WARF: P04336US
Time-Domain Inverse Scattering Techniques for Use in Microwave Imaging

INVENTORS

•

David Winters, Barry Van Veen, Susan Hagness

Microwave-based imaging methods, which exploit the contrast in dielectric properties between normal and malignant tissue, offer a promising alternative to ultrasound and X-ray techniques for imaging breast tissue. Malignant tissue is most accurately detected and localized when the processing algorithms include the average properties of the actual, heterogeneous breast tissue, which vary from patient to patient; however, current techniques for solving the equations to estimate these properties suffer from nonlinearity and ill-posedness of the full inverse scattering problem. UW-Madison researchers have developed an improved method for estimating the average dielectric properties of breast tissue. The method, an extension of a time-domain inverse scattering technique, greatly reduces the number of unknown parameters in the inverse scattering problem. This allows the average properties of the breast tissue to be more readily estimated for each patient, increasing the accuracy of malignant tissue detection and localization. Simulation studies have shown that this method results in significantly improved detection of malignant tissue.

• Detecting malignant breast tissue

• Greatly reduces the number of unknown parameters, reducing computational requirements and decreasing the level of ill-conditioning inherent in the inverse problem
• Much less sensitive to noise than other approaches
• Can obtain a solution where conventional inverse scattering methods fail
• Applicable to other subsurface microwave imaging problems, including land mine detection

For current licensing status, please contact our team at licensing@warf.org or phone 608.262.4924. (Clicking this link will open a contact form in a popup window. If you have problems viewing the form, try disabling your popup blocker software.)