GATE: a simulation toolkit for PET and SPECT

Feasibility of proton dosimetry overriding planning CT with daily CBCT elaborated through generative artificial intelligence tools

Radiotherapy commonly utilizes cone beam computed tomography (CBCT) for patient positioning and treatment monitoring. CBCT is deemed to be secure for patients, making it suitable for the delivery of fractional doses. However, limitations such as a narrow field of view, beam hardening, scattered radiation artifacts, and variability in pixel intensity hinder the direct use of raw CBCT for dose recalculation during treatment. To address this issue, reliable correction techniques are necessary to remove artifacts and remap pixel intensity into Hounsfield Units (HU) values. This study proposes a deep-learning framework for calibrating CBCT images acquired with narrow field of view (FOV) systems and demonstrates its potential use in proton treatment planning updates. Cycle-consistent generative adversarial networks (cGAN) processes raw CBCT to reduce scatter and remap HU. Monte Carlo simulation is used to generate CBCT scans, enabling the possibility to focus solely on the algorithm's ability to reduce artifacts and cupping effects without considering intra-patient longitudinal variability and producing a fair comparison between planning CT (pCT) and calibrated CBCT dosimetry. To showcase the viability of the approach using real-world data, experiments were also conducted using real CBCT. Tests were performed on a publicly available dataset of 40 patients who received ablative radiation therapy for pancreatic cancer. The simulated CBCT calibration led to a difference in proton dosimetry of less than 2%, compared to the planning CT. The potential toxicity effect on the organs at risk decreased from about 50% (uncalibrated) up the 2% (calibrated). The gamma pass rate at 3%/2 mm produced an improvement of about 37% in replicating the prescribed dose before and after calibration (53.78% vs 90.26%). Real data also confirmed this with slightly inferior performances for the same criteria (65.36% vs 87.20%). These results may confirm that generative artificial intelligence brings the use of narrow FOV CBCT scans incrementally closer to clinical translation in proton therapy planning updates.

Deep generative denoising networks enhance quality and accuracy of gated cardiac PET data

BACKGROUND

Cardiac positron emission tomography (PET) can visualize and quantify the molecular and physiological pathways of cardiac function. However, cardiac and respiratory motion can introduce blurring that reduces PET image quality and quantitative accuracy. Dual cardiac- and respiratory-gated PET reconstruction can mitigate motion artifacts but increases noise as only a subset of data are used for each time frame of the cardiac cycle.

AIM

The objective of this study is to create a zero-shot image denoising framework using a conditional generative adversarial networks (cGANs) for improving image quality and quantitative accuracy in non-gated and dual-gated cardiac PET images.

METHODS

Our study included retrospective list-mode data from 40 patients who underwent an 18F-fluorodeoxyglucose (18F-FDG) cardiac PET study. We initially trained and evaluated a 3D cGAN-known as Pix2Pix-on simulated non-gated low-count PET data paired with corresponding full-count target data, and then deployed the model on an unseen test set acquired on the same PET/CT system including both non-gated and dual-gated PET data.

RESULTS

Quantitative analysis demonstrated that the 3D Pix2Pix network architecture achieved significantly (p value<0.05) enhanced image quality and accuracy in both non-gated and gated cardiac PET images. At 5%, 10%, and 15% preserved count statistics, the model increased peak signal-to-noise ratio (PSNR) by 33.7%, 21.2%, and 15.5%, structural similarity index (SSIM) by 7.1%, 3.3%, and 2.2%, and reduced mean absolute error (MAE) by 61.4%, 54.3%, and 49.7%, respectively. When tested on dual-gated PET data, the model consistently reduced noise, irrespective of cardiac/respiratory motion phases, while maintaining image resolution and accuracy. Significant improvements were observed across all gates, including a 34.7% increase in PSNR, a 7.8% improvement in SSIM, and a 60.3% reduction in MAE.

CONCLUSION

The findings of this study indicate that dual-gated cardiac PET images, which often have post-reconstruction artifacts potentially affecting diagnostic performance, can be effectively improved using a generative pre-trained denoising network.

Scatter estimation and correction using time-of-flight and deconvolution in x-ray medical imaging

Objective.Time-of-flight (TOF) scatter rejection requires a total timing jitter, including the detector timing jitter and the x-ray source's pulses width, of 50 ps or less to mitigate most of the effects of scattered photons in radiography and CT imaging. However, since the total contribution of the source and detector to the timing jitter can be retrieved during an acquisition with nothing between the source and detector, it can be demonstrated that this contribution may be partially removed to improve the image quality.Approach.A scatter correction method using iterative deconvolution of the measured time point-spread function estimates the number of scattered photons detected in each pixel. To evaluate the quality of the estimation, GATE was used to simulate the radiography of a water cylinder with bone inserts, and a head and torso in a system with total timing jitters from 100 ps up to 500 ps full-width-at-half-maximum (FWHM).Main results.With a total timing jitter of 200 ps FWHM, 89% of the contrast degradation caused by scattered photons was recovered in a head and torso radiography, compared to 28% with a simple time threshold method. Corrected images using the estimation have a percent root-mean square error between 2% and 14% in both phantoms with timing jitters from 100 to 500 ps FWHM which is lower than the error achieved with scatter rejection alone at 100 ps FWHM.Significance.TOF x-ray imaging has the potential to mitigate the effects of the scattering contribution and offers an alternative to anti-scatter grids that avoids loss of primary photons. Compare to simple TOF scatter rejection using only a threshold, the deconvolution estimation approach has lower requirements on both the source and detector. These requirements are now within reach of state-of-the-art systems.

3D small-scale dosimetry and tumor control of 225Ac radiopharmaceuticals for prostate cancer

Radiopharmaceutical therapy using α -emitting225 Ac is an emerging treatment for patients with advanced metastatic cancers. Measurement of the spatial dose distribution in organs and tumors is needed to inform treatment dose prescription and reduce off-target toxicity, at not only organ but also sub-organ scales. Digital autoradiography with α -sensitive detection devices can measure radioactivity distributions at 20-40 μ m resolution, but anatomical characterization is typically limited to 2D. We collected digital autoradiographs across whole tissues to generate 3D dose volumes and used them to evaluate the simultaneous tumor control and regional kidney dosimetry of a novel therapeutic radiopharmaceutical for prostate cancer, [225Ac]Ac-Macropa-PEG4-YS5, in mice. 22Rv1 xenograft-bearing mice treated with 18.5 kBq of [225Ac]Ac-Macropa-PEG4-YS5 were sacrificed at 24 h and 168 h post-injection for quantitative α -particle digital autoradiography and hematoxylin and eosin staining. Gamma-ray spectroscopy of biodistribution data was used to determine temporal dynamics and213 Bi redistribution. Tumor control probability and sub-kidney dosimetry were assessed. Heterogeneous225 Ac spatial distribution was observed in both tumors and kidneys. Tumor control was maintained despite heterogeneity if cold spots coincided with necrotic regions.225 Ac dose-rate was highest in the cortex and renal vasculature. Extrapolation of tumor control suggested that kidney absorbed dose could be reduced by 41% while maintaining 90% TCP. The 3D dosimetry methods described allow for whole tumor and organ dose measurements following225 Ac radiopharmaceutical therapy, which correlate to tumor control and toxicity outcomes.

Design and proof of concept of a double-panel TOF-PET system

OBJECTIVE

Positron Emission Tomography (PET) is a well-known imaging technology for the diagnosis, treatment, and monitoring of several diseases. Most PET scanners use a Ring-Shaped Detector Configuration (RSDC), which helps obtain homogeneous image quality but are restricted to an invariable Field-of-View (FOV), scarce spatial resolution, and low sensitivity. Alternatively, few PET systems use Open Detector Configurations (ODC) to permit an accessible FOV adaptable to different target sizes, thus optimizing sensitivity. Yet, to compensate the lack of angular coverage in ODC-PET, developing a detector with high-timing performance is mandatory to enable Time-of-Flight (TOF) techniques during reconstruction. The main goal of this work is to provide a proof of concept PET scanner appropriate for constructing the new generation of ODC-PET suitable for biopsy guidance and clinical intervention during acquisition. The designed detector has to be compact and robust, and its requirements in terms of performance are spatial and time resolutions < 2 mm and < 200 ps, respectively.

METHODS

The present work includes a simulation study of an ODC-PET based on 2-panels with variable distance. The image quality (IQ) and Derenzo phantoms have been simulated and evaluated. The phantom simulations have also been performed using a ring-shaped PET for comparison purposes of the ODC approach with conventional systems. Then, an experimental evaluation of a prototype detector that has been designed following the simulation results is presented. This study focused on tuning the ASIC parameters and evaluating the scintillator surface treatment (ESR and TiO2), and configuration that yields the best Coincidence Time Resolution (CTR). Moreover, the scalability of the prototype to a module of 64 × 64mm2 and its preliminary evaluation regarding pixel identification are provided.

RESULTS

The simulation results reported sensitivity (%) values at the center of the FOV of 1.96, 1.63, and 1.18 for panel distances of 200, 250, and 300 mm, respectively. The IQ reconstructed image reported good uniformity (87%) and optimal CRC values, and the Derenzo phantom reconstruction suggests a system resolution of 1.6-2 mm. The experimental results demonstrate that using TiO2 coating yielded better detector performance than ESR. Acquired data was filtered by applying an energy window of ± 30% at the photopeak level. After filtering, best CTR of 230 ± 2 ps was achieved for an 8 × 8 LYSO pixel block with 2 × 2 × 12mm3 each. The detector performance remained constant after scaling-up the prototype to a module of 64 × 64mm2, and the flood map demonstrates the module's capabilities to distinguish the small pixels; thus, a spatial resolution < 2 mm (pixel size) is achieved.

CONCLUSIONS

The simulated results of this biplanar scanner show high performance in terms of image quality and sensitivity. These results are comparable to state-of-the-art PET technology and, demonstrate that including TOF information minimizes the image artifacts due to the lack of angular projections. The experimental results concluded that using TiO2 coating provide the best performance. The results suggest that this scanner may be suitable for organ study, breast, prostate, or cardiac applications, with good uniformity and CRC.

Deep learning applications for quantitative and qualitative PET in PET/MR: technical and clinical unmet needs

We aim to provide an overview of technical and clinical unmet needs in deep learning (DL) applications for quantitative and qualitative PET in PET/MR, with a focus on attenuation correction, image enhancement, motion correction, kinetic modeling, and simulated data generation. (1) DL-based attenuation correction (DLAC) remains an area of limited exploration for pediatric whole-body PET/MR and lung-specific DLAC due to data shortages and technical limitations. (2) DL-based image enhancement approximating MR-guided regularized reconstruction with a high-resolution MR prior has shown promise in enhancing PET image quality. However, its clinical value has not been thoroughly evaluated across various radiotracers, and applications outside the head may pose challenges due to motion artifacts. (3) Robust training for DL-based motion correction requires pairs of motion-corrupted and motion-corrected PET/MR data. However, these pairs are rare. (4) DL-based approaches can address the limitations of dynamic PET, such as long scan durations that may cause patient discomfort and motion, providing new research opportunities. (5) Monte-Carlo simulations using anthropomorphic digital phantoms can provide extensive datasets to address the shortage of clinical data. This summary of technical/clinical challenges and potential solutions may provide research opportunities for the research community towards the clinical translation of DL solutions.

Self-normalization for a 1 mm3resolution clinical PET system using deep learning

Objective.This work proposes, for the first time, an image-based end-to-end self-normalization framework for positron emission tomography (PET) using conditional generative adversarial networks (cGANs).Approach.We evaluated different approaches by exploring each of the following three methodologies. First, we used images that were either unnormalized or corrected for geometric factors, which encompass all time-invariant factors, as input data types. Second, we set the input tensor shape as either a single axial slice (2D) or three contiguous axial slices (2.5D). Third, we chose either Pix2Pix or polarized self-attention (PSA) Pix2Pix, which we developed for this work, as a deep learning network. The targets for all approaches were the axial slices of images normalized using the direct normalization method. We performed Monte Carlo simulations of ten voxelized phantoms with the SimSET simulation tool and produced 26,000 pairs of axial image slices for training and testing.Main results.The results showed that 2.5D PSA Pix2Pix trained with geometric-factors-corrected input images achieved the best performance among all the methods we tested. All approaches improved general image quality figures of merit peak signal to noise ratio (PSNR) and structural similarity index (SSIM) from ∼15 % to ∼55 %, and 2.5D PSA Pix2Pix showed the highest PSNR (28.074) and SSIM (0.921). Lesion detectability, measured with region of interest (ROI) PSNR, SSIM, normalized contrast recovery coefficient, and contrast-to-noise ratio, was generally improved for all approaches, and 2.5D PSA Pix2Pix trained with geometric-factors-corrected input images achieved the highest ROI PSNR (28.920) and SSIM (0.973).Significance.This study demonstrates the potential of an image-based end-to-end self-normalization framework using cGANs for improving PET image quality and lesion detectability without the need for separate normalization scans.

FAST (fast analytical simulator of tracer)-PET: an accurate and efficient PET analytical simulation tool

Objective.Simulation of positron emission tomography (PET) images is an essential tool in the development and validation of quantitative imaging workflows and advanced image processing pipelines. Existing Monte Carlo or analytical PET simulators often compromise on either efficiency or accuracy. We aim to develop and validate fast analytical simulator of tracer (FAST)-PET, a novel analytical framework, to simulate PET images accurately and efficiently.Approach. FAST-PET simulates PET images by performing precise forward projection, scatter, and random estimation that match the scanner geometry and statistics. Although the same process should be applicable to other scanner models, we focus on the Siemens Biograph Vision-600 in this work. Calibration and validation of FAST-PET were performed through comparison with an experimental scan of a National Electrical Manufacturers Association (NEMA) Image Quality (IQ) phantom. Further validation was conducted between FAST-PET and Geant4 Application for Tomographic Emission (GATE) quantitatively in clinical image simulations in terms of intensity-based and texture-based features and task-based tumor segmentation.Main results.According to the NEMA IQ phantom simulation, FAST-PET's simulated images exhibited partial volume effects and noise levels comparable to experimental images, with a relative bias of the recovery coefficient RC within 10% for all spheres and a coefficient of variation for the background region within 6% across various acquisition times. FAST-PET generated clinical PET images exhibit high quantitative accuracy and texture comparable to GATE (correlation coefficients of all features over 0.95) but with ∼100-fold lower computation time. The tumor segmentation masks comparison between both methods exhibited significant overlap and shape similarity with high concordance CCC > 0.97 across measures.Significance.FAST-PET generated PET images with high quantitative accuracy comparable to GATE, making it ideal for applications requiring extensive PET image simulations such as virtual imaging trials, and the development and validation of image processing pipelines.

Dose rate conversion coefficients for ocular contamination in nuclear medicine: A Monte Carlo simulation with experimental validation

BACKGROUND

Since 2011, the International Commission on Radiological Protection (ICRP) has recommended an annual eye lens dose limit of 20 mSv for radiation workers, averaged over 5 years, with no year exceeding 50 mSv. However, limited research has been conducted on dose rate conversion coefficients (DCCs) for direct contamination of the eye.

PURPOSE

This study aimed to accurately determine DCCs for the eye lens and cornea for ocular contamination with radionuclides used in nuclear medicine.

METHODS

DCCs for 37 radionuclides used in nuclear medicine were determined using two different methods. Method 1 involved conducting Monte Carlo (MC) simulations of an ICRU cylinder to determine the absorbed dose at a depth of 3 mm resulting from a point source. The accuracy of this simulation approach was validated by experimental thermoluminescent dosimeter (TLD) measurements for 18F, 68Ga, 99mTc, and 177Lu. In method 2, average DCCs were calculated for the eye lens (complete and radiosensitive parts) and the cornea for both a point source and thin surface contamination centered on the cornea using MC simulations on the adult mesh-type reference computational phantom of the eye from the ICRP (MRCP).

RESULTS

DCCs determined from TLD measurements showed excellent agreement (deviations: +1.4%, +4.7%, -3.1%, and -2.5% for 18F, 68Ga, 99mTc, and 177Lu, respectively) compared to MC simulations of the experimental set-up. For the 37 radionuclides, DCCs of the complete eye-lens for a point source ranged from 2.53 × 10-7 to 4.15 × 10-2 mGy MBq-1 s-1 for the adult MRCPs, being substantially smaller compared to DCCs determined via MC simulations of a ICRU cylinder. In general, point source and surface contamination showed comparable DCCs for the eye lens. Radionuclides emitting low-energy beta radiation or conversion electrons (e.g., 177Lu, 99mTc) showed low DCCs as the radiation does not penetrate to the depth of the eye lens, while radionuclides emitting high-energy beta radiation (e.g., 90Y) showed high DCCs. Overall, DCCs for the radiosensitive part of the eye lens were larger (up to a factor of 3) compared to the complete eye lens. DCCs for the cornea were larger than for the eye lens with a factor that strongly depended on the emitted radiation type. Especially alpha emitters (e.g., 211At, 223Ra) showed high DCCs for the cornea because of the short range of alpha radiation, leading to local maxima in the cornea and not reaching the eye lens.

CONCLUSION

DCCs at a depth of 3 mm in an ICRU cylinder and adult MRCP DCCs for both the complete and sensitive parts of the eye lens and cornea were determined for 37 radionuclides having applications in nuclear medicine. These DCCs are highly useful in radiation safety assessments and radiation dose calculations in ocular contamination incidents.

To Reconstruct or Discard: A Comparison of Additive and Subtractive Charge Sharing Correction Algorithms at High and Low X-ray Fluxes

Effective X-ray photon-counting spectral imaging (x-CSI) detector design involves the optimisation of a wide range of parameters both regarding the sensor (e.g., material, thickness and pixel pitch) and electronics (e.g., signal-processing chain and count-triggering scheme). Our previous publications have looked at the role of pixel pitch, sensor thickness and a range of additive charge sharing correction algorithms (CSCAs), and in this work, we compare additive and subtractive CSCAs to identify the advantages and disadvantages. These CSCAs differ in their approach to dealing with charge sharing: additive approaches attempt to reconstruct the original event, whilst subtractive approaches discard the shared events. Each approach was simulated on data from a wide range of x-CSI detector designs (pixel pitches 100-600 µm, sensor thickness 1.5 mm) and X-ray fluxes (106-109 photons mm-2 s-1), and their performance was characterised in terms of absolute detection efficiency (ADE), absolute photopeak efficiency (APE), relative coincidence counts (RCC) and binned spectral efficiency (BSE). Differences between the two approaches were explained mechanistically in terms of the CSCA's effect on both charge sharing and pule pileup. At low X-ray fluxes, the two approaches perform similarly, but at higher fluxes, they differ in complex ways. Generally, additive CSCAs perform better on absolute metrics (ADE and APE), and subtractive CSCAs perform better on relative metrics (RCC and BSE). Which approach to use will, thus, depend on the expected operating flux and whether dose efficiency or spectral efficiency is more important for the application in mind.

Characterization of accurate 3D collimator-detector response function for single- and multi-lofthole collimated SPECT cameras

PURPOSE

Collimator-detector response function (CDRF) of a SPECT scanner refers to the image generated from a point source of activity. This research aims to characterize the CDRF of a breast-dedicated SPECT imager equipped with a lofthole collimator using GATE Monte Carlo simulation.

MATERIALS AND METHODS

To do so, a cylindrical multi-lofthole collimation system with lofthole apertures dedicated to breast imaging was modeled using the GATE Monte Carlo simulator. The dependency of the CDRF on the source-to-collimator distance of a single-lofthole as well as 8-lofthole collimations was assessed and then compared. In addition, the 3D-sensitivity map of the 8-lofthole collimation was derived. Finally, fair comparisons were conducted between the response of the 8-lofthole collimator and that of an 8-pinhole and also existing analytical derivations. In all cases, a data acquisition period of 5.0 min with an in-air 99mTc point source was considered.

RESULTS

For the single-lofthole collimator, 4.5 times increasing the magnification factor leads to a 16- and twofold improvement in the sensitivity and spatial resolution, respectively. In the single-lofthole collimator, the resolution and sensitivity are degraded as the source-to-aperture distance increases. For the cylindrical 8-lofthole collimator, the findings confirm that CDRF strongly depends on source-to-aperture distance and angle of photon incidence. For a 30 mm in-plane offset point, a 25% increase in sensitivity is observed compared to that of the center of the FOV. Increasing the angle from 0∘ to 34∘ results in a 50% reduction in sensitivity. Furthermore, the findings illustrate that spatial resolution follows a quadratic function as10 - 3 d 2 + 2 × 10 - 4 d + R 0 where d is an offset along the x-, y-, and z-axis, and R0 is the spatial resolution at the center of the FOV.

CONCLUSION

In conclusion, both spatial resolution and sensitivity of the lofthole collimation are considerably angle- and offset-dependent within the FOV of single- and multi-lofthole collimated SPECT imagers.

Evaluating the performance of thermoplastic 3D bolus used in radiation therapy

A 3D-printed bolus is being developed to deliver accurate doses to superficial cancers. In this study, flexible thermoplastic filaments, specifically PLA, TPU, PETG, and HIPS, were fabricated into boluses and then compared to commercial bolus for the variation of the dose elevation region of photon beams. The experimental results indicate that the maximum dose depth is similar, and the consistent trend of the percentage depth dose confirms the potential usage as a build-up bolus.

The effects of intra-detector Compton scatter on low-frequency DQE for photon-counting CT using edge-on-irradiated silicon detectors

BACKGROUND

Edge-on-irradiated silicon detectors are currently being investigated for use in full-body photon-counting computed tomography (CT) applications. The low atomic number of silicon leads to a significant number of incident photons being Compton scattered in the detector, depositing a part of their energy and potentially being counted multiple times. Even though the physics of Compton scatter is well established, the effects of Compton interactions in the detector on image quality for an edge-on-irradiated silicon detector have still not been thoroughly investigated.

PURPOSE

To investigate and explain effects of Compton scatter on low-frequency detective quantum efficiency (DQE) for photon-counting CT using edge-on-irradiated silicon detectors.

METHODS

We extend an existing Monte Carlo model of an edge-on-irradiated silicon detector with 60 mm active absorption depth, previously used to evaluate spatial-frequency-based performance, to develop projection and image domain performance metrics for pure density and pure spectral imaging tasks with 30 and 40 cm water backgrounds. We show that the lowest energy threshold of the detector can be used as an effective discriminator of primary counts and cross-talk caused by Compton scatter. We study the developed metrics as functions of the lowest threshold energy for root-mean-square electronic noise levels of 0.8, 1.6, and 3.2 keV, where the intermediate level 1.6 keV corresponds to the noise level previously measured on a single sensor element in isolation. We also compare the performance of a modeled detector with 8, 4, and 2 optimized energy bins to a detector with 1-keV-wide bins.

RESULTS

In terms of low-frequency DQE for density imaging, there is a tradeoff between using a threshold low enough to capture Compton interactions and avoiding electronic noise counts. For 30 cm water phantom, 4 energy bins, and a root-mean-square electronic noise of 0.8, 1.6, and 3.2 keV, it is optimal to put the lowest energy threshold at 3, 6, and 1 keV, which gives optimal projection-domain DQEs of 0.64, 0.59, and 0.52, respectively. Low-frequency DQE for spectral imaging also benefits from measuring Compton interactions with respective optimal thresholds of 12, 12, and 13 keV. No large dependence on background thickness was observed. For the intermediate noise level (1.6 keV), increasing the lowest threshold from 5 to 35 keV increases the variance in a iodine basis image by 60%-62% (30 cm phantom) and 67%-69% (40 cm phantom), with 8 bins. Both spectral and density DQE are adversely affected by increasing the electronic noise level. Image-domain DQE exhibits similar qualitative behavior as projection-domain DQE.

CONCLUSIONS

Compton interactions contribute significantly to the density imaging performance of edge-on-irradiated silicon detectors. With the studied detector topology, the benefit of counting primary Compton interactions outweighs the penalty of multiple counting at all lowest threshold energies. Compton interactions also contribute significantly to the spectral imaging performance for measured energies above 10 keV.

Design and implementation of continuous bed motion (CBM) in Xtrim preclinical PET scanner for whole-body Imaging: MC simulation and experimental measurements

PURPOSE

Preclinical PET scanners often have limited axial field-of-view for whole-body (WB) scanning of the small-animal. Step-and-shoot(S&S) acquisition mode requires multiple bed positions (BPs) to cover the scan length. Alternatively, in Continuous Bed Motion(CBM) mode, data acquisition is performed while the bed is continuously moving. In this study, to reduce acquisition time and enhance image quality, the CBM acquisition protocol was optimized and implemented on the Xtrim-PET preclinical scanner for WB imaging.

METHODS

The over-scan percentage(OS%) in CBM mode was optimized by Monte Carlo simulation. Bed movement speed was optimized considering ranges from 0.1 to 2.0 mm s-1, and absolute system sensitivities with the optimal OS% were calculated. The performance of the scanner in CBM mode was measured, and compared with S&S mode based on the NEMA-NU4 standard.

RESULTS

The optimal trade-off between absolute sensitivity and uniformity of sensitivity profile was achieved at OS-50 %. In comparison to S&S mode with maximum ring differences (MRD) of 9 and 23, the calculated equivalent speeds in CBM(OS-50 %) mode were 0.3 and 0.14 mm s-1, respectively. In terms of data acquisition with equal sensitivity in both CBM(OS-50 %) and S&S(MRD-9) modes, the total scan time in CBM mode decreased by 25.9 %, 47.7 %, 54.7 %, and 58.2 % for scan lengths of 1 to 4 BPs, respectively.

CONCLUSION

The CBM mode enhances WB PET scans for small-animals, offering rapid data acquisition, high system sensitivity, and uniform axial sensitivity, leading to improved image quality. Its efficiency and customizable scan length and bed speed make it a superior alternative.

NaI gamma camera performance for high energies: Effects of crystal thickness, photomultiplier tube geometry and light guide thickness

BACKGROUND

Gamma camera imaging, including single photon emission computed tomography (SPECT), is crucial for research, diagnostics, and radionuclide therapy. Gamma cameras are predominantly based on arrays of photon multipliers tubes (PMTs) that read out NaI(Tl) scintillation crystals. In this way, standard gamma cameras can localize ɣ-rays with energies typically ranging from 30 to 360 keV. In the last decade, there has been an increasing interest towards gamma imaging outside this conventional clinical energy range, for example, for theragnostic applications and preclinical multi-isotope positron emission tomography (PET) and PET-SPECT. However, standard gamma cameras are typically equipped with 9.5 mm thick NaI(Tl) crystals which can result in limited sensitivity for these higher energies.

PURPOSE

Here we investigate to what extent thicker scintillators can improve the photopeak sensitivity for higher energy isotopes while attempting to maintain spatial resolution.

METHODS

Using Monte Carlo simulations, we analyzed multiple PMT-based configurations of gamma detectors with monolithic NaI (Tl) crystals of 20 and 40 mm thickness. Optimized light guide thickness together with 2-inch round, 3-inch round, 60 × 60 mm2 square, and 76 × 76 mm2 square PMTs were tested. For each setup, we assessed photopeak sensitivity, energy resolution, spatial, and depth-of-interaction (DoI) resolution for conventional (140 keV) and high (511 keV) energy ɣ using a maximum-likelihood algorithm. These metrics were compared to those of a "standard" 9.5 mm-thick crystal detector with 3-inch round PMTs.

RESULTS

Estimated photopeak sensitivities for 511 keV were 27% and 53% for 20 and 40 mm thick scintillators, which is respectively, 2.2 and 4.4 times higher than for 9.5 mm thickness. In most cases, energy resolution benefits from using square PMTs instead of round ones, regardless of their size. Lateral and DoI spatial resolution are best for smaller PMTs (2-inch round and 60 × 60 mm2 square) which outperform the more cost-effective larger PMT setups (3-inch round and 76 × 76 mm2 square), while PMT layout and shape have negligible (< 10%) effect on resolution. Best spatial resolution was obtained with 60 × 60 mm2 PMTs; for 140 keV, lateral resolution was 3.5 mm irrespective of scintillator thickness, improving to 2.8 and 2.9 mm for 511 keV with 20 and 40 mm thick crystals, respectively. Using the 3-inch round PMTs, lateral resolutions of 4.5 and 3.9 mm for 140 keV and of 3.5 and 3.7 mm for 511 keV were obtained with 20 and 40 mm thick crystals respectively, indicating a moderate performance degradation compared to the 3.5 and 2.9 mm resolution obtained by the standard detector for 140 and 511 keV. Additionally, DoI resolution for 511 keV was 7.0 and 5.6 mm with 20 and 40 mm crystals using 60 × 60 mm2 square PMTs, while with 3-inch round PMTs 12.1 and 5.9 mm were obtained.

CONCLUSION

Depending on PMT size and shape, the use of thicker scintillator crystals can substantially improve detector sensitivity at high gamma energies, while spatial resolution is slightly improved or mildly degraded compared to standard crystals.

Cloud-based serverless computing enables accelerated monte carlo simulations for nuclear medicine imaging

Objective.This study investigates the potential of cloud-based serverless computing to accelerate Monte Carlo (MC) simulations for nuclear medicine imaging tasks. MC simulations can pose a high computational burden-even when executed on modern multi-core computing servers. Cloud computing allows simulation tasks to be highly parallelized and considerably accelerated.Approach.We investigate the computational performance of a cloud-based serverless MC simulation of radioactive decays for positron emission tomography imaging using Amazon Web Service (AWS) Lambda serverless computing platform for the first time in scientific literature. We provide a comparison of the computational performance of AWS to a modern on-premises multi-thread reconstruction server by measuring the execution times of the processes using between105and2·1010simulated decays. We deployed two popular MC simulation frameworks-SimSET and GATE-within the AWS computing environment. Containerized application images were used as a basis for an AWS Lambda function, and local (non-cloud) scripts were used to orchestrate the deployment of simulations. The task was broken down into smaller parallel runs, and launched on concurrently running AWS Lambda instances, and the results were postprocessed and downloaded via the Simple Storage Service.Main results.Our implementation of cloud-based MC simulations with SimSET outperforms local server-based computations by more than an order of magnitude. However, the GATE implementation creates more and larger output file sizes and reveals that the internet connection speed can become the primary bottleneck for data transfers. Simulating 109decays using SimSET is possible within 5 min and accrues computation costs of about $10 on AWS, whereas GATE would have to run in batches for more than 100 min at considerably higher costs.Significance.Adopting cloud-based serverless computing architecture in medical imaging research facilities can considerably improve processing times and overall workflow efficiency, with future research exploring additional enhancements through optimized configurations and computational methods.

Patient-specific radiation dose for Chinese pediatric patients undergoing whole-body PET/CT examinations

Objective.To assess potential variations in the absorbed dose between Chinese and Caucasian children exposed to18F-FDG PET scan and to investigate the factors contributing to dose differences, this work employed patient-specific phantoms and our compartment model for calculating the patient-specific absorbed dose in Chinese children.Approach.Data of 29 Chinese pediatric patients undergoing whole-body18F-FDG PET/CT studies were retrospectively collected, including PET images for activity distributions and corresponding CT images for organ segmentation and phantom construction. A biokinetic compartment model was implemented to obtain cumulated activities. Absorbed radiation dose for both CT and PET component were calculated using Monte Carlo simulations. Regression models were fitted to time integrated activity coefficient (TIAC) and organ absorbed dose for each patient.Main results.TIACs of all the organs in our compartment model and the organ dose for 12 organs were correlated with patients' weight. Young children have significantly large uptake in brain compared to adults. The distinctions of anatomical and biological characteristics between Chinese and Caucasian children contribute to variations in the absorbed dose of18F-FDG PET scans. PET contributed more in organ dose than CT did in most organs, especially in brain and bladder. The average effective dose (± SD) was 4.5 mSv (± 1.12 mSv), 7.8 mSv (± 3.2 mSv) and 12.3 mSv (± 3.5 mSv) from CT, PET and their sum respectively. PET contributed 1.7 times higher than CT.Significance.To the best of our knowledge, this work represents the first attempt to estimate patient-specific radiation doses from PET/CT for Chinese pediatric patients. TIACs derived from our methodology in both age groups exhibited significant differences from the that reported in ICRP 128. Substantial differences in absorbed and effective doses were observed between Chinese and Caucasian children across all age groups. These disparities are attributed to markedly distinct anatomical and pharmacokinetic characteristics among adults and pediatric patients, and different racial groups. The application of data derived from adults to pediatric patients introduces considerable uncertainty. Our methodology offers a valuable approach not only for estimating pharmacokinetic characteristics and patient-specific radiation doses in pediatric patients undergoing18F-FDG studies but also for other cohorts with similar characteristics.

Quantifying partial volume effect in SPECT and planar imaging: optimizing region of interest for activity concentration estimation in different sphere sizes

INTRODUCTION

To quantify the partial volume effect in single photon emission tomography (SPECT) and planar images of Carlson phantom as well as providing an optimum region of interest (ROI) required to more accurately estimate the activity concentration for different sphere sizes.

METHODS

131 I solution with the 161.16 kBq/ml concentration was uniformly filled into the different spheres of Carlson phantom (cold background condition) with the diameters of 7.3, 9.2, 11.4, 14.3, 17.9, 22.4 and 29.9 mm, and there was no background activity. In the hot background condition, the spheres were filled with the solution of 131 I with the 1276.5 kBq/ml addition to the background activity concentration of 161.16 kBq/ml in all the phantoms. The spheres were mounted inside the phantom and underwent SPECT and planar images. ROI was drawn closely on the boundary of each sphere image and it was extended to extract the true count.

RESULTS

In the cold background condition, the recovery coefficient (RC) value for SPECT images ranged between 0.8 and 1.03. However, in planar imaging, the RC value was 0.72 for the smallest sphere size and it increased for larger spheres until 0.98 for 29.9 mm. In the hot background condition, the RC value for sphere diameters larger than 20 mm was overestimated more than in the cold background condition. The ROI/size required to more accurately determine activity concentration for the cold background ranged from 1.18 to 2.7. However, in the hot background condition, this ratio varied from 1.34 to 4.05.

CONCLUSION

In the quantification of partial volume effects, the spill-out effect seems to play a crucial role in the distribution of the image counts beyond the boundaries of the image pixels. However, more investigations are needed to accurately characterize limitations regarding the object size, background levels, and other factors.

Depth-encoding using optical photon TOF in a prism-PET detector with tapered crystals

BACKGROUND

High-resolution brain positron emission tomography (PET) scanner is emerging as a significant and transformative non-invasive neuroimaging tool to advance neuroscience research as well as improve diagnosis and treatment in neurology and psychiatry. Time-of-flight (TOF) and depth-of-interaction (DOI) information provide markedly higher PET imaging performance by increasing image signal-to-noise ratio and mitigating spatial resolution degradation due to parallax error, respectively. PET detector modules that utilize light sharing can inherently carry DOI information from the multiple timestamps that are generated per gamma event. The difference between two timestamps that are triggered by scintillation photons traveling in opposite directions signifies the event's depth-dependent optical photon TOF (oTOF). However, light leak at the crystal-readout interface substantially degrades the resolution of this oTOF-based depth encoding.

PURPOSE

We demonstrate the feasibility of oTOF-based depth encoding by mitigating light leak in single-ended-readout Prism-PET detector modules using tapered crystals. Minimizing light leak also improved both energy-based DOI and coincidence timing resolutions.

METHODS

The tapered Prism-PET module consists of a 16  × $\times$  16 array of 1.5  × $\times$  1.5  × $\times$  20 mm 3 ${\rm {mm}}^3$ lutetium yttrium oxyorthosillicate (LYSO) crystals, which are tapered down to 1.2  × $\times$  1.2 mm 2 ${\rm {mm}}^2$ at the crystal-readout interface. The LYSO array couples 4-to-1 to an 8  × $\times$  8 array of 3  × $\times$  3 mm 2 ${\rm {mm}}^2$ silicon photomultiplier (SiPM) pixels on the tapered end and to a segmented prismatoid light guide array on the opposite end. Performance of tapered and non-tapered Prism-PET detectors was experimentally characterized and evaluated by measuring flood histogram, energy resolution, energy-, and oTOF-based DOI resolutions, and coincidence timing resolution. Sensitivities of scanners using different Prism-PET detector designs were simulated using Geant4 application for tomographic emission (GATE).

RESULTS

For the tapered (non-tapered) Prism-PET module, the measured full width at half maximum (FWHM) energy, timing, energy-based DOI, and oTOF-based DOI resolutions were 8.88 (11.18)%, 243 (286) ps, 2.35 (3.18) mm, and 5.42 (13.87) mm, respectively. The scanner sensitivities using non-tapered and tapered crystals, and 10 rings of detector modules, were simulated to be 30.9 and 29.5 kcps/MBq, respectively.

CONCLUSIONS

The tapered Prism-PET module with minimized light leak enabled the first experimental report of oTOF-based depth encoding at the detector module level. It also enabled the utilization of thinner (i.e., 0.1 mm) inter-crystal spacing with barium sulfate as the reflector while also improving energy-based DOI and timing resolutions.

GPU optimization techniques to accelerate optiGAN-a particle simulation GAN

The demand for specialized hardware to train AI models has increased in tandem with the increase in the model complexity over the recent years. Graphics processing unit (GPU) is one such hardware that is capable of parallelizing operations performed on a large chunk of data. Companies like Nvidia, AMD, and Google have been constantly scaling-up the hardware performance as fast as they can. Nevertheless, there is still a gap between the required processing power and processing capacity of the hardware. To increase the hardware utilization, the software has to be optimized too. In this paper, we present some general GPU optimization techniques we used to efficiently train the optiGAN model, a Generative Adversarial Network that is capable of generating multidimensional probability distributions of optical photons at the photodetector face in radiation detectors, on an 8GB Nvidia Quadro RTX 4000 GPU. We analyze and compare the performances of all the optimizations based on the execution time and the memory consumed using the Nvidia Nsight Systems profiler tool. The optimizations gave approximately a 4.5x increase in the runtime performance when compared to a naive training on the GPU, without compromising the model performance. Finally we discuss optiGANs future work and how we are planning to scale the model on GPUs.

Virtual cylindrical PET for efficient DOI image reconstruction with sub-millimetre resolution

Objective.Image reconstruction in high resolution, narrow bore PET scanners with depth of interaction (DOI) capability presents a substantial computational challenge due to the very high sampling in detector and image space. The aim of this study is to evaluate the use of a virtual cylinder in reducing the number of lines of response (LOR) for DOI-based reconstruction in high resolution PET systems while maintaining uniform sub-millimetre spatial resolution.Approach.Virtual geometry was investigated using the awake animal mousePET as a high resolution test case. Using GEANT4 Application for Tomographic Emission (GATE), we simulated the physical scanner and three virtual cylinder implementations with detector size 0.74 mm, 0.47 mm and 0.36 mm (vPET1, vPET2 and vPET3, respectively). The virtual cylinder condenses physical LORs stemming from various crystal pairs and DOI combinations, and which intersect a single virtual detector pair, into a single virtual LOR. Quantitative comparisons of the point spread function (PSF) at various positions within the field of view (FOV) were compared for reconstructions based on the vPET implementations and the physical scanner. We also assessed the impact of the anisotropic PSFs by reconstructing images of a micro Derenzo phantom.Main results.All virtual cylinder implementations achieved LOR data compression of at least 50% for DOI PET reconstruction. PSF anisotropy in radial and tangential profiles was chiefly influenced by DOI resolution and only marginally by virtual detector size. Spatial degradation introduced by virtual cylinders was most prominent in the axial profile. All virtual cylinders achieved sub-millimetre volumetric resolution across the FOV when 6-bin DOI reconstructions (3.3 mm DOI resolution) were performed. Using vPET2 with 6 DOI bins yielded nearly identical reconstructions to the non-virtual case in the transaxial plane, with an LOR compression factor of 86%. Resolution modelling significantly reduced the effects of the asymmetric PSF arising from the non-cylindrical geometry of mousePET.Significance.Narrow bore and high resolution PET scanners require detectors with DOI capability, leading to computationally demanding reconstructions due to the large number of LORs. In this study, we show that DOI PET reconstruction with 50%-86% LOR compression is possible using virtual cylinders while maintaining sub-millimetre spatial resolution throughout the FOV. The methodology and analysis can be extended to other scanners with DOI capability intended for high resolution PET imaging.

Anatomy-guided multi-resolution image reconstruction in PET

Objective. In this paper, we propose positron emission tomography image reconstruction using a multi-resolution triangular mesh. The mesh can be adapted based on patient specific anatomical information that can be in the form of a computed tomography or magnetic resonance imaging image in the hybrid imaging systems. The triangular mesh can be adapted to high resolution in localized anatomical regions of interest (ROI) and made coarser in other regions, leading to an imaging model with high resolution in the ROI with clearly reduced number of degrees of freedom compared to a conventional uniformly dense imaging model.Approach.We compare maximum likelihood expectation maximization reconstructions with the multi-resolution model to reconstructions using a uniformly dense mesh, a sparse mesh and regular rectangular pixel mesh. Two simulated cases are used in the comparison, with the first one using the NEMA image quality phantom and the second the XCAT human phantom.Main results.When compared to the results with the uniform imaging models, the locally refined multi-resolution mesh retains the accuracy of the dense mesh reconstruction in the ROI while being faster to compute than the reconstructions with the uniformly dense mesh. The locally dense multi-resolution model leads also to more accurate reconstruction than the pixel-based mesh or the sparse triangular mesh.Significance.The findings suggest that triangular multi-resolution mesh, which can be made patient and application specific, is a potential alternative for pixel-based reconstruction.

Extension of the cone-beam CT field-of-view using two complementary short scans

BACKGROUND

Robotic C-arm cone-beam computed tomography (CBCT) scanners provide fast in-room imaging in radiotherapy. Their mobility extends beyond performing a gantry rotation, but they might encounter obstructions to their motion which limit the gantry angle range. The axial field-of-view (FOV) of a reconstructed CBCT image depends on the acquisition geometry. When imaging a large anatomical location, such as the thorax, abdomen, or pelvis, a centered cone beam might be insufficient to acquire untruncated projection images. Some CBCT scanners can laterally displace their detector and collimate the beam to increase the FOV, but the gantry must then perform a360 ∘ $360^{\circ}$ rotation to provide complete data for reconstruction.

PURPOSE

To extend the FOV of a CBCT image with a single short scan (gantry angle range of180 ∘ + $180^{\circ}+$ fan angle) using two complementary short scans.

METHODS

We defined an acquisition protocol using two short scans during which the source follows the same trajectory and where the detector has equal and opposite tilt and/or offset between the two scans, which we refer to as complementary scans. We created virtual acquisitions using a Monte Carlo simulator on a digital anthropomorphic phantom and on a computed tomography (CT) scan of a patient abdomen. For our proposed method, each simulation produced two complementary sets of projections, which were weighted for redundancies and used to reconstruct one CBCT image. We compared the resulting images to the ground truth phantoms and simulations of conventional scans.

RESULTS

Reconstruction artifacts were slightly more prominent in the complementary scans w.r.t. a complete scan with untruncated projections but matched those in a single short scan without truncation. When analyzing reconstructed scans from simulated projections with scatter and corrected with prior CT information, we found a global agreement between complementary and conventional scan approaches.

CONCLUSIONS

When dealing with a limited range of motion of the gantry of a CBCT scanner, two complementary short scans are a technically valid alternative to a full 360∘ $^{\circ}$ scan with equal FOV. This approach enables FOV extension without collisions or hardware upgrades.

Joint EURADOS-EANM initiative for an advanced computational framework for the assessment of external dose rates from nuclear medicine patients

BACKGROUND

In order to ensure adequate radiation protection of critical groups such as staff, caregivers and the general public coming into proximity of nuclear medicine (NM) patients, it is necessary to consider the impact of the radiation emitted by the patients during their stay at the hospital or after leaving the hospital. Current risk assessments are based on ambient dose rate measurements in a single position at a specified distance from the patient and carried out at several time points after administration of the radiopharmaceutical to estimate the whole-body retention. The limitations of such an approach are addressed in this study by developing and validating a more advanced computational dosimetry approach using Monte Carlo (MC) simulations in combination with flexible and realistic computational phantoms and time activity distribution curves from reference biokinetic models.

RESULTS

Measurements of the ambient dose rate equivalent Ḣ*(10) at 1 m from the NM patient have been successfully compared against MC simulations with 5 different codes using the ICRP adult reference computational voxel phantoms, for typical clinical procedures with 99mTc-HDP/MDP, 18FDG and Na131I. All measurement data fall in the 95% confidence intervals, determined for the average simulated results. Moreover, the different MC codes (MCNP-X, PHITS, GATE, GEANT4, TRIPOLI-4®) have been compared for a more realistic scenario where the effective dose rate Ė of an exposed individual was determined in positions facing and aside the patient model at 30 cm, 50 cm and 100 cm. The variation between codes was lower than 8% for all the radiopharmaceuticals at 1 m, and varied from 5 to 16% for the face-to face and side-by-side configuration at 30 cm and 50 cm. A sensitivity study on the influence of patient model morphology demonstrated that the relative standard deviation of Ḣ*(10) at 1 m for the range of included patient models remained under 16% for time points up to 120 min post administration.

CONCLUSIONS

The validated computational approach will be further used for the evaluation of effective dose rates per unit administered activity for a variety of close-contact configurations and a range of radiopharmaceuticals as part of risk assessment studies. Together with the choice of appropriate dose constraints this would facilitate the setting of release criteria and patient restrictions.

Fano cavity test and investigation of the response of the Roos chamber irradiated by proton beams in perpendicular magnetic fields up to 1 T

Objective. The aim of this work is to investigate the response of the Roos chamber (type 34001) irradiated by clinical proton beams in magnetic fields.Approach. At first, a Fano test was implemented in Monte Carlo software package GATE version 9.2 (based on Geant4 version 11.0.2) using a cylindrical slab geometry in a magnetic field up to 1 T. In accordance to an experimental setup (Fuchset al2021), the magnetic field correction factorskQB⃗of the Roos chamber were determined at different energies up to 252 MeV and magnetic field strengths up to 1 T, by separately simulating the ratios of chamber signalsMQ/MQB⃗,without and with magnetic field, and the dose-conversion factorsDw,QB⃗/Dw,Qin a small cylinder of water, with and without magnetic field. Additionally, detailed simulations were carried out to understand the observed magnetic field dependence.Main results. The Fano test was passed with deviations smaller than 0.25% between 0 and 1 T. The ratios of the chamber signals show both energy and magnetic field dependence. The maximum deviation of the dose-conversion factors from unity of 0.22% was observed at the lowest investigated proton energy of 97.4 MeV andB⃗= 1 T. The resultingkQB⃗factors increase initially with the applied magnetic field and decrease again after reaching a maximum at around 0.5 T; except for the lowest 97.4 MeV beam that show no observable magnetic field dependence. The deviation from unity of the factors is also larger for higher proton energies, where the maximum lies at 1.0035(5), 1.0054(7) and 1.0069(7) for initial energies ofE0= 152, 223.4 and 252 MeV, respectively.Significance. Detailed Monte Carlo studies showed that the observed effect can be mainly attributed to the differences in the transport of electrons produced both outside and inside of the air cavity in the presence of a magnetic field.

Impact of scatter correction on personalized dosimetry in selective internal radiotherapy using 166Ho-PLLA: a single-center study including Monte-Carlo simulation, phantom and patient imaging

BACKGROUND

Developments in transarterial radioembolization led to the conception of new microspheres loaded with holmium-166 (166Ho). However, due to the complexity of the scatter components in 166Ho single photon emission computed tomography (SPECT), questions about image quality and dosimetry are emerging. The aims of this work are to investigate the scatter components and correction methods to propose a suitable solution, and to evaluate the impact on image quality and dosimetry including Monte-Carlo (MC) simulations, phantom, and patient data.

METHODS

Dual energy window (DEW) and triple energy window (TEW) methods were investigated for scatter correction purposes and compared using Contrast Recovery Coefficients (CRC) and Contrast to Noise Ratios (CNR). First, MC simulations were carried out to assess all the scatter components in the energy windows used, also to confirm the choice of the parameter needed for the DEW method. Then, MC simulations of acquisitions of a Jaszczak phantom were conducted with conditions mimicking an ideal scatter correction. These simulated projections can be reconstructed and compared with real acquisitions corrected by both methods and then reconstructed. Finally, both methods were applied on patient data and their impact on personalized dosimetry was evaluated.

RESULTS

MC simulations confirmed the use of k = 1 for the DEW method. These simulations also confirmed the complexity of scatter components in the main energy window used with a high energy gamma rays component of about half of the total counts detected, together with a negligible X rays component and a negligible presence of fluorescence. CRC and CNR analyses, realized on simulated scatter-free projections of the phantom and on scatter corrected acquisitions of the same phantom, suggested an increased efficiency of the TEW method, even at the price of higher level of noise. Finally, these methods, applied on patient data, showed significant differences in terms of non-tumoral liver absorbed dose, non-tumoral liver fraction under 50 Gy, tumor absorbed dose, and tumor fraction above 150 Gy.

CONCLUSIONS

This study demonstrated the impact of scatter correction on personalized dosimetry on patient data. The use of a TEW method is proposed for scatter correction in 166Ho SPECT imaging.

Efficient image reconstruction for a small animal PET system with dual-layer-offset detector design

BACKGROUND

A compact PET/SPECT/CT system Inliview-3000B has been developed to provide multi-modality information on small animals for biomedical research. Its PET subsystem employed a dual-layer-offset detector design for depth-of-interaction capability and higher detection efficiency, but the irregular design caused some difficulties in calculating the normalization factors and the sensitivity map. Besides, the relatively larger (2 mm) crystal cross-section size also posed a challenge to high-resolution image reconstruction.

PURPOSE

We present an efficient image reconstruction method to achieve high imaging performance for the PET subsystem of Inliview-3000B.

METHODS

List mode reconstruction with efficient system modeling was used for the PET imaging. We adopt an on-the-fly multi-ray tracing method with random crystal sampling to model the solid angle, crystal penetration and object attenuation effect, and modify the system response model during each iteration to improve the reconstruction performance and computational efficiency. We estimate crystal efficiency with a novel iterative approach that combines measured cylinder phantom data with simulated line-of-response (LOR)-based factors for normalization correction before reconstruction. Since it is necessary to calculate normalization factors and the sensitivity map, we stack the two crystal layers together and extend the conventional data organization method here to index all useful LORs. Simulations and experiments were performed to demonstrate the feasibility and advantage of the proposed method.

RESULTS

Simulation results showed that the iterative algorithm for crystal efficiency estimation could achieve good accuracy. NEMA image quality phantom studies have demonstrated the superiority of random sampling, which is able to achieve good imaging performance with much less computation than traditional uniform sampling. In the spatial resolution evaluation based on the mini-Derenzo phantom, 1.1 mm hot rods could be identified with the proposed reconstruction method. Reconstruction of double mice and a rat showed good spatial resolution and a high signal-to-noise ratio, and organs with higher uptake could be recognized well.

CONCLUSION

The results validated the superiority of introducing randomness into reconstruction, and demonstrated its reliability for high-performance imaging. The Inliview-3000B PET subsystem with the proposed image reconstruction can provide rich and detailed information on small animals for preclinical research.

Dual-isotope imaging method for Actinium-225 and Bismuth-213 using alpha imaging detector

Recently, 225Ac has received considerable attention for its use in targeted alpha therapy because it has a relatively long half-life and yields four more alpha-particles from the daughter nuclides. The performance evaluation should separately assess the distribution of 225Ac and 213Bi because daughter nuclides, including 213Bi, can cause renal toxicity, which may hinder the widespread use of 225Ac for targeted alpha therapy. In this study, we describe and validate a spectrum decomposition method for dual-isotope imaging of 225Ac and 213Bi using an alpha imaging detector. We implemented an experiment to demonstrate the feasibility of using the alpha imaging detector to obtain distribution images using therapeutic amounts of 225Ac. In addition, we designed and conducted a Monte Carlo simulation under realistic conditions based on the experimental results to evaluate the spectrum decomposition method for dual-isotope imaging. The alpha imaging detector exhibited a detection efficiency of 18.5% and an energy resolution of 13.4% at 5.5 MeV. In the simulation, the distributions of 225Ac and 213Bi were obtained in each region with a relative error of 5%. The results of this study confirmed the feasibility of the dual-isotope imaging method for discriminating alpha-emitters using small amounts of 225Ac.

Measurement of the12C(p,n)12N reaction cross section below 150 MeV

Objective. Proton therapy currently faces challenges from clinical complications on organs-at-risk due to range uncertainties. To address this issue, positron emission tomography (PET) of the proton-induced11C and15O activity has been used to provide feedback on the proton range. However, this approach is not instantaneous due to the relatively long half-lives of these nuclides. An alternative nuclide,12N (half-life 11 ms), shows promise for real-timein vivoproton range verification. Development of12N imaging requires better knowledge of its production reaction cross section.Approach. The12C(p,n)12N reaction cross section was measured by detecting positron activity of graphite targets irradiated with 66.5, 120, and 150 MeV protons. A pulsed beam delivery with 0.7-2 × 108protons per pulse was used. The positron activity was measured during the beam-off periods using a dual-head Siemens Biograph mCT PET scanner. The12N production was determined from activity time histograms.Main results. The cross section was calculated for 11 energies, ranging from 23.5 to 147 MeV, using information on the experimental setup and beam delivery. Through a comprehensive uncertainty propagation analysis, a statistical uncertainty of 2.6%-5.8% and a systematic uncertainty of 3.3%-4.6% were achieved. Additionally, a comparison between measured and simulated scanner sensitivity showed a scaling factor of 1.25 (±3%). Despite this, there was an improvement in the precision of the cross section measurement compared to values reported by the only previous study.Significance. Short-lived12N imaging is promising for real-timein vivoverification of the proton range to reduce clinical complications in proton therapy. The verification procedure requires experimental knowledge of the12N production cross section for proton energies of clinical importance, to be incorporated in a Monte Carlo framework for12N imaging prediction. This study is the first to achieve a precise measurement of the12C(p,n)12N nuclear cross section for such proton energies.

Monte Carlo simulation study of an in vivo four-dimensional tracking system with a diverging collimator for monitoring radiation source (Ir-192) location during brachytherapy: proof of concept and feasibility

Introduction: The aim of this study was to demonstrate the potential of an in vivo four-dimensional (4D) tracking system to accurately localize the radiation source, Iridium-192 (Ir-192) in high-dose rate brachytherapy. Methods: To achieve time-dependent 3D positioning of the Ir-192 source, we devised a 4D tracking system employing multiple compact detectors. During the system's design phase, we conducted comprehensive optimization and analytical evaluations of the diverging collimator employed for detection purposes. Subsequently, we executed 3D reconstruction and positioning procedures based on the 2D images obtained by six detectors, each equipped with an optimized diverging collimator. All simulations for designing and evaluating the 4D tracking system were performed using the open-source GATE (v9.1) Monte Carlo platform based on the GEANT4 (v10.7) toolkit. In addition, to evaluate the accuracy of the proposed 4D tracking system, we conducted simulations and 3D positioning using a solid phantom and patient data. Finally, the error between the reconstructed position coordinates determined by the tracking system and the original coordinates of the Ir-192 radiation source was analyzed. Results: The parameters for the optimized diverging collimator were a septal thickness of 0.3 mm and a collimator height of 30 mm. A tracking system comprising 6 compact detectors was designed and implemented utilizing this collimator. Analysis of the accuracy of the proposed Ir-192 source tracking system found that the average of the absolute values of the error between the 3D reconstructed and original positions for the simulation with the solid phantom were 0.440 mm for the x coordinate, 0.423 mm for the y coordinate, and 0.764 mm for the z coordinate, and the average Euclidean distance was 1.146 mm. Finally, in a simulation based on data from a patient who underwent brachytherapy, the average Euclidean distance between the original and reconstructed source position was 0.586 mm. Discussion: These results indicated that the newly designed in vivo 4D tracking system for monitoring the Ir-192 source during brachytherapy could determine the 3D position of the radiation source in real time during treatment. We conclude that the proposed positioning system has the potential to make brachytherapy more accurate and reliable.

Crystal scatter effects in a large-area dual-panel Positron Emission Mammography system

Positron Emission Mammography (PEM) is a valuable molecular imaging technique for breast studies using pharmaceuticals labeled with positron emitters and dual-panel detectors. PEM scanners normally use large scintillation crystals coupled to sensitive photodetectors. Multiple interactions of the 511 keV annihilation photons in the crystals can result in event mispositioning leading to a negative impact in radiopharmaceutical uptake quantification. In this work, we report the study of crystal scatter effects of a large-area dual-panel PEM system designed with either monolithic or pixelated lutetium yttrium orthosilicate (LYSO) crystals using the Monte Carlo simulation platform GATE. The results show that only a relatively small fraction of coincidences (~20%) arise from events where both coincidence photons undergo single interactions (mostly through photoelectric absorption) in the crystals. Most of the coincidences are events where at least one of the annihilation photons undergoes a chain of Compton scatterings: approximately 79% end up in photoelectric absorption while the rest (<1%) escape the detector. Mean positioning errors, calculated as the distance between first hit and energy weighted (assigned) positions of interaction, were 1.70 mm and 1.92 mm for the monolithic and pixelated crystals, respectively. Reconstructed spatial resolution quantification with a miniDerenzo phantom and a list mode iterative reconstruction algorithm shows that, for both crystal types, 2 mm diameter hot rods were resolved, indicating a relatively small effect in spatial resolution. A drastic reduction in peak-to-valley ratios for the same hot-rod diameters was observed, up to a factor of 14 for the monolithic crystals and 7.5 for the pixelated ones.

A time- and space-saving Monte Carlo simulation method using post-collimation generative adversarial network for dose calculation of an O-ring gantry Linac

PURPOSE

This study explores the feasibility of employing Generative Adversarial Networks (GANs) to model the RefleXion X1 Linac. The aim is to investigate the accuracy of dose simulation and assess the potential computational benefits.

METHODS

The X1 Linac is a new radiotherapy machine with a binary multi-leaf collimation (MLC) system, facilitating innovative biology-guided radiotherapy. A total of 34 GAN generators, each representing a desired MLC aperture, were developed. Each generator was trained using a phase space file generated underneath the corresponding aperture, enabling the generation of particles and serving as a beam source for Monte Carlo simulation. Dose distributions in water were simulated for each aperture using both the GAN and phase space sources. The agreement between dose distributions was evaluated. The computational time reduction from bypassing the collimation simulation and storage space savings were estimated.

RESULTS

The percentage depth dose at 10 cm, penumbra, and full-width half maximum of the GAN simulation agree with the phase space simulation, with differences of 0.4 % ± 0.2 %, 0.32 ± 0.66 mm, and 0.26 ± 0.44 mm, respectively. The gamma passing rate (1 %/1mm) for the planar dose exceeded 90 % for all apertures. The estimated time-saving for simulating an plan using 5766 beamlets was 530 CPU hours. The storage usage was reduced by a factor of 102.

CONCLUSION

The utilization of the GAN in simulating the X1 Linac demonstrated remarkable accuracy and efficiency. The reductions in both computational time and storage requirements make this approach highly valuable for future dosimetry studies and beam modeling.

Time-of-flight scatter rejection in x-ray radiography

Objective.Time-of-flight (TOF) scatter rejection allows for identifying and discarding scattered photons without the use of an anti-scatter grid (ASG). Although TOF scatter rejection was initially presented for cone-beam computed tomography, we propose, herein, to extend this approach to x-ray radiography. This work aims to evaluate with simulations if TOF scatter rejection can outperform ASGs for radiography.Approach.GATE was used to simulate the radiography of a head and a torso and a water cylinder with bone inserts in a system with total timing jitters from 0 ps up to 500 ps full-width-at-half-maximum. The transmission factor of TOF scatter rejection for primary and scattered photons was evaluated as if it were a virtual ASG.Main results.With a total timing jitter of 50 ps, TOF scatter rejection can reach a selectivity of 4.93 with a primary photons transmission of 99%. Reducing the timing jitter close to 0 ps increases the selectivity up to 15.85 for a head and torso radiography, outperforming typical ASGs which usually have a selectivity from 2.5 to 10 with a primary photons transmission from 50% to 70%.Significance.This suggests that TOF scatter rejection may be suitable to replace ASGs in applications requiring lower radiation exposure if sufficiently low timing jitter is achieved.

Quantitative calibration of Tb-161 SPECT/CT in view of personalised dosimetry assessment studies

BACKGROUND

Terbium-161 (161Tb)-based radionuclide therapy poses an alternative to current Lutetium-177 (177Lu) approaches with the additional benefit of secondary Auger and conversion electron emissions capable of delivering high doses of localised damage to micro-metastases including single cells. Quantitative single-photon emission computed tomography, paired with computed tomography (SPECT/CT), enables quantitative measurement from post-therapy imaging. In view of dosimetry extrapolations, a Tb-161 sensitivity SPECT/CT camera calibration was performed using a method previously validated for 177Lu.

METHODS

Serial imaging of a NEMA/IEC body phantom with Tb-161 was performed on SPECT/CT with low-energy high-resolution collimators employing a photopeak of 75 keV with a 20% width. Quantitative stability and recovery coefficients were investigated over a sequence of 19 scans with buffered 161Tb solution at total phantom activity ranging from 70 to 4990 MBq.

RESULTS

Sphere recovery coefficients were 0.60 ± 0.05, 0.52 ± 0.07, 0.45 ± 0.07, 0.39 ± 0.07, 0.28 ± 0.08, and 0.20 ± 0.08 for spheres 37, 28, 22, 17, 13, and 10mm, respectively, when considered across all activity and scan durations with dual-energy window scatter correction. Whole-field reconstructed sensitivity was calculated as 1.42E-5 counts per decay. Qualitatively, images exhibited no visual artefacts and were comparable to 177Lu SPECT/CT.

CONCLUSIONS

Quantitative SPECT/CT of 161Tb is feasible over a range of activities enabling dosimetry analogous to 177Lu whilst also producing suitable imaging for clinical review. This has been incorporated into a prospective trial of 161Tb-PSMA for men with metastatic prostate cancer.

Application of NEMA protocols to verify GATE models based on the Digital Biograph Vision and the Biograph Vision Quadra scanners

PURPOSE

The Monte Carlo method is an effective tool to simulate and verify PET systems. Furthermore, it can help in the design and optimization of new medical imaging devices and algorithms. In this framework, the goal of this work is to verify the GATE toolkit performance when applied to simulate two Siemens Healthineers PET scanners: a standard axial field-of-view Biograph Vision scanner and the new long axial field-of-view Biograph Vision Quadra scanner.

METHODS

The simulation toolkit GATE is based on GEANT4, comprising its main functionalities and a set of domain-specific features in the field of medical physics. To accomplish our purpose, the guidelines described in the NEMA NU 2-2018 protocol are reproduced. Then the simulated results are compared to experimental data available in the literature for both PET scanners. The assessment of the models includes different studies of sensitivity, count rate performances, spatial resolution and image quality. These tests are intended to evaluate the image quality of PET devices.

RESULTS

In the spatial resolution test, relative errors lower than 8% are obtained between the experiments and GATE models. The sensitivity is 17.2 cps/kBq (Vision) and 175.9 cps/kBq (Quadra), representing relative differences with the experiment of 6% and 0.3%, respectively. Deviations from peak NECR are less than 9%. In the Image Quality test, the contrast recovery coefficient for hot spheres, with 8 iterations and 5 subsets, ranges between 57-83% for Vision and 54-86% for Quadra. These values represent a maximum deviation between the simulations and the experiments of 10% for the Quadra scanner. In the case of the Vision scanner, the highest difference is observed for the 10 mm sphere (∼38%) due to the higher contrast recovery of the experiment caused by the Gibbs artifact from the use of PSF reconstruction.

CONCLUSIONS

The results of the simulations have provided evidence of a good agreement between the experimental data and the results obtained with GATE. Thus, this work supports the capability of this MC toolkit to accurately simulate the models of the Vision and Quadra scanners. This study has laid the basis for further research in this field and has identified several areas that could be explored.

Resolving inter-crystal scatter in a light-sharing depth-encoding PET detector

Objective.Inter-crystal scattering (ICS) in light-sharing positron emission tomography (PET) detectors leads to ambiguity in positioning the initial interaction, which significantly degrades the contrast, quantitative accuracy, and spatial resolution of the resulting image. Here, we attempt to resolve the positioning ambiguity of ICS in a light-sharing depth-encoding detector by exploiting the confined, deterministic light-sharing enabled by the segmented light guide unique to Prism-PET.Approach.We first considered a test case of ICS between two adjacent crystals using an analytical and a neural network approach. The analytical approach used a Bayesian estimation framework constructed from a scatter absorption model-the prior-and a detector response model-the likelihood. A simple neural network was generated for the same scenario, to provide mutual validation for the findings. Finally, we generalized the solution to three-dimensional event positioning that handles all events in the photopeak using a convolutional neural network with unique architecture that separately predicts the identity and depth-of-interaction (DOI) of the crystal containing the first interaction.Main results.The analytical Bayesian method generated an estimation error of 20.5 keV in energy and 3.1 mm in DOI. Further analysis showed that the detector response model was sufficiently robust to achieve adequate performance via maximum likelihood estimation (MLE), without prior information. We then found convergent results using a simple neural network. In the generalized solution using a convolutional neural network, we found crystal identification accuracy of 83% and DOI estimation error of 3.0 mm across all events. Applying this positioning algorithm to simulated data, we demonstrated significant improvements in image quality over the baseline, centroid-based positioning approach, attaining 38.9% improvement in intrinsic spatial resolution and enhanced clarity in hot spots of diameters 0.8 to 2.5 mm.Significance.The accuracy of our findings exceeds those of previous reports in the literature. The Prism-PET light guide, mediating confined and deterministic light-sharing, plays a key role in ICS recovery, as its mathematical embodiment-the detector response model-was the essential driver of accuracy in our results.

Automatic planning for head and neck seed implant brachytherapy based on deep convolutional neural network dose engine

BACKGROUND

Seed implant brachytherapy (SIBT) is an effective treatment modality for head and neck (H&N) cancers; however, current clinical planning requires manual setting of needle paths and utilizes inaccurate dose calculation algorithms.

PURPOSE

This study aims to develop an accurate and efficient deep convolutional neural network dose engine (DCNN-DE) and an automatic SIBT planning method for H&N SIBT.

METHODS

A cohort of 25 H&N patients who received SIBT was utilized to develop and validate the methods. The DCNN-DE was developed based on 3D-unet model. It takes single seed dose distribution from a modified TG-43 method, the CT image and a novel inter-seed shadow map (ISSM) as inputs, and predicts the dose map of accuracy close to the one from Monte Carlo simulations (MCS). The ISSM was proposed to better handle inter-seed attenuation. The accuracy and efficacy of the DCNN-DE were validated by comparing with other methods taking MCS dose as reference. For SIBT planning, a novel strategy inspired by clinical practice was proposed to automatically generate parallel or non-parallel potential needle paths that avoid puncturing bone and critical organs. A heuristic-based optimization method was developed to optimize the seed positions to meet clinical prescription requirements. The proposed planning method was validated by re-planning the 25 cases and comparing with clinical plans.

RESULTS

The absolute percentage error in the TG-43 calculation for CTV V100 and D90 was reduced from 5.4% and 13.2% to 0.4% and 1.1% with DCNN-DE, an accuracy improvement of 93% and 92%, respectively. The proposed planning method could automatically obtain a plan in 2.5 ± 1.5 min. The generated plans were judged clinically acceptable with dose distribution comparable with those of the clinical plans.

CONCLUSIONS

The proposed method can generate clinically acceptable plans quickly with high accuracy in dose evaluation, and thus has a high potential for clinical use in SIBT.

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.

Lifetime attributable risks (LARs) of cancer in the fetus associated with maternal radiography examinations

PURPOSE

For various reasons, pregnant women are occasionally exposed to ionizing radiation during radiology examinations. In these situations, it is essential to determine the radiation dose to the fetus and any associated risks. The present study attempts to calculate the mean dose for the fetus to estimate the possible cancer induction and cancer mortality risks resulting from maternal radiography exams.

MATERIAL AND METHODS

The GATE Monte Carlo platform and a standard voxelized pregnant phantom were employed to calculate fetal radiation dose during maternal radiography exams. The data published in Biological Effects of Ionizing Radiation VII were used to convert fetal dose to lifetime attributable risks (LARs) of cancer incidence and cancer-related mortality.

RESULTS

The fetal doses and LARs of cancer incidence and cancer-related mortality for the radiographs of the chest and skull were negligible. The maximum LAR values for the lateral view of the abdomen in computed and digital radiography are 5598.29 and 2238.95 per 100,000 individuals, respectively. The computed radiography of the lateral view of the abdomen revealed the highest LAR of cancer-related mortality (2074.30 deaths for every 100,000 people).

CONCLUSION

The radiation dose incurred by the fetus due to chest and skull radiographs was minimal and unlikely to cause any abnormalities in the fetus. The discernible elevation in the lifetime attributable risk associated with cancer incidence and mortality arising from lateral computed radiography examinations of the abdomen warrants careful consideration within the realm of maternal radiography examinations.

Improvement of Whole-body Bone Planar Images on a Bone-dedicated Single-photon Emission Computed Tomography Scanner by Blind Deconvolution Algorithm

Purpose

We have developed a bone-dedicated collimator with higher sensitivity but slightly degraded resolution on single-photon emission computed tomography (SPECT) for planar bone scintigraphy, compared with conventional low-energy high-resolution collimator. In this work, we investigated the feasibility of using the blind deconvolution algorithm to improve the resolution of planar images on bone scintigraphy.

Materials and Methods

Monte Carlo simulation was performed with the NCAT phantom for modeling bone scintigraphy on the clinical dual-head SPECT scanner (Imagine NET 632, Beijing Novel Medical Equipment Ltd.) equipped with the bone-dedicated collimator. Maximum likelihood estimation method was used for the blind deconvolution algorithm. The initial estimation of point spread function (PSF) and iteration number for the method were determined by comparing the deblurred images obtained from different input parameters. We simulated different tumors in five different locations and with five different diameters to evaluate the robustness of the initial inputs. Furthermore, we performed chest phantom studies on the clinical SPECT scanner. The quantified increased contrast ratio (CR) between the tumor and the background was evaluated.

Results

The 2 mm PSF kernel and 10 iterations provided a practical and robust deblurred image on our system. Those two inputs can generate robust deblurred images in terms of the tumor location and size with an average increased CR of 21.6%. The phantom studies also demonstrated the ability of blind deconvolution, using those two inputs, with increased CRs of 17%, 17%, 22%, 20%, and 13% for lesions with diameters of 1 cm, 2 cm, 3 cm, 4 cm, and 5 cm, respectively.

Conclusions

It is feasible to use the blind deconvolution algorithm to deblur the planar images for SPECT bone scintigraphy. The appropriate values of the PSF kernel and the iteration number for the blind deconvolution can be determined using simulation studies.

Application of quantum entanglement induced polarization for dual-positron and prompt gamma imaging

The intrinsic resolution of Positron Emission Tomography (PET) imaging is bound by positron range effects, wherein the radioactive decay of the imaging tracer occurs at a disjoint location from positron annihilation. Compounding this issue are the variable ranges positrons achieve, depending on tracer species (the energy they are emitted with) and the medium they travel in (bone vs soft tissue, for example) - causing the range to span more than an order of magnitude across various study scenarios (~0.19 mm to ~6.4 mm). Radioisotopes, such as Zr-89, exhibit dual emissions of positron and prompt gammas, offering an opportunity for accurate tracer positioning as prompt gammas originate from the tracer location. These multi-emission radiotracers have historically suffered from increased noise corresponding to the third gamma interfering in annihilation gamma coincidence pairing. Recent advancements, however, have brought to light the unique property of annihilation gammas having scattering kinematics distinct from random gamma pairs. These properties are born from the singular quantum entanglement state available to the gamma pair following para-positronium decay which prescribes linearly orthogonal polarization. Such coherent polarization is not shared by prompt gamma emissions, offering an opportunity for their discrimination. We present an investigation into this technique, comparing the distribution of relevant scattering kinematics of entangled annihilation gammas and corresponding prompt gammas via a Monte Carlo simulation.

Exploration of differentiability in a proton computed tomography simulation framework

Objective.Gradient-based optimization using algorithmic derivatives can be a useful technique to improve engineering designs with respect to a computer-implemented objective function. Likewise, uncertainty quantification through computer simulations can be carried out by means of derivatives of the computer simulation. However, the effectiveness of these techniques depends on how 'well-linearizable' the software is. In this study, we assess how promising derivative information of a typical proton computed tomography (pCT) scan computer simulation is for the aforementioned applications.Approach.This study is mainly based on numerical experiments, in which we repeatedly evaluate three representative computational steps with perturbed input values. We support our observations with a review of the algorithmic steps and arithmetic operations performed by the software, using debugging techniques.Main results.The model-based iterative reconstruction (MBIR) subprocedure (at the end of the software pipeline) and the Monte Carlo (MC) simulation (at the beginning) were piecewise differentiable. However, the observed high density and magnitude of jumps was likely to preclude most meaningful uses of the derivatives. Jumps in the MBIR function arose from the discrete computation of the set of voxels intersected by a proton path, and could be reduced in magnitude by a 'fuzzy voxels' approach. The investigated jumps in the MC function arose from local changes in the control flow that affected the amount of consumed random numbers. The tracking algorithm solves an inherently non-differentiable problem.Significance.Besides the technical challenges of merely applying AD to existing software projects, the MC and MBIR codes must be adapted to compute smoother functions. For the MBIR code, we presented one possible approach for this while for the MC code, this will be subject to further research. For the tracking subprocedure, further research on surrogate models is necessary.

Towards Neural Charged Particle Tracking in Digital Tracking Calorimeters With Reinforcement Learning

We propose a novel technique for reconstructing charged particles in digital tracking calorimeters using reinforcement learning aiming to benefit from the rapid progress and success of neural network architectures without the dependency on simulated or manually-labeled data. Here we optimize by trial-and-error a behavior policy acting as an approximation to the full combinatorial optimization problem, maximizing the physical plausibility of sampled trajectories. In modern processing pipelines used in high energy physics and related applications, tracking plays an essential role allowing to identify and follow charged particle trajectories traversing particle detectors. Due to the high multiplicity of charged particles and their physical interactions, randomly deflecting the particles, the reconstruction is a challenging undertaking, requiring fast, accurate and robust algorithms. Our approach works on graph-structured data, capturing track hypotheses through edge connections between particles in the detector layers. We demonstrate in a comprehensive study on simulated data for a particle detector used for proton computed tomography, the high potential as well as the competitiveness of our approach compared to a heuristic search algorithm and a model trained on ground truth. Finally, we point out limitations of our approach, guiding towards a robust foundation for further development of reinforcement learning based tracking.

Deep learning-based fast denoising of Monte Carlo dose calculation in carbon ion radiotherapy

BACKGROUND

Plan verification is one of the important steps of quality assurance (QA) in carbon ion radiotherapy. Conventional methods of plan verification are based on phantom measurement, which is labor-intensive and time-consuming. Although the plan verification method based on Monte Carlo (MC) simulation provides a more accurate modeling of the physics, it is also time-consuming when simulating with a large number of particles. Therefore, how to ensure the accuracy of simulation results while reducing simulation time is the current difficulty and focus.

PURPOSE

The purpose of this work was to evaluate the feasibility of using deep learning-based MC denoising method to accelerate carbon-ion radiotherapy plan verification.

METHODS

Three models, including CycleGAN, 3DUNet and GhostUNet with Ghost module, were used to denoise the 1 × 106 carbon ions-based MC dose distribution to the accuracy of 1 × 108 carbon ions-based dose distribution. The CycleGAN's generator, 3DUNet and GhostUNet were all derived from the 3DUNet network. A total of 59 cases including 29 patients with head-and-neck cancers and 30 patients with lung cancers were collected, and 48 cases were randomly selected as the training set of the CycleGAN network and six cases as the test set. For the 3DUNet and GhostUNet models, the numbers of training set, validation set, and test set were 47, 6, and 6, respectively. Finally, the three models were evaluated qualitatively and quantitatively using RMSE and three-dimensional gamma analysis (3 mm, 3%).

RESULTS

The three end-to-end trained models could be used for denoising the 1 × 106 carbon ions-based dose distribution, and their generalization was proved. The GhostUNet obtained the lowest RMSE value of 0.075, indicating the smallest difference between its denoised and 1 × 108 carbon ions-based dose distributions. The average gamma passing rate (GPR) between the GhostUNet denoising-based versus 1 × 108 carbon ions-based dose distributions was 99.1%, higher than that of the CycleGAN at 94.3% and the 3DUNet at 96.2%. Among the three models, the GhostUNet model had the fewest parameters (4.27 million) and the shortest training time (99 s per epoch) but achieved the best denoising results.

CONCLUSION

The end-to-end deep network GhostUNet outperforms the CycleGAN, 3DUNet models in denoising MC dose distributions for carbon ion radiotherapy. The network requires less than 5 s to denoise a sample of MC simulation with few particles to obtain a qualitative and quantitative result comparable to the dose distribution simulated by MC with relatively large number particles, offering a significant reduction in computation time.

A simulation study of the system characteristics for a long axial FOV PET design based on monolithic BGO flat panels compared with a pixelated LSO cylindrical design

BACKGROUND

Although a new generation of tomographs with a longer axial field-of-view called total-body PET have been developed, they are not widely utilized due to their high cost compared to conventional scanners. The newly designed walk-through total-body PET scanner is introduced as a high-throughput and cost-efficient alternative to total-body PET scanners, by making use of a flat panel geometry and lower cost, depth-of-interaction capable, monolithic BGO detectors. The main aim of the presented study is to evaluate through Monte Carlo simulation the system characteristics of the walk-through total-body PET scanner by comparing it with a Quadra-like total-body PET of similar attributes to the Siemens Biograph Vision Quadra.

METHODS

The walk-through total-body PET is comprised of two flat detector panels, spaced 50 cm apart. Each panel, 70 [Formula: see text] 106 cm[Formula: see text] in size, consists of 280 BGO-based monolithic detectors. The Quadra-like TB-PET has been simulated based on the characteristics of the Biograph Vision Quadra, one of the most common total-body PET scanners with 106 cm of axial field-of-view, which is constructed with pixelated LSO scintillation crystals. The spatial resolution, sensitivity, count rate performance, scatter fractions, and image quality of both scanners are simulated in the GATE simulation toolkit for comparison.

RESULTS

Due to the DOI-capable detectors used in the walk-through total-body PET, the values of the spatial resolution of this scanner were all below 2 mm along directions parallel to the panels, and reached a maximum of 3.36 mm in the direction perpendicular to the panels. This resolution is a large improvement compared to the values of the Quadra-like TB-PET. The walk-through total-body PET uses its maximum sensitivity (154 cps/kBq) for data acquisition and image reconstruction.

CONCLUSION

Based on the combination of very good spatial resolution and high sensitivity of the walk-through total-body PET, along with a 2.2 times lower scintillation crystal volume and 1.8 times lower SiPM surface, this scanner can be a very cost-efficient alternative for total-body PET scanners in cases where concomitant CT is not required.

Radiobiological Assessment of Targeted Radionuclide Therapy with [177Lu]Lu-PSMA-I&T in 2D vs. 3D Cell Culture Models

In vitro therapeutic efficacy studies are commonly conducted in cell monolayers. However, three-dimensional (3D) tumor spheroids are known to better represent in vivo tumors. This study used [177Lu]Lu-PSMA-I&T, an already clinically applied radiopharmaceutical for targeted radionuclide therapy against metastatic castrate-resistant prostate cancer, to demonstrate the differences in the radiobiological response between 2D and 3D cell culture models of the prostate cancer cell lines PC-3 (PSMA negative) and LNCaP (PSMA positive). After assessing the target expression in both models via Western Blot, cell viability, reproductive ability, and growth inhibition were assessed. To investigate the geometric effects on dosimetry for the 2D vs. 3D models, Monte Carlo simulations were performed. Our results showed that PSMA expression in LNCaP spheroids was highly preserved, and target specificity was shown in both models. In monolayers of LNCaP, no short-term (48 h after treatment), but only long-term (14 days after treatment) radiobiological effects were evident, showing decreased viability and reproductive ability with the increasing activity. Further, LNCaP spheroid growth was inhibited with the increasing activity. Overall, treatment efficacy was higher in LNCaP spheroids compared to monolayers, which can be explained by the difference in the resulting dose, among others.

An automatic pipeline for PET/MRI attenuation correction validation in the brain

PURPOSE

Challenges in PET/MRI quantitative accuracy for neurological uses arise from PET attenuation correction accuracy. We proposed and evaluated an automatic pipeline to assess the quantitative accuracy of four MRI-derived PET AC methods using analytically simulated PET brain lesions and ROIs as ground truth for PET activity.

METHODS

Our proposed pipeline, integrating a synthetic lesion insertion tool and the FreeSurfer neuroimaging framework, inserts simulated spherical and brain ROIs into PET projection space, reconstructing them via four PET MRAC techniques. Utilizing an 11-patient brain PET dataset, we compared the quantitative accuracy of four MRACs (DIXON, DIXONbone, UTE AC, and DL-DIXON) against the gold standard PET CTAC, evaluating MRAC to CTAC activity bias in spherical lesions and brain ROIs with and without background activity against original (lesion free) PET reconstructed images.

RESULTS

The proposed pipeline yielded accurate results for spherical lesions and brain ROIs, adhering to the MRAC to CTAC pattern of original brain PET images. Among the MRAC methods, DIXON AC exhibited the highest bias, followed by UTE, DIXONBone, and DL-DIXON showing the least. DIXON, DIXONbone, UTE, and DL-DIXON showed MRAC to CTAC biases of - 5.41%, - 1.85%, - 2.74%, and 0.08% respectively for ROIs inserted in background activity; - 7.02%, - 2.46%, - 3.56%, and - 0.05% for lesion ROIs without background; and - 6.82%, - 2.08%, - 2.29%, and 0.22% for the original brain PET images' 16 FreeSurfer brain ROIs.

CONCLUSION

The proposed pipeline delivers accurate results for synthetic spherical lesions and brain ROIs, with and without background activity consideration, enabling the evaluation of new attenuation correction approaches without utilizing measured PET emission data. Additionally, it offers a consistent method to generate realistic lesion ROIs, potentially applicable in assessing further PET correction techniques.

A stochastic alternative technique for Compton Maximum Likelihood Expectation-Maximization (MLEM) reconstruction

Even though iterative methods and particularly Maximum Likelihood Expectation-Maximization (MLEM) algorithms have been established in reconstruction with Compton data, their detailed design with respect to physical rules and processes dominate their plain implementation in the form of a system matrix. A new elementary but efficient alternative for the well-known system matrix with respect to Compton Camera image reconstruction is presented in this work. For each detected event there is a weighting factor inserted as an accumulated probability which carries all the necessary information. This probability which involves only the Compton scattering angle of the incident photon corresponds to a map that correlates all events within all possible source origins. Based on maximizing likelihood principles, the proposed model weights in a stochastic way the difference of the scatterer-to-source angle θ0, as it is determined by the deposited energy on the absorber, and any other potential scattering angle θJ, specified by the position coordinates on the reconstruction matrix. Obtained image spatial resolution, angular distortions and response to focal length determination are a few of the studied cases for the algorithms' evaluation via simulations in GEANT4/GATE with a set of radioactive sources and phantoms with in- and out-of-plane arrangement.

Electron contamination suppression in transmission detectors for radiotherapy

Objective.Higher energy and intensity radiotherapy beams are being used, in part, due to the increased spatial accuracy of treatments. However, higher intensity beams can result in a larger total dose error, motivating the increasing need for real-time dose monitoring. We are developing a thin, real-time upstream monolithic active pixel sensor based system for beam monitoring with excellent precision on measuring the beam shape. Here we present a method to additionally provide dosimetry by adding thin conversion material in strips to the surface of the detector, a grating structure.Approach.By modulating the thickness of the conversion material to minimally disturb the contamination electron signal while enhancing the photon signal, the difference in these signals can be used to extract a photon-only signal, and hence dose. The simulation software Gate, based on Geant4, is utilised to study whether well functioning gratings can be better made from aluminium or copper and to optimise the thickness of a copper grating.Main results.It is possible to enhance the photon signal by a factor 6.7 (7.7) compared to the bare sensor for a 5.8 (6.7) MV beam, without modulation of the signal due to beam electrons.Significance.The grating can be used to perform dosimetry in real-time using a thin upstream detector.

3D phantom for image quality assessment of mammography systems

Objective.To present an innovative approach for the design of a 3D mammographic phantom for medical equipment quality assessment, estimation of the glandular tissue percentage in the patient's breast, and emulation of microcalcification (μC) breast lesions.Approach.Contrast-to noise ratio (CNR) measurements, as well as spatial resolution and intensity-to-glandularity calibrations under mammography conditions were performed to assess the effectiveness of the phantom. CNR measurements were applied to different groups of calcium hydroxyapatite (HA) and aluminum oxide (AO)μCs ranging from 200 to 600μm. Spatial resolution was characterized using an aluminum plate contained in the phantom and standard linear figures of merit, such as the line spread function and modulation transfer function (MTF). The intensity-to-glandularity calibration was developed using an x-ray attenuation matrix within the phantom to estimate the glandular tissue percentage in a breast with a compressed thickness of 4 cm.Main results.For the prototype studied, the minimum confidence level for detecting HAμCs is 95.4%, while for AOμCs is above 68.3%. It was also possible to determine that the MTF of the commercial mammography machine used for this study at the Nyquist frequency is 41%. Additionally, a one-to-one intensity-to-glandularity calibration was obtained and verified with Monte-Carlo simulation results.Significance.The phantom provides traditional arrangements presented in accreditation phantoms, which makes it competitive with available devices, but excelling in regarding affordability, modularity, and inlays distribution. Moreover, its design allows to be positioned in close proximity to the patient's breast during a medical screening for a simultaneous x-ray imaging, such that the features of the phantom can be used as reference values to specify characteristics of the real breast tissue, such as proportion of glandular/adipose composition and/orμC type and size lesions.