Detection and characterization of breast lesions: Color-coded signal intensity curve software for magnetic resonance-based breast imaging

This chapter explores several innovative techniques employed in the study of breast cancer. One of the technologies, contrast-enhanced magnetic resonance (MR) mammography (CE-MRM) can help to distinguish between benign and malignant lesions on the basis of lesion morphology and in most cases through the profile of the curve for lesion signal intensity enhancement over time (SI/T curve). Magnetic resonance (MR) images acquired with high spatial resolution are needed for accurate assessment of the morphologic characteristics and internal architecture of lesions, while fast imaging protocols with high temporal resolution are needed to evaluate the enhancement kinetics of lesions. Two principal strategies have evolved to improve specificity: rapid dynamic imaging after gadolinium enhancement and high spatial resolution imaging. CE- magnetic resonance imaging (MRI), besides detecting morphological information of the lesion, can provide dynamic information on the contrast medium behavior related to the intensity of MRI signal over time. The signal intensity-time curves of malignant lesions have characteristic washout profiles, whereas the curves of benign lesions more frequently have a steady increase or plateau profile. A digital imaging and communication in medicine (DICOM)-compatible software package is developed, which is able to automatically load a complete breast MRI study and to automatically register and subtract images before creating a false color map (FCM) for each scan plane. The approach compares the qualitative parameters of the signal intensity enhancement over time (SI/T) evolution of each point with the standard lesion curve types and modulates the color brightness with the earliness of the enhancement peak. This qualitative approach to lesion classification greatly reduces the computational effort and the time required for examination analysis even on an individual PC-based system. © 2008 Elsevier Inc.