Speaker Abstract: S-42
Computer Aided Diagnosis (CAD): The Current State and Future for this Technology
Nicholas Petrick, Ph.D., Center for Devices and Radiological Health/FDA
Computer techniques that assist clinicians in making medical decisions are quickly being disseminated into clinical practice in many hospitals and medical centers. These computer tools are collectively known as computer-aided diagnosis (CAD) devices. One of the important areas where CAD devices are receiving FDA approval is as adjuncts in reading medical images. These devices aid a radiologist in either detecting or characterizing a disease state based on image features. This talk will present an overview of some current CAD technologies by providing an introduction to various image processing, image features and classification strategies utilized by these computer schemes. In particular, actual CAD algorithms developed for detecting and characterizing breast lesions on planar mammographic images, as well as algorithms developed to detect lung nodules or colonic polyps on computer tomography (CT) volume sets will be described. I will also discuss the performance of these CAD algorithms and prognosticate a bit about how CAD technology may continue to evolve in the near future.
2004 FDA Science Forum | FDA Chapter, Sigma Xi | CFSAN | FDA
Last updated on 2004-MAY-28 by frf