Skip Ribbon Commands Skip to main content
Help (new window)
University of Balamand > Academics > Research > Seminars > Bilal Ghazal

Classification of Contrast Ultrasound Images Using Gaussian Mixture Models

Tuesday, December 4, 2007 from 12:30PM to 2:00 PM at the Jacobo Auditorium

SUBMITTED BY: Mr. Bilal Ghazal
University of Balamand
bilalghazal00@hotmail.com

ABSTRACT: The use of ultrasound is increasing in medical applications, but ultrasound images are characterized by a noise called speckle. The use of contrast agents enhances the echo from blood with respect to the surrounding tissues, and the discrimination between contrast agents and tissues allows a better visualization of the contrast ultrasound images, yields an easy observation of the wall of blood vessels, and permits better measurement of blood perfusion. Nevertheless, images are still not quite clear enough to be directly adopted in medical diagnosis. This major drawback may be tackled by newly implemented image-treatment approaches that enhance the contrast echo and present the capability for classification.

We have applied a new technique based on the autoregressive model coupled to the Gaussian mixture model GMM. The GMM has modeled both agent and tissue behaviors using Gaussian functions. We have then processed the resultant image by a classification method based on a fixed window size in order to obtain a satisfying differentiation of the ultrasound image into two classes. Our study has also adopted the Agent to Tissue Ratio (ATR) factor and the Fisher criterion to study the performance of this approach.
facebook    LinkedIn    youTube    Social Media
University of Balamand,
Balamand Al Kurah,
Lebanon

Tel:  +961-6-930250
Fax: +961-6-930278