Accession Number : ADA336818

Title :   Computer-Aided Diagnosis and Feature-Guided Data Reduction Systems in Mammography

Descriptive Note : Annual rept 23 Sep 96-22 Sep 97

Corporate Author : MICHIGAN UNIV ANN ARBOR

Personal Author(s) : Chan, Heang-Ping

PDF Url : ADA336818

Report Date : OCT 1997

Pagination or Media Count : 21

Abstract : The goals of this project are: (1) Implement and evaluate a computer-aided diagnosis (CAD) system to assist radiologists in mammographic interpretation. (2) Develop and evaluate a feature-guided data compression technique to facilitate implementation of digital mammography. For the first sub-project, we plan to implement our CAD algorithms for detection and classification of microcalcifications in a high speed workstation, develop user interface for efficient operation of the CAD programs, and perform a pilot clinical trial in three mammographic screening sites. For the second sub-project, we plan to evaluate different data compression techniques. We will select the most efficient loss less technique and the lossy technique that can provide maximum compression ratio without noticeable loss of information for mammography. In the first year of the funding period, we have performed the following studies: (1) development of a graphical user interface (GUI) for the CAD, (2) conversion of CAD software from VMS to UNIX operating system, (3) improvement in the automated algorithms for detection of masses and microcalcifications, (4) study of data compression methods for mammograms, and (5) design methods for data collection and data analysis for the pilot clinical trial. These studies are the necessary steps to accomplish the goals of this project.

Descriptors :   *COMPUTER AIDED DIAGNOSIS, *BREAST CANCER, *MAMMOGRAPHY, ALGORITHMS, DATA PROCESSING, DIGITAL SYSTEMS, INTERFACES, CLINICAL MEDICINE, DATA ACQUISITION, COMPUTER APPLICATIONS, DATA COMPRESSION, COMPRESSION RATIO, RADIOLOGISTS.

Subject Categories : Anatomy and Physiology
      Medicine and Medical Research
      Cybernetics

Distribution Statement : APPROVED FOR PUBLIC RELEASE