Comprehensive Characterization of Breast Tissue Using Multi-modal MRI

点击翻译按钮获取中文摘要

ABSTRACT Diffusion-weighted MRI (DWI) has become a pillar of clinical MRI with its uniqueness to probe tissue properties well beyond the conventional MRI can achieve with its millimetric resolution. Most DWI applications rely on apparent diffusion coefficient (ADC) which has been associated with tissue cellularity1,2. However, the comprehensive properties of tissue cannot be adequately revealed by a single measurement3. Recent publications indicated these hidden properties can be probed using specific portions of the b-value “spectrum” in DWI4. For example, intravoxel incoherent motion model5–7 with low-b-values can characterize micro- vascularity without contrast agent, presenting a potential for complementing dynamic contrast-enhanced MRI (DCE-MRI)8. High-b-value non-Gaussian models9–24 can probe microstructures as in the case with continuous- time random walk model17,19,21,25,26 which can reflect tissue micro-structural heterogeneity, thus enable studying microstructures beyond cellularity and vascularity. Recognizing the complexity of the breast tissue, the breast imaging community has shown an increasing interest in incorporating advanced DWI into their breast MRI protocols27. However, expansion of advanced DWI to the breast has not been fully achieved28,29 due to two major challenges. Firstly, commercial single-shot echo-planar imaging (ss-EPI) is prone to image distortion, which is exacerbated in the breast due to off-center setting30–37. Secondly, an integrated approach with multiple MRI- based metrics has not been established for breast tissue characterization38,39. The main goal of this project is to develop two imaging tools to enable a comprehensive characterization of breast tissue: (1) a novel, distortion- resilient, and time-efficient DWI acquisition technique, and (2) an integrated multi-modal MRI analysis framework. Our overarching motivation is to bring the advancements in multi-high-b-value DWI to the breast. The hypothesis is that breast neoplasm can be evaluated by characterizing tissue cellularity, vascularity, and heterogeneity collectively; and these properties can be comprehensively probed by utilizing a set of DWI parameters from the entire spectrum of b-values together with DCE-MRI metrics. The specific aims are: 1. To develop a novel, distortion-resilient, and time-efficient pulse sequence, Steer-PROP for breast DWI – SPREAD – that will enable DWI with a full spectrum of b-values from 0 to 3000 s/mm2. 2. To establish MRI-based metrics for characterizing tissue cellularity, vascularity, and heterogeneity; and to develop an integrated multi-modal MRI analysis framework – TERMINAL – for the breast. 3. To demonstrate a potential application of SPREAD and TERMINAL in the context of predicting response to neoadjuvant chemotherapy (NAC) in breast cancer. This project will provide novel tools to enable DWI acquisition with a full b-value spectrum and DWI analysis for assessing malignancy and evaluating treatment response in breast cancer. Successful completion of the project will serve as a prototype for the expansion of advanced DWI into breast imaging.