There is an urgent need for cost-effective, non-invasive diagnostic techniques for the evaluation of breast lesions. The primary objective of this project is to develop a multiparameter MRI based tissue model to differentiate benign from malignant tissue that may be used for the purpose of selecting treatment options. While many imaging techniques (particularly contrast-enhanced MRI) now provide high sensitivity for the detection of breast lesions, specificity remains relatively low, resulting in many biopsies of lesions that have a benign final diagnosis. Our proposed approach investigates a mechanistic hypothesis that multiparameter MRI, by accounting for the multiple biophysical states of water in tissue and tissue perfusion, _ yield a more complete classification of breast tissue than any single MRI parameter. Potentially malignant tissue can be identified by combining multiparameter MRI data [Tl-weighted (T1WI), T2-weighted (T2WI), pre- and post- dynamic contrast enhanced gadolinium (GdDTPA), soditma, and spectroscopic MR images] in conjunction with cluster analysis methods (ISODATA and angle analysis) into a single vector, to improve differentiation between benign or malignant breast lesions and breast tissue. Based on our development of MRI based tissue models to differentiate and diagnose breast cancer, we propose to accomplish these objectives using a multiparameter MRI data approach in conjunction with sophisticated data analysis routines. This approach Hill enable us to monitor and predict which patients will respond to chemotherapy and those patients that will not respond by monitoring the ISODATA vector characteristics of the breast tissue in chemotherapy.
Showing the most recent 10 out of 40 publications
