CLINICAL ULTRASONIC CHARACTERIZATION OF THE PROSTATE

Riverside Research Institute, New York NY

Investigators

Linked publications & trials

Abstract

[unreadable] DESCRIPTION (provided by applicant): The proposed 5-year study continues our on-going project that has demonstrated the feasibility of differentiating cancerous from non-cancerous tissues of the prostate using spectrum analysis of radio-frequency ultrasonic echo signals. Our current studies show a clear superiority of the spectrum analysis methods being developed over conventional ultrasonic imaging for identifying cancerous tissue; for example, they produce ROC-curve areas approaching 0.80 in contrast to matching ROC-curve areas of 0.66 for conventional methods of identifying cancerous tissues non-invasively. The proposed studies will build upon the prior studies with the purposes of improving 1) biopsy guidance, hence enabling early detection and timely treatment of prostate cancers, 2) staging, hence treatment planning and monitoring, and 3) understanding of the mechanisms of ultrasound scattering in prostate tissue, hence establishing a basis for advanced methods of distinguishing cancerous from non-cancerous prostate tissues. These purposes are consistent with the priorities defined in the August 1998 Report of the Prostate Cancer Review Group (PRG) to the Advisory Committee of the Director of the National Cancer Institute. The proposed studies will be undertaken by a consortium consisting of the applicant organization, Riverside Research Institute, in collaboration with the Memorial Sloan-Kettering Cancer Center (MSKCC), as in prior studies, and the University Hospital and Medical Center (UHMC) of the State University of New York (SUNY) at Stony Brook. The addition of collaborators at UHMC will improve the study markedly in terms of the racial diversity of its human-subject populations, its rate of patient recruitment, its surgical-specimen resources, and its medical perspectives and approaches. The proposed studies will continue to develop and apply spectrum analysis of ultrasonic echo signals and will investigate complementary methods such as elastography for the purposes of tissue typing and presenting tissue-typing information in clinically useful images. The efficacy of these presentations will be assessed using standard ROC methods, and the classification techniques underlying them will include non-linear methods of classification such as neural-network tools.