Computational Science on Medicine-Modelling, Simulation and Application


2D, 3D, and Full-field Breast Ultrasound Computer-aided Diagnosis
Prof. Ruey-Feng Chang ( Department of Computer Science and Information Engineering , National Taiwan University )

2008 - 04 - 25 (Fri.)
13:30 - 14:30
308, Mathematics Research Center Building (ori. New Math. Bldg.)

In our previous 2D CAD studies, the texture and speckle information are used to diagnose tumors and these features are classified by the neural network, support vector machine, data mining, and retrieval techniques. Recently, we dealt with 3D CAD techniques including 3D texture analysis, spiculation detection, and microcalcification identification. From these studies, using 3D techniques for CAD represents a potentially significant advantage over 2D. Vessels provide blood supply for tumor expansion and also increase the opportunity for tumor cells to enter the blood or lymph circulation. In contrast to normal vessels, tumor vasculature is highly disorganized, tortuous and dilated, with uneven diameter and excessive branching. We propose a method extracting morphological features from 3-D breast power Doppler ultrasound and a CAD system adopting these features to classify tumor. Ultrasound elastography is a newly developed technique of imaging the tissue elasticity and it has been used clinically to examine a variety of breast lesions in patients. In general, the breast cancer tissue is harder than the adjacent normal breast tissue. Recently, we have developed a CAD for color elastography from Hitachi EUB-8500. Recently, several companies such as Aloka in Japan and U-systems in USA have developed new whole breast ultrasound screening machines. We have developed a computer-aided detection system for these fully automatic screening systems. By using the full-view images, our fully automatic CAD system for whole breast screening could be used to find suspicious frames and locate the approximate positions of tumors in the US images. By the way, an automatic rib detection will be introduced in this talk.

Time:March 7 - June 27, 2008
Organizer:Tony Wen-Hann Sheu ( Department of Engineering Science and Ocean Engineering, National Taiwan University )

Available Talk List