Title |
A sparse reconstruction algorithm for superparamagnetic relaxometry
|
Institution |
UNIVERSITY OF TX MD ANDERSON CAN CTR, HOUSTON, TX
|
Principal Investigator |
LOUPOT, SARA
|
NCI Program Director |
Perkins
|
Cancer Activity |
Training
|
Division |
CCT
|
Funded Amount |
$28,044
|
Project Dates |
08/01/2016 - 07/31/2018
|
Fiscal Year |
2017
|
Project Type |
Grant
|
Research Topics w/ Percent Relevance |
Cancer Types w/ Percent Relevance |
Bioengineering (100.0%)
Cancer (100.0%)
|
N/A
|
Research Type |
Resources and Infrastructure Related to Detection, Diagnosis, or Prognosis
|
Abstract |
"Project Summary Superparamagnetic relaxometry (SPMR) is a novel nanoparticle imaging technique that utilizes the magnetic properties of biologically targeted superparamagnetic nanoparticles to potentially detect as few as 15,000 cancer cells. Source reconstruction in SPMR requires solving the ill-posed magnetic inverse problem. There is currently a gap in knowledge about how to solve this inverse problem in order to determine which of the many possible solutions represents the true location of bound particles. The long-term goal of this project is to translate SPMR into the clinic as an early detection technique for cancer. The objective for this project is to develop an algorithm that can reconstruct the location of cancer-bound nanoparticles in 3 dimensions without any prior knowledge of the number of sites with bound particles. The hypothesis of the work is that a sparse reconstruction algorithm based on physics models and tuned to the SPMR environment will reliably reconstruct the 3-dimensional distribution of cancer-bound nanoparticles. We plan to test this hypothesis with these specific aims: Specific Aim 1: Develop an experimentally informed forward model. The forward model for the sparse reconstruction algorithm will be based on the application of the Biot-Savart law to the physical conditions of the MRX device. The model will then be adjusted to best simulate data collected from the device. Specific Aim 2: Apply and characterize the performance of the inverse algorithm. A sparse reconstruction algorithm will be implemented to reconstruct the distribution of particles from the signal returned by the detectors. The sensitivity, resolution and accuracy of the algorithm across a range of environmental and user-defined variables will then be characterized and optimized. The expected outcome of these aims is a novel reconstruction algorithm that will significantly improve source localization and quantification in magnetic relaxometry. The development of a robust and well characterized reconstruction method will positively impact the field of SPMR by opening it up to possible applications in image guidance and novel early detection techniques. The knowledge that gained of the minimum detectability of the algorithm and the characterization of its response with respect to environmental noise and optimization parameters will inform the design of future experiments towards preclinical studies. The robust reconstruction algorithm developed by this project will bring this novel technology one step closer to realizing its potential to detect early disease with unparalleled sensitivity and specificity." |