Quantitative 4D-PET Reconstruction for Small Animal Using SMEIR-Reconstructed 4D-CBCT

Respiratory motions in small animals positron emission tomography (PET) cause image degradation during reconstruction. This paper aims to develop a motion compensated 4D-PET reconstruction method using accurate motion corrections and attenuation corrections (ACs) from 4D-cone beam computed tomography (CBCT) images reconstructed using a simultaneous motion estimation and image reconstruction (SMEIR) method. Projections of 4D-CBCT were calculated using a ray-tracing method on a digital 4D rat phantom, and list-mode data of 4D-PET with matched respiratory phases were simulated using the GATE Monte Carlo package. The respiratory rate was set at 1.0 s per cycle with 10 phases of 30 projection images each. 4D-CBCT images were reconstructed using the SMEIR method and motion information and linear attenuation from 4D-CBCT were subsequently used for motion compensated 4D-PET reconstruction and ACs. We quantitatively evaluate the reconstructed 4D-PET using the errors of tumor volume and standard uptake values of tumors with different sizes. The tumor motion was successfully reconstructed and showed good agreement with the original phantom. The proposed method reduced tumor volume errors and standard uptake value errors. For tumor diameters of 3.0, 4.5, and 6.0 mm, the tumor volume errors are 32.5%, 29.2%, and 19.4%, respectively, with motion compensation and 45.1%, 37.5%, and 20.2%, respectively, without compensation.