BREAST DIAGNOSIS: QUANTITATIVE IMAGING BY ULTRASOUND

Project: Research project

Project Details

Description

The goal of this proposal is to develop and investigate methods and instruments for obtaining reliable quantitative ultrasound images of several acoustic paramenters within two-dimensional coronal cross sections of the breasts of women. The instrument and associated programs will synergistically combine quantitative multimodal images to obtain diagnoses of breast disease. Ultrasound computer-assisted transmission tomography will be utilized to obtain quantitative images of acoustic speed, acoustic attenuation, acoustic absorption and scatter and reflection where appropriate. Both analytic and iterative methods of correcting aberrations in the images caused by inhomogeneous refractive index will be investigated. Compound B-scan images will also be obtained in the same coronal planes. Our clinical scanner will be modified to obtain data necessary for high fidelity images derived using newly developed inverse scattering methods. Patients scheduled for surgical biopsy will be scanned and the resulting images compared to the histology results from biopsy and/or surgical mastectomy. Comparisons between surgical specimens and acoustic images will be made using methods of automated computer-assisted pattern classification and recognition combined with histological methods of tissue preparation specifically designed to aid in comparing ultrasound images of the coronal plane of the intact breast with similar planes through the excised breast. This parallel investigation of the clinical instrument, inverse scattering methods, histologic tissue analysis techniques, and pattern recognition programs should result in complete evaluation of the prototype clinical scanner and may result in an instruemnt design capable for use in the clinical environment.
StatusFinished
Effective start/end date1/1/901/1/90

Funding

  • National Cancer Institute

Fingerprint

Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.