MENU

Click here to

×

Are you sure ?

Yes, do it No, cancel
  1. Home
  2. Meeting Program
  3. All
  4. CT Dose Calculation Via a Fast, Deterministic Boltzmann Transport Equation Solver: Validation Against AAPM Task Group 195 Monte Carlo Reference Data

CT Dose Calculation Via a Fast, Deterministic Boltzmann Transport Equation Solver: Validation Against AAPM Task Group 195 Monte Carlo Reference Data

A Wang1*, A Maslowski1 , T Wareing1 , T Schmidt2 , J Star-Lack1 , (1) Varian Medical Systems, Palo Alto, CA, (2) Marquette University, Milwaukee, WI

Presentations

(Thursday, 8/2/2018) 10:00 AM - 12:00 PM

Room: Room 202

Purpose: To benchmark a software tool for computing CT dose against AAPM TG195 Monte Carlo reference data.

Methods: Monte Carlo methods have been well-established for calculating CT dose maps, but require immense computational resources to achieve sufficiently high statistical accuracy. To overcome this limitation, we implemented a deterministic method to solve the same underlying Boltzmann Transport Equation (BTE) that governs photon interactions. The computation was discretized in spatial location, energy, and angle, and an efficient GPU implementation of the deterministic BTE solver was applied to compute the object’s photon fluence distribution, which does not exhibit stochastic noise. The performance of our BTE solver was benchmarked using the CT dose calculation task in AAPM TG195 for a voxelated anthropomorphic phantom (Case 5), which has a volume size of 50x32x26 cm³ with 1 mm isotropic voxels. The influence of voxel size (ranging from 4 – 20 mm) used in the BTE solver to compute photon fluence was assessed for accuracy (organ dose error) and calculation time. Additional acquisition types such as an axial cone-beam CT and helical CT were assessed against a previously validated Geant4 Monte Carlo implementation, which used 1 mm voxels to establish the reference organ doses.

Results: For the BTE solver, organ dose error of <1% was achievable with 5 mm voxels and calculation time of 12 seconds for the axial cone-beam CT. Similar levels of accuracy were achieved for the simulated helical CT. We found increasing error and decreasing calculation time with larger voxels, as expected. The run times are a vast reduction from the 30 CPU-hours for 1% uncertainty reported by TG195 for analog Geant4.

Conclusion: A deterministic BTE solver offers a fast, accurate alternative to Monte Carlo methods for computing CT dose. Such technology may enable routine estimation of patient-specific organ doses for every CT scan.

Funding Support, Disclosures, and Conflict of Interest: This work was funded by NIH U01EB023822. TS receives research funding from GE Healthcare. AW, AM, TW, and JSL are employees of Varian Medical Systems.

Keywords

Dose, Monte Carlo, CT

Taxonomy

IM- CT: Radiation dosimetry & risk

Contact Email

Additional Links