ARRS 2022 Abstracts

RETURN TO ABSTRACT LISTING


1426. Dual-Energy vs. Photon-Counting CT in Characterizing Multiple Myeloma
Authors * Denotes Presenting Author
  1. Ehsan Abadi *; Duke University
  2. Cindy McCabe; Duke University
  3. William Segars; Duke University
  4. Fides Schwartz; Duke University
  5. Erin McCrum; Duke University
  6. Emily Vinson; Duke University
  7. Ehsan Samei; Duke University
Objective:
The treatment of multiple myeloma, a cancer of the plasma cells originating in the bone marrow, is highly dependent on accurate early-stage diagnosis and characterizations. CT technologies of dual-energy and photon-counting CT exhibit great potential. The purpose of this study was to understand the comparative advantages of the spectral and spatial characterization of the two technologies for the characterization of multiple myeloma with the use of virtual imaging trials, a platform which provides realism and ability to compare computational images of human models across multiple imaging scenarios.

Materials and Methods:
For the patient model, we used an anthropomorphic computational (XCAT) phantom of an adult male with a BMI of 28.5 kg/m2. We incorporated various myeloma lesions of differing shapes and sizes within the vertebrae of the XCAT phantom. The lesion morphology was extracted from clinical patient data with materials defined as a mixture of yellow and red marrows and trabecular bone. The phantom was virtually imaged using a validated CT simulator (DukeSim), modeling both a commercial dual-source, dual-energy energy CT scanner (Force, Siemens) and a new, investigational dual-source photon-counting prototype (NAEOTOM Alpha, Siemens) at multiple tube voltages, dose levels, and detector energy thresholds (for photon-counting). Using the spectral sinogram images, virtual monoenergetic images were reconstructed with a vendor-specific reconstruction software (ReconCT, Siemens) at multiple reconstruction kernels and settings (e.g., slice thickness and pixel size). The acquired images were quantitatively evaluated in terms of lesion detection and classification from normal marrow and edema.

Results:
Photon-counting images offered improved lesion detection particularly for smaller lesions, attributable to PCCT’s high spatial resolution (0.2 mm detector size), and lesion classification, attributable to improved energy separation. The results offered choices to render each technology at its best task performance for multiple myeloma.

Conclusion:
Dual-energy CT and the emerging photon-counting CT have the potential to improve the diagnostic accuracy in multiple myeloma. In this study, we comprehensively evaluated with findings that pave the way towards more optimized and accurate patient treatment and care.