NOVEL DIFFUSION TENSOR IMAGING CAD FOR BREAST CANCER

signals in 6 directions. If random motion is assumed isotropic, then Diffusion Weighted Imaging (DWI), which only requires acquisition of signals in 3 directions, can be used, Diagnostic values obtained from diffusion analysis are Apparent Diffusion Coefficient (ADC), obtained from DWI, and Fractional Anisotropy (FA), obtained from DTI. The specific aim of this study is to establish feasibility of a novel morphological feature derived from DWI than can be used with ADC and FA values to improve the sensitivity and specificity of detecting and diagnosing breast cancer using a DTI computer-aided-diagnosis (CAD) system. DTI is acquired in substantially less time than dynamic-contrast-enhanced (DCE) imaging and does not require use of potentially toxic enhancingagents. The short term benefit of an effective breast DTI CAD is improvement in clinical accuracy of breast cancer detection and diagnosis when used in conjunction with existing DCE. Long term potential benefits include substantially reducing cost and eliminating potential patient harm by enabling DTI to be used in place of DCE for detection of specific types of breast cancer for which DTI is shown to have higher sensitivity than DCE.