Click here to


Are you sure ?

Yes, do it No, cancel

Expedient System Optimization of X-Ray Fluorescence Computed Tomography (XFCT) Using Analytical Point Spread Function and Model Observer

H Tseng1*, S Vedantham2 , S Cho3 , A Karellas4 , (1) University of Arizona, Tucson, AZ, (2) University of Arizona - Banner University Medical Center, Tucson, AZ, (3) UT MD Anderson Cancer Center, Houston, TX, (4) University of Arizona, Tucson, AZ


(Sunday, 7/14/2019) 4:00 PM - 5:00 PM

Room: 303

Purpose: To accelerate the optimization process for X-Ray Fluorescence Computed Tomography (XFCT) system design using analytical Point Spread Functions (PSF) for four different types of pinhole collimator designs, along with the Channelized Hotelling Observer (CHO), for applications in small-animal imaging.

Methods: Single pinhole (SPH) collimator and three types of multi-pinhole (MPH) collimators - non-focusing right-circular double-cone (NFRCDC), focusing right-circular double-cone (FRCDC), and focusing oblique-circular double-cone (FOCDC) - were examined in this study. Two types of septal materials, gold and lead were examined. A digital cylindrical phantom (0.5mm voxels) of 25 mm diameter with centrally-located 6 mm diameter insert producing signal was modeled. System matrices were implemented by analytical PSF of the pinhole collimators for forward and back projection. Poisson noise was added to the projection data (16 equiangular views) before image reconstruction using Maximum-Likelihood Expectation-Maximization (MLEM) algorithm with 10 iterations. Signal-present and signal-absent images were generated and separated into training and testing image data sets for the detection task performed by CHO with the well-known 10 Dense Difference-of-Gaussian channels. The area under the ROC curve was used as the image quality metric for system optimization.

Results: The computational time for generating the system matrix was reduced from 50 minutes (Monte Carlo simulations) to about 1 second when analytical PSF was used. Results indicate that the MPH-XFCT system has superior image quality compared to SPH-XFCT system. Overall, FOCDC based collimator provided the best performance for XFCT. This result was consistent with the theoretical expectation. The septal material can be chosen based on the energy of the x-ray fluorescent nanoparticle.

Conclusion: The computational acceleration achieved by the generation of PSF-based system matrix avoids the pre-calculation and the storage space for forward and back projection. This expedites the tuning of parameters when designing and optimizing XFCT systems.


Not Applicable / None Entered.


Not Applicable / None Entered.

Contact Email