MENU

Click here to

×

Are you sure ?

Yes, do it No, cancel

Comparison of Two Reconstruction Methods for Bound-Nanoparticle Detection Using Superparamagnetic Relaxometry

S Thrower*, S Kandala , D Fuentes , K Mathieu , W Stefan , A Kulp , J Hazle , UT MD Anderson Cancer Center, Houston, TX

Presentations

(Sunday, 7/29/2018) 4:00 PM - 6:00 PM

Room: Davidson Ballroom B

Purpose: Superparamagnetic relaxometry (SPMR) is an emerging technology that can detect small quantities of immobilized nanoparticles with high sensitivity and specificity. The current approach to source reconstruction in SPMR, multiple source analysis (MSA), does not adequately detect small tumors in small animal models due to the presence of high nanoparticle uptake in the liver. We compare the ability of the sparsity averaging reweighting analysis (SARA) to detect small amounts of bound nanoparticles in close proximity to large amounts of bound nanoparticles to that of MSA in a user-blinded phantom study.

Methods: Using both SARA and MSA, we asked a user to classify 60 SPMR measurements of phantom configurations that simulated no binding (no phantom, n=6), binding in the liver only (1 source n=20), and binding in the liver and tumor on the left flank (2 sources, left n=17) or right flank (2 sources, right, n=17). The phantoms consist of cotton swabs with between 0.39µg and 50µg of 25nm spherical SPIO nanoparticles dried on the tip. The performance of the algorithm was quantified by the overall classification accuracy.

Results: The overall classification accuracy was 43% (26/60) for MSA and 82% (49/60) for SARA. Both algorithms correctly identified 6/6 cases with no source. MSA correctly classified 18/20 cases with one source, and 2/34 cases with two sources. SARA correctly classified 12/20 cases with one source and 31/34 cases with two sources.

Conclusion: SARA improved the overall classification accuracy by 39 percentage points over the current method of reconstruction and substantially improved the detection of two source cases (6.25% vs 91% correct), with only a small decrease in accuracy for one source cases due to false positives. This will greatly improve our ability to conduct the small animal trials necessary to translate this promising technology into the clinic.

Funding Support, Disclosures, and Conflict of Interest: This work was undertaken as part of a Sponsored Research Agreement with Imagion Biosystems, Inc. and, in part, facilitated by collaboration with the Center for Integrated Nanotechnologies, an Office of Science User Facility operated for the US Department of Energy (DOE) Office of Science.

Keywords

Reconstruction, Optimization, Low Contrast Detectability

Taxonomy

IM- Other (General): Nanoparticles (general)

Contact Email