Dynamic contrast-enhanced (DCE) magnetic resonance imaging (MRI) of breast cancer patients has shown considerable promise in aiding diagnoses of breast lesions and characterizing treatment response. The challenge in DCE breast imaging is the need for both good temporal resolution to capture tracer kinetic properties and good spatial resolution for visualizing morphology. Traditional dynamic methods in MRI acquire incomplete k-space data at each time point, and use k-space temporal interpolation (or data sharing) to form "complete" k-space datasets prior to Fourier reconstruction. We propose to investigate model-based image reconstruction methods that avoid k-space interpolation by estimating the object model parameters that best fit the available k-space data. These reconstruction methods will incorporate parallel imaging techniques. They will also be extended to account for nonrigid deformations due to patient motion during the scan using novel methods for joint estimation of motion parameters and image intensity parameters. The methods will be evaluated using computer simulations, phantom studies, and human DCE-MRI scan data. The human data will be collected as part of Project 1 and will include DCE-MRI scans of breast cancer patients undergoing neoadjuvant chemotherapy, where early prediction of tumor response is of clinical importance. The proposed methods have the potential to improve image quality both in breast DCE-MRI as well as other dynamic MR applications.

Public Health Relevance

The relevance of this research to public health is that improving the quality of MR images through more sophisticated data processing may lead to more accurate diagnosis and treatment of patients with breast cancer and other diseases.

Agency
National Institute of Health (NIH)
Institute
National Cancer Institute (NCI)
Type
Research Program Projects (P01)
Project #
5P01CA087634-10
Application #
8445394
Study Section
Special Emphasis Panel (ZCA1-GRB-P)
Project Start
Project End
2015-02-28
Budget Start
2013-03-01
Budget End
2014-02-28
Support Year
10
Fiscal Year
2013
Total Cost
$209,071
Indirect Cost
$64,965
Name
University of Michigan Ann Arbor
Department
Type
DUNS #
073133571
City
Ann Arbor
State
MI
Country
United States
Zip Code
48109
Nataraj, Gopal; Nielsen, Jon-Fredrick; Fessler, Jeffrey (2016) Optimizing MR Scan Design for Model-Based T1, T2 Estimation from Steady-State Sequences. IEEE Trans Med Imaging :
Larson, Eric D; Lee, Won-Mean; Roubidoux, Marilyn A et al. (2016) Automated Breast Ultrasound: Dual-Sided Compared with Single-Sided Imaging. Ultrasound Med Biol 42:2072-82
Berisha, Visar; Wisler, Alan; Hero, Alfred O et al. (2016) Empirically Estimable Classification Bounds Based on a Nonparametric Divergence Measure. IEEE Trans Signal Process 64:580-591
Hero, Alfred O; Rajaratnam, Bala (2016) Foundational Principles for Large-Scale Inference: Illustrations Through Correlation Mining. Proc IEEE Inst Electr Electron Eng 104:93-110
Galbán, Craig J; Ma, Bing; Malyarenko, Dariya et al. (2015) Multi-site clinical evaluation of DW-MRI as a treatment response metric for breast cancer patients undergoing neoadjuvant chemotherapy. PLoS One 10:e0122151
Brisset, Jean-Christophe; Hoff, Benjamin A; Chenevert, Thomas L et al. (2015) Integrated multimodal imaging of dynamic bone-tumor alterations associated with metastatic prostate cancer. PLoS One 10:e0123877
Muckley, Matthew J; Noll, Douglas C; Fessler, Jeffrey A (2015) Fast parallel MR image reconstruction via B1-based, adaptive restart, iterative soft thresholding algorithms (BARISTA). IEEE Trans Med Imaging 34:578-88
Zhao, Feng; Fessler, Jeffrey A; Wright, Steven M et al. (2014) Regularized estimation of magnitude and phase of multi-coil b1 field via Bloch-Siegert B1 mapping and coil combination optimizations. IEEE Trans Med Imaging 33:2020-30
Boes, Jennifer L; Hoff, Benjamin A; Hylton, Nola et al. (2014) Image registration for quantitative parametric response mapping of cancer treatment response. Transl Oncol 7:101-10
Watanabe, Takanori; Kessler, Daniel; Scott, Clayton et al. (2014) Disease prediction based on functional connectomes using a scalable and spatially-informed support vector machine. Neuroimage 96:183-202

Showing the most recent 10 out of 85 publications