Encrypted login | home

Program Information

A 3D Reconstruction Algorithm for Superparamagnetic Relaxometry

S Loupot

S Loupot*, D Fuentes , W Stefan , J Sovizi , K Mathieu , J Hazle , UT MD Anderson Cancer Center, Houston, TX


TH-AB-708-1 (Thursday, August 3, 2017) 7:30 AM - 9:30 AM Room: 708

Purpose: Superparamagnetic relaxometry (SPMR) using biologically targeted iron oxide nanoparticles is an emerging technology for the early detection of cancer. Currently, the number of tumors and an estimate of their location is required to reconstruct the targeted particle distribution. This is not practical for clinical applications of the technology. The goal of this work is to validate a volumetric reconstruction algorithm for locating the iron oxide nanoparticles without prior knowledge of the tumor location.

Methods: The volumetric reconstruction algorithm was compared to current point-based methods for reconstructing the iron oxide dipoles in phantoms with known quantities of bound particles. A blinded study was performed to determine the specificity and sensitivity of each algorithm to detect a tumor in the presence of a strong background signal. These verification tests provide confidence for testing the algorithm feasibility in vivo.

Results: Our algorithm agreed with the current method on a linear titration of particle quantities in phantoms, and could locate a source with 3mm accuracy. In the blinded study, our algorithm showed 100% specificity and 86% sensitivity, while the conventional method had 86% sensitivity but only 86% specificity. The algorithm was able to locate a tumor in vivo.

Conclusion: This work represents a significant improvement on the current method of source reconstruction in SPMR by overcoming the requirement of prior knowledge of the tumor location. Additionally, it brings SPMR from a point-based measurement to a volumetric representation of a bound particle distribution. This algorithm will likely enable translation of SPMR technology to human trials as a novel method of early cancer detection.

Funding Support, Disclosures, and Conflict of Interest: This research was supported by the National Cancer Institute of the National Institutes of Health under Award Number F31CA210434 and by Imagion Biosystems.

Contact Email: