Project Abstract
Overview: In 2020, over 276,000 women will be diagnosed with invasive breast cancer, and over 48,000 will
die from it. The ultimate goal of this project is to provide fast, accurate, accessible non-contrast-enhanced MRI
screening methods to safely detect breast cancer in the high-risk population.
Relevance: Screening mammography is successful in reducing breast cancer mortality, but misses many
important cancers, especially in women with a high risk of breast cancer, and in the 27 million women with dense
breasts who undergo screening. MRI has been shown to be more than twice as sensitive for screening in many
scenarios, but high cost, discomfort, and poor compliance due to the need for IV gadolinium-based contrast
agents with side effects limit its impact for screening. This project aims to change this paradigm by developing a
rapid, inexpensive, comfortable, non-contrast-enhanced breast MRI screening exam.
Approach: Based on substantial prior work, our group will develop high-resolution 3D and 2D diffusion MRI
methods that can depict important morphologic features and quantify diffusion heterogeneity in breast lesions.
To support these methods, we will ¿rst implement ¿exible, closely-¿tting soft breast coil arrays that maximize
SNR and parallel imaging capabilities while reducing positioning time and increasing comfort. We will combine
the ef¿ciency and motion-insensitivity of non-Cartesian 3D cones imaging with the double-echo steady-state dif-
fusion approach to offer ef¿cient 3D diffusion-weighted imaging (DWI). Additionally, we will redesign echo-planar
imaging (EPI) DWI including asymmetric encoding with bulk motion insensitivity, full k-space readouts, and novel
locally low-rank reconstruction to provide high-resolution quantitative DWI that is robust to small patient motion.
Finally, collaborating with two other major centers, we will study these methods in high-risk screening patients,
aiming to demonstrate improved non-contrast-enhanced sensitivity from 45% to over 70% among women with
negative mammograms, while retaining speci¿city over 91%. If successful, these rates are suf¿cient to make
non-contrast-enhanced MRI screening viable for clinical use.
Summary: Using advances to 2D and 3D DWI, combined with ¿exible coil arrays, we aim to provide accurate,
low-cost, comfortable, MRI screening without intravenous contrast, in a 10 minute exam. This will ultimately en-
able more effective and comfortable breast cancer screening for millions women for whom x-ray mammography
is insuf¿cient.