Purpose: To develop a patient-specific dedicated breast CT (DBCT) method and apparatus that enhances the contrast resolution of small microcalcifications, and to validate the improvements via a dose-matched comparative study against three clinically investigated cone-beam and helical CT systems.
Methods: A new DBCT design, Anatomically Specified Conformation Computed Tomography (ASC-CT), is introduced. The architecture of the apparatus comprises of decoupling of the robotics of the source assembly from the detection assembly by providing two independently rotating gantries, along with utilization a multiplicity of Time-Delay-Integration (TDI) mode line detectors in the detection assembly. The method of image acquisition entails adjusting the layout of detectors according to the specific shape of the breast of interest, followed by capturing collimated beams by detector panels while sweeping circular trajectories around the breast.Feasibility of ASC-CT was verified through a series of Monte Carlo (MC) simulations. First, we quantified the dependency of microcalcification contrast resolution (assessed in terms of signal-difference-to-noise-radio, SDNR) on scattered photon acquisition. In another MC study, we compared the performance of ASC-CT in detecting microcalcifications with different sizes against three well-studied DBCT systems. Radiation dose was matched in all scan procedures. An average homogeneous breast (14 cm in diameter, 15% glandular fraction) was simulated and used in all MC studies.
Results: Depending the position of the microcalcifications in the breast, SDNR drops 26% to 52% due to high-frequency components of scattered radiation. The comparative study revealed 115% and 62% gains in average SDNR of 250-micron microcalcifications for ASC-CT over cone-beam CT and spiral CT, respectively.
Conclusion: Both signal level and contrast resolution of microcalcifications are highly dependent on the acquisition of the scattered photons. The higher the scatter level, the lower the reconstructed microcalcification signal. SDNR is maximized in the case of no-scatter image acquisition outlined in ASC-CT.
Funding Support, Disclosures, and Conflict of Interest: The research disclosed in this presentation is fully funded by Malcova LLC.
Not Applicable / None Entered.