Meeting Banner
Abstract #0200

Separation of Benign and Malignant Breast Lesions Using Pharmacokinetic Analysis for a Biopsy Cohort

Melanie Freed1, Ana Paula Klautau Leite1, 2, Jin Zhang1, Melanie Moccaldi3, Kai Tobias Block4, Linda Moy3, Sungheon Kim1

1Radiology, Center for Biomedical Imaging, New York University School of Medicine, New York, NY, United States; 2Radiology, Hosptial das Clinicas-FMUSP, School of Medicine - University of Sao Paolo, Sao Paolo, SP, Brazil; 3Radiology, Cancer Institute, New York University School of Medicine, New York, NY, United States; 4Radiology, Center for Biomedical Imaging, NYU Langone Medical Center, New York, NY, United States

Although DCE-MRI is an important tool for high-risk screening of breast cancer lesions, its suboptimal specificity results in many unnecessary biopsies. In this study, we investigate the use of pharmacokinetic analysis for discrimination of lesion types in a population of patients referred for biopsy. This population is particularly problematic since standard clinical analysis has already failed for this group. Our analysis shows that Ktrans and ve show noticeable differences between lesion types in this patient group, however, refinement of the analysis methods is necessary to further improve the separation of lesion types for this challenging cohort.

Abstract PDF Presentation Video

Keywords

accrue acquisition additional adequate administered agent already although application arterial artery assumed assuming background beginning benign berg biomedical biopsies biopsy black block blood blue breast cancer challenging channel clinic clinical cohort coil concentration considered constant contrast converted core create currently curves dedicated delayed detailed diagnosis diagnostic difficult discrimination dots dynamic enhanced enhancement example expand extracellular failed fifth fitted five forty fraction frames freed generalized gradient guided histology identified inadequate initial intensity internal kinetic lesion lesions linear malignant mammary manual manually measures median medicine model models modified necessary noted pathologist patient patients plasma population problematic promising protocol radial radiology recommended reconstructed referred refs relies represent resolution respectively risk sampling scheme school screening selected sensitivity separate separation shutter simple since society song space specificity speed spoiled spokes stack stars statistically suffers suspicious system temporal third tissue tool trained trend type types undergoing unilateral uniquely unnecessary utility variable vascular vibe view views volume women