Scatter Correction Based on GPU-Accelerated Full Monte Carlo Simulation for Brain PET/MRI

MCGPU-PET: An Open-Source Real-Time Monte Carlo PET Simulator

Monte Carlo (MC) simulations are commonly used to model the emission, transmission, and/or detection of radiation in Positron Emission Tomography (PET). In this work, we introduce a new open-source MC software for PET simulation, MCGPU-PET, which has been designed to fully exploit the computing capabilities of modern GPUs to simulate the acquisition of more than 100 million coincidences per second from voxelized sources and material distributions. The new simulator is an extension of the PENELOPE-based MCGPU code previously used in cone-beam CT and mammography applications. We validated the accuracy of the accelerated code by comparing it to GATE and PeneloPET simulations achieving an agreement within 10 percent approximately. As an example application of the code for fast estimation of PET coincidences, a scan of the NEMA IQ phantom was simulated. A fully 3D sinogram with 6382 million true coincidences and 731 million scatter coincidences was generated in 54 seconds in one GPU. MCGPU-PET provides an estimation of true and scatter coincidences and spurious background (for positron-gamma emitters such as 124I) at a rate 3 orders of magnitude faster than CPU-based MC simulators. This significant speed-up enables the use of the code for accurate scatter and prompt-gamma background estimations within an iterative image reconstruction process.

UMC-PET: a fast and flexible Monte Carlo PET simulator

Objective.The GPU-based Ultra-fast Monte Carlo positron emission tomography simulator (UMC-PET) incorporates the physics of the emission, transport and detection of radiation in PET scanners. It includes positron range, non-colinearity, scatter and attenuation, as well as detector response. The objective of this work is to present and validate UMC-PET as a a multi-purpose, accurate, fast and flexible PET simulator.Approach.We compared UMC-PET against PeneloPET, a well-validated MC PET simulator, both in preclinical and clinical scenarios. Different phantoms for scatter fraction (SF) assessment following NEMA protocols were simulated in a 6R-SuperArgus and a Biograph mMR scanner, comparing energy histograms, NEMA SF, and sensitivity for different energy windows. A comparison with real data reported in the literature on the Biograph scanner is also shown.Main results.NEMA SF and sensitivity estimated by UMC-PET where within few percent of PeneloPET predictions. The discrepancies can be attributed to small differences in the physics modeling. Running in a 11 GB GeForce RTX 2080 Ti GPU, UMC-PET is ∼1500 to ∼2000 times faster than PeneloPET executing in a single core Intel(R) Xeon(R) CPU W-2155 @ 3.30 GHz.Significance.UMC-PET employs a voxelized scheme for the scanner, patient adjacent objects (such as shieldings or the patient bed), and the activity distribution. This makes UMC-PET extremely flexible. Its high simulation speed allows applications such as MC scatter correction, faster SRM estimation for complex scanners, or even MC iterative image reconstruction.

PET scatter estimation using deep learning U-Net architecture

Objective.Positron emission tomography (PET) image reconstruction needs to be corrected for scatter in order to produce quantitatively accurate images. Scatter correction is traditionally achieved by incorporating an estimated scatter sinogram into the forward model during image reconstruction. Existing scatter estimated methods compromise between accuracy and computing time. Nowadays scatter estimation is routinely performed using single scatter simulation (SSS), which does not accurately model multiple scatter and scatter from outside the field-of-view, leading to reduced qualitative and quantitative PET reconstructed image accuracy. On the other side, Monte-Carlo (MC) methods provide a high precision, but are computationally expensive and time-consuming, even with recent progress in MC acceleration.Approach.In this work we explore the potential of deep learning (DL) for accurate scatter correction in PET imaging, accounting for all scatter coincidences. We propose a network based on a U-Net convolutional neural network architecture with 5 convolutional layers. The network takes as input the emission and computed tomography (CT)-derived attenuation factor (AF) sinograms and returns the estimated scatter sinogram. The network training was performed using MC simulated PET datasets. Multiple anthropomorphic extended cardiac-torso phantoms of two different regions (lung and pelvis) were created, considering three different body sizes and different levels of statistics. In addition, two patient datasets were used to assess the performance of the method in clinical practice.Main results.Our experiments showed that the accuracy of our method, namely DL-based scatter estimation (DLSE), was independent of the anatomical region (lungs or pelvis). They also showed that the DLSE-corrected images were similar to that reconstructed from scatter-free data and more accurate than SSS-corrected images.Significance.The proposed method is able to estimate scatter sinograms from emission and attenuation data. It has shown a better accuracy than the SSS, while being faster than MC scatter estimation methods.

Fast 3D kernel computation method for positron range correction in PET

Objective. The positron range is a fundamental, detector-independent physical limitation to spatial resolution in positron emission tomography (PET) as it causes a significant blurring of underlying activity distribution in the reconstructed images. A major challenge for positron range correction methods is to provide accurate range kernels that inherently incorporate the generally inhomogeneous stopping power, especially at tissue boundaries. In this work, we propose a novel approach to generate accurate three-dimensional (3D) blurring kernels both in homogenous and heterogeneous media to improve PET spatial resolution.Approach. In the proposed approach, positron energy deposition was approximately tracked along straight paths, depending on the positron stopping power of the underlying material. The positron stopping power was derived from the attenuation coefficient of 511 keV gamma photons according to the available PET attenuation maps. Thus, the history of energy deposition is taken into account within the range of kernels. Special emphasis was placed on facilitating the very fast computation of the positron annihilation probability in each voxel.Results. Positron path distributions of¹⁸F in low-density polyurethane were in high agreement with Geant4 simulation at an annihilation probability larger than 10^-2∼ 10^-3of the maximum annihilation probability. The Geant4 simulation was further validated with measured¹⁸F depth profiles in these polyurethane phantoms. The tissue boundary of water with cortical bone and lung was correctly modeled. Residual artifacts from the numerical computations were in the range of 1%. The calculated annihilation probability in voxels shows an overall difference of less than 20% compared to the Geant4 simulation.Significance. The proposed method is expected to significantly improve spatial resolution for non-standard isotopes by providing sufficiently accurate range kernels, even in the case of significant tissue inhomogeneities.

Feasibility of positron range correction in 82-Rubidium cardiac PET/CT

Background

Myocardial perfusion imaging (MPI) using positron emission tomography (PET) tracers is an essential tool in investigating diseases and treatment responses in cardiology. ⁸²Rubidium (⁸²Rb)-PET imaging is advantageous for MPI due to its short half-life, but cannot be used for small animal research due to the long positron range. We aimed to correct for this, enabling MPI with ⁸²Rb-PET in rats.

Methods

The effect of positron range correction (PRC) on ⁸²Rb-PET was examined using two phantoms and in vivo on rats. A NEMA NU-4-inspired phantom was used for image quality evaluation (%standard deviation (%SD), spillover ratio (SOR) and recovery coefficient (RC)). A cardiac phantom was used for assessing spatial resolution. Two rats underwent rest ⁸²Rb-PET to optimize number of iterations, type of PRC and respiratory gating.

Results

NEMA NU-4 metrics (no PRC vs PRC): %SD 0.087 versus 0.103; SOR (air) 0.022 versus 0.002, SOR (water) 0.059 versus 0.019; RC (3 mm) 0.219 versus 0.584, RC (4 mm) 0.300 versus 0.874, RC (5 mm) 0.357 versus 1.197. Cardiac phantom full width at half maximum (FWHM) and full width at tenth maximum (FWTM) (no PRC vs. PRC): FWTM 6.73 mm versus 3.26 mm (true: 3 mm), FWTM 9.27 mm versus 7.01 mm. The in vivo scans with respiratory gating had a homogeneous myocardium clearly distinguishable from the blood pool.

Conclusion

PRC improved the spatial resolution for the phantoms and in vivo at the expense of slightly more noise. Combined with respiratory gating, the spatial resolution achieved using PRC should allow for quantitative MPI in small animals.

Advances in Preclinical PET

The classical intent of PET imaging is to obtain the most accurate estimate of the amount of positron-emitting radiotracer in the smallest possible volume element located anywhere in the imaging subject at any time using the least amount of radioactivity. Reaching this goal, however, is confounded by an enormous array of interlinked technical issues that limit imaging system performance. As a result, advances in PET, human or animal, are the result of cumulative innovations across each of the component elements of PET, from data acquisition to image analysis. In the report that follows, we trace several of these advances across the imaging process with a focus on small animal PET.

High-throughput, accurate Monte Carlo simulation on CPU hardware for PET applications

Monte Carlo simulations (MCS) represent a fundamental approach to modelling the photon interactions in Positron Emission Tomography (PET). A variety of PET-dedicated MCS tools are available to assist and improve PET imaging applications. Of these, GATE has evolved into one of the most popular software for PET MCS because of its accuracy and flexibility. However, simulations are extremely time-consuming. The use of graphics processing units (GPU) has been proposed as a solution to this, with reported acceleration factors about 400-800. These factors refer to GATE benchmarks performed on a single CPU core. Consequently, CPU-based MCS can also be easily accelerated by one order of magnitude or beyond when exploiting multi-threading on powerful CPUs. Thus, CPU-based implementations become competitive when further optimisations can be achieved. In this context, we have developed a novel, CPU-based software called the PET Physics Simulator (PPS), which combines several efficient methods to significantly boost the performance. PPS flexibly applies GEANT4 cross-sections as a pre-calculated database, thus obtaining results equivalent to GATE. This is demonstrated for an elaborated PET scanner with 3-layer block detectors. All code optimisations yield an acceleration factor of  20 (single core). Multi-threading on a high-end CPU workstation (96 cores) further accelerates the PPS by a factor of 80. This results in a total speed-up factor of 1600, which outperforms comparable GPU-based MCS by a factor of  2. Optionally, the proposed method of coincidence multiplexing can further enhance the throughput by an additonal factor of 15. The combination of all optimisations corresponds to an acceleration factor of  24000. In this way, the PPS can simulate complex PET detector systems with an effective throughput of photon pairs in less than 10 milliseconds.

SimPET-An open online platform for the Monte Carlo simulation of realistic brain PET data. Validation for 18 F-FDG scans

Purpose

SimPET (www.sim‐pet.org) is a free cloud‐based platform for the generation of realistic brain positron emission tomography (PET) data. In this work, we introduce the key features of the platform. In addition, we validate the platform by performing a comparison between simulated healthy brain FDG‐PET images and real healthy subject data for three commercial scanners (GE Advance NXi, GE Discovery ST, and Siemens Biograph mCT).

Methods

The platform provides a graphical user interface to a set of automatic scripts taking care of the code execution for the phantom generation, simulation (SimSET), and tomographic image reconstruction (STIR). We characterize the performance using activity and attenuation maps derived from PET/CT and MRI data of 25 healthy subjects acquired with a GE Discovery ST. We then use the created maps to generate synthetic data for the GE Discovery ST, the GE Advance NXi, and the Siemens Biograph mCT. The validation was carried out by evaluating Bland‐Altman differences between real and simulated images for each scanner. In addition, SPM voxel‐wise comparison was performed to highlight regional differences. Examples for amyloid PET and for the generation of ground‐truth pathological patients are included.

Results

The platform can be efficiently used for generating realistic simulated FDG‐PET images in a reasonable amount of time. The validation showed small differences between SimPET and acquired FDG‐PET images, with errors below 10% for 98.09% (GE Discovery ST), 95.09% (GE Advance NXi), and 91.35% (Siemens Biograph mCT) of the voxels. Nevertheless, our SPM analysis showed significant regional differences between the simulated images and real healthy patients, and thus, the use of the platform for converting control subject databases between different scanners requires further investigation.

Conclusions

The presented platform can potentially allow scientists in clinical and research settings to perform MC simulation experiments without the need for high‐end hardware or advanced computing knowledge and in a reasonable amount of time.