PET attenuation correction and non-specific uptake normalization for emission guided radiation therapy

Objectives Emission guided radiation therapy(EGRT) is a new modality that uses PET emissions in real-time for direct tumor tracking during radiation delivery. Radiation beamlets are delivered along PET lines of response(LOR’s) by a fast rotating ring therapy unit consisting of a linear accelerator and PET detectors. Attenuation and non-specific tracer uptake cause inhomogeneous LOR detection and response, thereby limiting EGRT’s capability of incorporating intensity modulated radiation therapy(IMRT) to achieve dose planning goals. Here we develop both PET attenuation correction and non-specific uptake normalization methods.

Methods We sample EGRT’s detection and delivery space into 3D parallel sinogram space. Every detected LOR is mapped into one of the sinogram bins. Non-uniform LOR detection is compensated by applying a LOR response probability distribution in the same sinogram space so that responses to LOR’s with lower attenuation or larger non-specific activity integrals are less likely, and vice versa. Simulations are performed on a 4D digital XCAT patient with a lung tumor. The radiation delivery is simulated by a voxel-based Monte Carlo algorithm and the emission data is generated using the Geant4 Application for Tomographic Emission package. 3D IMRT, EGRT with or w/o attenuation correction and activity normalization are simulated.

Results Compared with 3D IMRT, when evaluating the dose to 95% of the gross tumor volume, EGRT w/o corrections, with attenuation correction, and with correction & normalization yields an 11%, 18%, and 21% relative increase respectively. All studies are normalized to have the same lung dose.

Conclusions As in PET imaging, attenuation correction improves the performance of EGRT. Unlike imaging however, additional normalization for non-specific uptake is necessary to achieve a more uniform dose distribution. These methods provide the basis for EGRT to accomplish sophisticated dose modulation to achieve treatment planning goals.

Research Support This work is supported by Georgia Institute of Technology new faculty startup fund, RefleXion Medical and the National Cancer Institute (R43CA153466).