Radiologists who want to look under the hood and study the engine driving this technological break-through will notice “FIRST means fast”. Toshiba accelerated computational throughput to bring their iterative reconstruction technique to the clinic where extensive reconstruction times are not acceptable. In FIRST, a reconstruction from the original (measured) projection data is performed with traditional FBP to create a seed image. This image then undergoes forward projection to create a forward projected dataset (essentially a projection summation of HU values along each ray path projection). The forward projection data is then compared against the original projection data to identify the differences between the two. This essentially identifies errors in the FBP reconstruction, which are compensated for. A subsequent correction step is then performed to update the seed image. This is a sophisticated process because additional information concerning the focal spot size (optics model), photoelectric noise (statistical model), the cone beam trajectory (cone beam model), and the exact scan parameters (system model) are utilized in the update process.
This process is iteratively repeated until the differences between the original (measured) projection data and the forward projection data are minimal, ensuring that a highly accurate image of the patient is produced with increased spatial resolution and low contrast media. Because this process is coupled with anatomically based noise regularization, FIRST is incredibly robust for data with extremely low photon counts, which translates to an opportunity to achieve significant reductions in radiation dose for all examinations.
Available for the Aquilion ONE family of CT systems, FIRST improves high-contrast spatial resolution while making exams safer for patients by providing ultra-low dose examinations.
Following Toshiba’s long-standing philosophy of minimizing dose for all patients, FIRST:
- Is integrated into SUREExposure 3D, ensuring automatic dose reduction
- Requires as little as 3 minutes for full volumetric reconstruction
- Reconstructs in parallel with InstaView realtime reconstruction
- Delivers unsurpassed natural-looking image quality due to anatomical modeling depending on the patient’s anatomy
FIRST IN ROUTINE CLINICAL PRACTICE