Room: Room 205
Purpose: To quantify accuracy of kidney stone characterization as a function of radiation dose for dual-source (DS), dual-energy (DE) CT scanning protocols.
Methods: 88 kidney stones of known composition were suspended in gelatin, placed in water phantoms, and scanned using second- and third-generation DS scanners. Matching of kV pairings and phantom lateral widths (both scanners: 35cm, 40cm, 45cm, & 50cm; plus 55cm in third-generation) was performed using routine clinical protocols. Each phantom was scanned at 100%, 75% and 50% of routine clinical dose on both scanners, as well as 33% on the third-generation scanner. All scans used automatic exposure control. Kidney stone composition was determined using commercial software, and accuracy was evaluated as a function of phantom size, dose, kV combination, and use of iterative reconstruction.
Results: Routine dose ranged from CTDIvol=7.07mGy to 33.27mGy based on scanner type, phantom size, and kV combination. On the second-generation DS scanner, dose levels of 50% and 75% in the 35cm and 40cm phantoms, respectively, resulted in 100% and 97% correct stone characterization; performance in larger phantoms fell to as low as 47% as dose decreased. For the third-generation DS scanner, dose levels as low as 33% led to 100%, 99%, and 97% correct stone characterization in the 35cm, 40cm, and 50cm phantoms, respectively. Dose levels as low as 50% were acceptable in the 45cm phantom (98% correct). Any dose reduction for larger (55cm) phantoms resulted in 85% or lower stone characterization accuracy. For the second-generation scanner, using iterative reconstruction increased accuracy by up to 33% (at 50% dose in 50cm phantom), while in the third-generation scanner it did not improve accuracy.
Conclusion: Using kidney stone characterization as an indicator of DE performance, radiation dose in DE scans can be reduced to as low as 33%-50% of routine dose while maintaining diagnostic performance.
Funding Support, Disclosures, and Conflict of Interest: Irene Duba and Dr. Halaweish are employees of Siemens Healthineers. Dr. McCollough receives research funding support from Siemens Healthineers.