Back to the list
Congress: ECR25
Poster Number: C-21700
Type: Poster: EPOS Radiologist (scientific)
Authorblock: A. Choux1, Z. Yin2, C. L. Kim2, A. Pourmorteza1; 1Atlanta, GA, GA/US, 2Niskayuna, NY/US
Disclosures:
Arnaud Choux: Research/Grant Support: GE HealthCare
Zhye Yin: Employee: GE HealthCare
Chang Lyong Kim: Employee: GE HealthCare
Amir Pourmorteza: Grant Recipient: GE HealthCare
Keywords: Computer applications, CT, CT-Angiography, CT-Quantitative, Physics, Cancer
Methods and materials

A cylindrical test object (diameter=20cm, height=5cm) was filled with ballistic gel and contained 137 2-mL test tubes filled with calibrated concentrations of materials including: iodine-, gadolinium, tantalum, bismuth-based contrast agents and their mixtures and dilutions in water and simulated blood.

Fig 2: Test object used in the study. Left: Photo. Right: CT reconstruction.
Each material and its water dilutions is represented as a line in the multi-energy space. The angle between these lines was chosen as the metric of spectral separability. We scanned the phantom on a prototype deep-Si PCD-CT [2]. The first 3 thresholds are dynamic and capture Compton effects. The last 5 bins are adjustable and were used for material separation. They were once set to optimize the number of detected photons in each energy bin and optimize iodine detectability (setting1), and once to capture K-edges of multiple CAs (setting2). The phantom was scanned at 120 kVp,300mAs. Energy-bin images were reconstructed with filtered-backprojection and a standard kernel.
Table 1: Image acquisition and reconstruction parameters.
GALLERY