ULTRASONIC SPECKLE IS THE MOST IMPORTANT SOURCE OF NOISE DEGRADING CLINICAL ULTRASOUND IMAGES AND IT LIKELY IMPAIRS THE DETECTABILITY OF LOW-CONTRAST TUMORS.

ULTRASONIC SPECKLE IS THE MOST IMPORTANT SOURCE OF NOISE DEGRADING CLINICAL ULTRASOUND IMAGES AND IT LIKELY IMPAIRS THE DETECTABILITY OF LOW-CONTRAST TUMORS. PHASE I AND PHASE II OF THIS RESEARCH EFFORT ARE DEVOTED TO THE EVALUATION OF THE IMPROVEMENT IN TUMOR DETECTABILITY WHICH MAY BE ACHIEVED BY SEVERAL SPECKLE REDUCTION TECHNIQUES IN ULTRASONIC IMAGES, AND DEFINITION OF THOSE WAVEFORM PROCESSING MODALITIES THAT LEAD TO OPTIMUM IMPROVEMENT IN IMAGE QUALITY. IN PHASE I, ULTRASONIC DATA WILL BE ACQUIRED FROM A CONTRAST RESOLUTION PHANTOM AND STORED ON FLOPPY DISK. FREQUENCY DIVERSITY TECHNIQUES, AVERAGING OF DIGITALLY FILTERED SCAN DATA ACQUIRED EMPLOYING BROAD-BAND TRANSDUCERS, WILL BE USED FOR SPECKLE REDUCTION. THE RESULTING SPECKLE REDUCED IMAGES WILL BE EVALUATED FOR THE DETECTABILITY OF LOW-CONTRAST TARGETS. THE RESULTS OF PHASE I WILL BE USED TO GUIDE PHASE II, A MORE INVOLVED STUDY OF OBSERVERS' ABILITY TO DETECT LOW-CONTRAST TARGETS IN SPECKLE-REDUCED IMAGES FROM PHANTOMS AND PATHOLOGIC SPECIMENS. A SPECIALLY DESIGNED DATA ACQUISITION SYSTEM WILL BE USED FOR SPATIAL AND FREQUENCY DIVERSITY AVERAGING. THE FRAMEWORK OF KNOWLEDGE CREATED WILL INDICATE POSSIBLE HARDWARE AND PROCESSING IMPROVEMENTS IN CLINICAL SCANNERS FOR