The long-term goal of the research described in this proposal is to develop innovative magnetic resonance (MR) imaging techniques for early detection of breast cancer. MR imaging may provide a useful alternative to conventional X-ray mammography for detecting cancer in women with predominantly fibro-glandular breasts which are not easily penetrated by X-rays. In particular, this research will explore the application of spin locking to MR imaging of the breast. Spin locking depends on the novel tissue contrast parameter T(1p) which can be used to improve contrast of breast lesions with respect to normal breast tissues. Furthermore, T(1p) depends on magnetization transfer, an important relaxation mechanism which may be useful for detecting breast cancer. This proposal describes research designed to investigate the potential of T(1p) and magnetization transfer for distinguishing malignant from benign breast lesions in vivo. The goal will be a non-invasive MR imaging technique for confirmation of breast cancer without the need for costly, uncomfortable and cosmetically-deforming surgical biopsy. Specifically, this research aims to develop and optimize off-resonance spin-locking techniques for improved detection of breast lesions. The techniques will be developed on a clinical MR imaging system modified for spin locking and optimized using T(1p) relaxation measurements obtained from excised breast tissue specimens, both normal tissues as well as benign and malignant breast lesions as confirmed by pathological examination. The techniques will be evaluated using a specially designed anthropomorphic breast imaging phantom which mimics the T(1p) and magnetization transfer measured in the excised tissues. Clinical evaluation of the spin-locking techniques will be carried out on patient volunteers with known breast lesions scheduled for surgical biopsy. Particular emphasis will be placed on women with predominantly fibro-glandular breasts with lesions not well seen by X-ray mammography. Spin-locked image contrast will be compared to conventional MR image contrast and mammographic findings and grouped according to diagnosis to determine the benefit of spin locking for improving detection of breast cancer. The spin-locked breast images will be combined with conventional MR breast images to provide improved 3-D visualization of breast cancer. Furthermore, a theoretical model of magnetization transfer will be applied to the T(1p) relaxation measurements of the excised breast tissues in order to extract magnetization transfer rates and proton fractions. T(1p) and magnetization transfer information will be grouped according to diagnosis to determine the potential of spin locking for in vivo discrimination of malignant from benign breast lesions.
