1990. The author then studied the standardization of histogram, and found the histogram base width was stable to the sensitivity control. Then the histogram base width was standardized by dividing with full length of the brightness, and the gray level histogram width (GLHW) value was completed. As the GLHW was not influenced by the sensitivity control, but changed by image contrast, the contrast level was fixed at 2 among multiple levels. The GLHW was stable among various B-mode devices, but a machine was tested using the ultrasound phantom. Now the GLHW could be applied for the tissue characterization with common B-mode machine without special computer or program [1]. The region of interest (ROI) of subject image was 1x1cm in the trial study, but found the ROI size did not influence on the results. Automatic determination was investigated and confirmed no difference to the manual determination. The automation was introduced into the histogram analysis of Aloka B-mode machine, which was expressed by