Fast GPU-based Monte Carlo code for SPECT/CT reconstructions generates improved 177Lu images

Abstract

BackgroundFull Monte Carlo (MC)-based SPECT reconstructions have a strong potential for correcting for image degrading factors, but the reconstruction times are long. The objective of this study was to develop a highly parallel Monte Carlo code for fast, ordered subset expectation maximum (OSEM) reconstructions of SPECT/CT images. The MC code was written in the Compute Unified Device Architecture language for a computer with four graphics processing units (GPUs) (GeForce GTX Titan X, Nvidia, USA). This enabled simulations of parallel photon emissions from the voxels matrix (1283 or 2563). Each computed tomography (CT) number was converted to attenuation coefficients for photo absorption, coherent scattering, and incoherent scattering. For photon scattering, the deflection angle was determined by the differential scattering cross sections. An angular response function was developed and used to model the accepted angles for photon interaction with the crystal, and a detector scattering kernel was used for modeling the photon scattering in the detector. Predefined energy and spatial resolution kernels for the crystal were used. The MC code was implemented in the OSEM reconstruction of clinical and phantom 177Lu SPECT/CT images. The Jaszczak image quality phantom was used to evaluate the performance of the MC reconstruction in comparison with attenuated corrected (AC) OSEM reconstructions and attenuated corrected OSEM reconstructions with resolution recovery corrections (RRC).ResultThe performance of the MC code was 3200 million photons/s. The required number of photons emitted per voxel to obtain a sufficiently low noise level in the simulated image was 200 for a 1283 voxel matrix. With this number of emitted photons/voxel, the MC-based OSEM reconstruction with ten subsets was performed within 20 s/iteration. The images converged after around six iterations. Therefore, the reconstruction time was around 3 min. The activity recovery for the spheres in the Jaszczak phantom was clearly improved with MC-based OSEM reconstruction, e.g., the activity recovery was 88% for the largest sphere, while it was 66% for AC-OSEM and 79% for RRC-OSEM.ConclusionThe GPU-based MC code generated an MC-based SPECT/CT reconstruction within a few minutes, and reconstructed patient images of 177Lu-DOTATATE treatments revealed clearly improved resolution and contrast.

Adaptive scan duration in SPECT: Evaluation for radioembolization

“…The current clinical reconstruction software takes approximately a minute to finish, which is likely sufficient for use in clinical practice. Further speed‐up may, for instance, be achieved by employing GPU‐based software 15 or deep learning approaches 16 …”

Establishment of a clinical SPECT/CT protocol for imaging of 161Tb

“…SPECT acquisitions were reconstructed using three ordered subset expectation maximization (OSEM) reconstruction projection models, namely, (1) attenuation-corrected (A-OSEM), (2) attenuation-corrected and scatter-corrected (AS- OSEM), and (3) attenuation-corrected, scatter-corrected, and detector-collimator response-corrected (ASC-OSEM). The latter was performed by MC simulation using the Sahlgrenska Academy Reconstruction code (SARec) [24]. The triple-energy window (TEW) method was used for scatter correction for EM1, while the dual-energy window (DEW) was used for EM2 in the AS-OSEM reconstruction.…”

“…Further image improvements may be achieved with Monte Carlo-based (MC) reconstruction techniques [22,23]; which inherently incorporates compensation for attenuation, scatter, and detector-collimator response in the forward projection. While the simulation times have been considered to be too long for clinical implementation, we, nevertheless, recently implemented a fast MC-based reconstruction code that generated improved resolution compared to the in-depth resolution corrections method offered by the vendor of the SPECT/CT equipment [24]. This methodology was recently used for the evaluation of bone marrow dosimetry in 177 Lu-DOTATATE treatments [25].…”

“…The SPECT/CT protocols were evaluated in phantom studies. Three different reconstruction methods were compared, namely, attenuation-corrected (AC) ordered subset expectation maximization (A-OSEM) reconstruction, attenuation-and scatter-corrected (SC) OSEM reconstruction (AS-OSEM), and AC, SC, and collimator-detector response (C)-corrected OSEM (ASC-OSEM), using MC-based reconstruction with the Sahlgrenska Academy Reconstruction code (SARec) [24].…”

Can SSTR2 expression in previously resected SI-NETs predict overall survival after PRRT treatment of remaining lesions?

“…In PRRT treated patients (cohort A) an estimation of the uptake of radionuclide was done by measuring the activity concentration in the tumours. In SPECT images acquired 24 h after the first PRRT treatment reconstructions were done with the recently developed Monte Carlo based ordered subset expectation maximization algorithm SARec [19]. Tumours were identified by visual inspection and the three tumours containing the highest maximum voxel values in each patient were chosen for assessment.…”