Encrypted login | home

Program Information

Gold Nanoparticle and Iodine Prediction of Concentration Using Dual Energy Computed Tomography in Phantoms

A Melancon

L Lu1 , M Jacobsen2 , T Li1 , M Jonathan4 , E Tasciotti4 , R Layman2 , M Melancon1 , A Melancon3*, Departments of (1) Interventional Radiology, (2) Imaging Physics, (3) and Radiation Physics, University of Texas M D Anderson Cancer Center, Houston, TX, (4) Methodist Research Hospital, Houston, TX


MO-RAM-GePD-I-6 (Monday, July 31, 2017) 9:30 AM - 10:00 AM Room: Imaging ePoster Lounge

Purpose: To quantify gold nanoparticle (AuNP) concentrations in AuNP-iodine phantom using data available from dual-energy computed tomography (DECT) for the development of radiopaque inferior vena cava filters (IVCF).

Methods: Multiple agar and blood phantoms of varying concentrations of hydrophilic AuNP and iodine were constructed. The phantoms were scanned in a GE DECT scanner and reconstructed with the GSI software. The phantoms were then prepared for ICP-OES to obtain concentrations of AuNP. A standard curve relating measured AuNP concentration to water-iodine material decomposition was created. A bisector line of a AuNP vs iodine concentration graph that gave the best prediction for the concentrations was created using MATLAB. HU vs concentration graphs were created to predict the HU values of mixed AuNP-iodine samples given concentrations and to predict concentration given total HU and total concentration.

Results: There was a positive correlation of AuNP concentration from ICP to water-iodine material decomposition (R² = 0.953 in agar and 0.4135 in blood). The bisector line also predicted concentrations with a positive correlation (R² = 0.882 in agar and 0.856 in blood). Prediction of HU values showed good correlation but with errors of <5% for the agar phantoms and systematic errors around 30 units for the blood phantoms. The results of the concentration predictions were similar except for AuNP concentration much less than 1 mg/ml where our model performed poorly.

Conclusion: The ability to discriminate resorbable IVC filters in the presence of iodinated blood is important to ensure its integrity and location. This work shows the potential to predict AuNP concentrations in iodine using currently available spectral imaging software. Preliminary tests have shown correlations between AuNP and water-iodine concentrations, and HU and concentration. Further investigation will involve developing correction factors to eliminate error and extending the procedure to imaging IVCF in large animals.

Contact Email: