Ali S. Saad's Home page

 First Project:

 Research Title: Wavelets multi-resolution for semi-automatic segmentation and filtering of ultrasound images for better diagnoses

ًprincipale investigatore: Ali saad College of applied medical sciences/biomedical technology Assistant professor

co-investigatore: Ali Al-Mejrad College of applied medical sciences/biomedical technology Assistant professor

  Research Problem:

Ultrasound images are very noisy, they are affected by speckle which originate from waves arrives at a sensor point out of phase the interference among these waves causes the granular aspect termed speckle. The presence of speckle in the image prevents the physician from doing an accurate diagnosis for patients.

Research Significance:

 In order to help physicians for better diagnoses a digital image processing method is proposed here to first improve the visual quality of the image by reducing speckle and then give an automatic segmentation of the image using wavelet techniques at different level of resolution which certainly gives a more accurate estimation of boundaries of the object under investigation., Therefore a better estimation of the size of the object. Using this quantitative method of estimation physician can easily estimated the increase or decrease of tumors for better diagnoses, or the growing of a fetus.

Research Objectives:

Is to develop software that can realize the segmentation of ultrasound images a different level of resolution using wavelet techniques for speckle reduction and unsupervised fuzzy clustering for segmentation.

 Research Methodology:

Wavelets decomposition is used to transform the original image into several components: one low-resolution component called “approximation”, and the other components called “details”. At each level after using a soft threshold wavelet filtering a pre-segmentation by fuzzy clustering is applied then a refinement of the segmentation is achieved by statistical maximum a posterior method.