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.

Deep learning using contrast-enhanced ultrasound images to predict the nuclear grade of clear cell renal cell carcinoma.

PURPOSE: To assess the effectiveness of a deep learning model using contrastenhanced ultrasound (CEUS) images in distinguishing between low-grade (grade I and II) and high-grade (grade III and IV) clear cell renal cell carcinoma (ccRCC). METHODS: A retrospective study was conducted using CEUS images of 177 Fuhrmangraded ccRCCs (93 low-grade and 84 high-grade) from May 2017 to December 2020. A total of 6412 CEUS images were captured from the videos and normalized for subsequent analysis. A deep learning model using the RepVGG architecture was proposed to differentiate between low-grade and high-grade ccRCC. The model's performance was evaluated based on sensitivity, specificity, positive predictive value, negative predictive value and area under the receiver operating characteristic curve (AUC). Class activation mapping (CAM) was used to visualize the specific areas that contribute to the model's predictions. RESULTS: For discriminating high-grade ccRCC from low-grade, the deep learning model achieved a sensitivity of 74.8%, specificity of 79.1%, accuracy of 77.0%, and an AUC of 0.852 in the test set. CONCLUSION: The deep learning model based on CEUS images can accurately differentiate between low-grade and high-grade ccRCC in a non-invasive manner.

Advanced deep learning models for phenotypic trait extraction and cultivar classification in lychee using photon-counting micro-CT imaging.

Introduction: In contemporary agronomic research, the focus has increasingly shifted towards non-destructive imaging and precise phenotypic characterization. A photon-counting micro-CT system has been developed, which is capable of imaging lychee fruit at the micrometer level and capturing a full energy spectrum, thanks to its advanced photon-counting detectors. Methods: For automatic measurement of phenotypic traits, seven CNN-based deep learning models including AttentionUNet, DeeplabV3+, SegNet, TransUNet, UNet, UNet++, and UNet3+ were developed. Machine learning techniques tailored for small-sample training were employed to identify key characteristics of various lychee species. Results: These models demonstrate outstanding performance with Dice, Recall, and Precision indices predominantly ranging between 0.90 and 0.99. The Mean Intersection over Union (MIoU) consistently falls between 0.88 and 0.98. This approach served both as a feature selection process and a means of classification, significantly enhancing the study's ability to discern and categorize distinct lychee varieties. Discussion: This research not only contributes to the advancement of non-destructive plant analysis but also opens new avenues for exploring the intricate phenotypic variations within plant species.

Ultrasound image segmentation of renal tumors based on UNet++ with fusion of multiscale residuals and dual attention.

Objective. Laparoscopic renal unit-preserving resection is a routine and effective means of treating renal tumors. Image segmentation is an essential part before tumor resection. The current segmentation method mainly relies on doctors manual delineation, which is time-consuming, labor-intensive, and influenced by their personal experience and ability. And the image quality of segmentation is low, with problems such as blurred edges, unclear size and shape, which are not conducive to clinical diagnosis.Approach. To address these problems, we propose an automated segmentation method, i.e. the UNet++ algorithm fusing multiscale residuals and dual attention (MRDA_UNet++). It replaces two consecutive 3 × 3 convolutions in UNet++ with the 'MultiRes block' module, which incorporates coordinate attention to fuse features from different scales and suppress the impact of background noise. Furthermore, an attention gate is also added at the short connections to enhance the ability of the network to extract features from the target area.Main results. The experimental results show that MRDA_UNet++ achieves 93.18%, 92.87%, 93.66%, and 92.09% on the real-world dataset for MIoU, Dice, Precision, and Recall, respectively. Compared to the baseline model UNet++ on three public datasets, the MIoU, Dice, and Recall metrics improved by 6.00%, 7.90% and 18.09% respectively for BUSI, 0.39%, 0.27% and 1.03% for Dataset C, and 1.37%, 1.75% and 1.30% for DDTI.Significance. The proposed MRDA_UNet++ exhibits obvious advantages in feature extraction, which can not only significantly reduce the workload of doctors, but also further decrease the risk of misdiagnosis. It is of great value to assist doctors diagnosis in the clinic.

Guided block matching and 4-D transform domain filter projection denoising method for dynamic PET image reconstruction.

PURPOSE: Dynamic PET is an essential tool in oncology due to its ability to visualize and quantify radiotracer uptake, which has the potential to improve imaging quality. However, image noise caused by a low photon count in dynamic PET is more significant than in static PET. This study aims to develop a novel denoising method, namely the Guided Block Matching and 4-D Transform Domain Filter (GBM4D) projection, to enhance dynamic PET image reconstruction. METHODS: The sinogram was first transformed using the Anscombe method, then denoised using a combination of hard thresholding and Wiener filtering. Each denoising step involved guided block matching and grouping, collaborative filtering, and weighted averaging. The guided block matching was performed on accumulated PET sinograms to prevent mismatching due to low photon counts. The performance of the proposed denoising method (GBM4D) was compared to other methods such as wavelet, total variation, non-local means, and BM3D using computer simulations on the Shepp-Logan and digital brain phantoms. The denoising methods were also applied to real patient data for evaluation. RESULTS: In all phantom studies, GBM4D outperformed other denoising methods in all time frames based on the structural similarity and peak signal-to-noise ratio. Moreover, GBM4D yielded the lowest root mean square error in the time-activity curve of all tissues and produced the highest image quality when applied to real patient data. CONCLUSION: GBM4D demonstrates excellent denoising and edge-preserving capabilities, as validated through qualitative and quantitative assessments of both temporal and spatial denoising performance.

CNN-based automatic segmentations and radiomics feature reliability on contrast-enhanced ultrasound images for renal tumors.

Objective: To investigate the feasibility and efficiency of automatic segmentation of contrast-enhanced ultrasound (CEUS) images in renal tumors by convolutional neural network (CNN) based models and their further application in radiomic analysis. Materials and methods: From 94 pathologically confirmed renal tumor cases, 3355 CEUS images were extracted and randomly divided into training set (3020 images) and test set (335 images). According to the histological subtypes of renal cell carcinoma, the test set was further split into clear cell renal cell carcinoma (ccRCC) set (225 images), renal angiomyolipoma (AML) set (77 images) and set of other subtypes (33 images). Manual segmentation was the gold standard and serves as ground truth. Seven CNN-based models including DeepLabV3+, UNet, UNet++, UNet3+, SegNet, MultilResUNet and Attention UNet were used for automatic segmentation. Python 3.7.0 and Pyradiomics package 3.0.1 were used for radiomic feature extraction. Performance of all approaches was evaluated by the metrics of mean intersection over union (mIOU), dice similarity coefficient (DSC), precision, and recall. Reliability and reproducibility of radiomics features were evaluated by the Pearson coefficient and the intraclass correlation coefficient (ICC). Results: All seven CNN-based models achieved good performance with the mIOU, DSC, precision and recall ranging between 81.97%-93.04%, 78.67%-92.70%, 93.92%-97.56%, and 85.29%-95.17%, respectively. The average Pearson coefficients ranged from 0.81 to 0.95, and the average ICCs ranged from 0.77 to 0.92. The UNet++ model showed the best performance with the mIOU, DSC, precision and recall of 93.04%, 92.70%, 97.43% and 95.17%, respectively. For ccRCC, AML and other subtypes, the reliability and reproducibility of radiomic analysis derived from automatically segmented CEUS images were excellent, with the average Pearson coefficients of 0.95, 0.96 and 0.96, and the average ICCs for different subtypes were 0.91, 0.93 and 0.94, respectively. Conclusion: This retrospective single-center study showed that the CNN-based models had good performance on automatic segmentation of CEUS images for renal tumors, especially the UNet++ model. The radiomics features extracted from automatically segmented CEUS images were feasible and reliable, and further validation by multi-center research is necessary.

Patient-specific fetal radiation dosimetry for pregnant patients undergoing abdominal and pelvic CT imaging.

BACKGROUND: Accurate estimation of fetal radiation dose is crucial for risk-benefit analysis of radiological imaging, while the radiation dosimetry studies based on individual pregnant patient are highly desired. PURPOSE: To use Monte Carlo calculations for estimation of fetal radiation dose from abdominal and pelvic computed tomography (CT) examinations for a population of patients with a range of variations in patients' anatomy, abdominal circumference, gestational age (GA), fetal depth (FD), and fetal development. METHODS: Forty-four patient-specific pregnant female models were constructed based on CT imaging data of pregnant patients, with gestational ages ranging from 8 to 35 weeks. The simulation of abdominal and pelvic helical CT examinations was performed on three validated commercial scanner systems to calculate organ-level fetal radiation dose. RESULTS: The absorbed radiation dose to the fetus ranged between 0.97 and 2.24 mGy, with an average of 1.63 ± 0.33 mGy. The CTDIvol -normalized fetal dose ranged between 0.56 and 1.30, with an average of 0.94 ± 0.25. The normalized fetal organ dose showed significant correlations with gestational age, maternal abdominal circumference (MAC), and fetal depth. The use of ATCM technique increased the fetal radiation dose in some patients. CONCLUSION: A technique enabling the calculation of organ-level radiation dose to the fetus was developed from models of actual anatomy representing a range of gestational age, maternal size, and fetal position. The developed maternal and fetal models provide a basis for reliable and accurate radiation dose estimation to fetal organs.

Development of a voxel S-value database for patient internal radiation dosimetry.

PURPOSE: Personalized dosimetry with high accuracy drew great attention in clinical practices. Voxel S-value (VSV) convolution has been proposed to speed up absorbed dose calculations. However, the VSV method is efficient for personalized internal radiation dosimetry only when there are pre-calculated VSVs of the radioisotope. In this work, we propose a new method for VSV calculation based on the developed mono-energetic particle VSV database of γ, ß, α, and X-ray for any radioisotopes. METHODS: Mono-energetic VSV database for γ, ß, α, and X-ray was calculated using Monte Carlo methods. Radiation dose was first calculated based on mono-energetic VSVs for [F-18]-FDG in 10 patients. The estimated doses were compared with the values obtained from direct Monte Carlo simulation for validation of the proposed method. The number of VSVs used in calculation was optimized based on the estimated dose accuracy and computation time. RESULTS: The generated VSVs showed a great consistency with the results calculated using direct Monte Carlo simulation. For [F-18]-FDG, the proposed VSV method with number of VSV of 9 shows the best relative average organ absorbed dose uncertainty of 3.25% while the calculation time was reduced by 99% and 97% compared to the Monte Carlo simulation and traditional multiple VSV methods, respectively. CONCLUSIONS: In this work, we provided a method to generate the VSV kernels for any radioisotope based on the pre-calculated mono-energetic VSV database and significantly reduced the time cost for the multiple VSVs dosimetry approach. A software was developed to generate VSV kernels for any radioisotope in 19 mediums.

Automatic organ completion with image stitching for personalized radiation dosimetry in CT examinations.

PURPOSE: Computed tomography (CT) image-based patient-specific voxel-based dosimetry has difficulties complementing missing tissues for organs located partially inside or completely outside the image volume. Previous studies constructed patient-specific whole-body models by rescaling reference phantoms or extending regional CT images with manually adjusted phantoms. This study proposes a methodology for automatic organ completion of regional CT images for CT dosimetry using a stitching approach. METHODS: Virtual clinical trials were performed by truncating whole-body CT images to generate virtual clinical chest and abdominopelvic CT images. Corresponding anchor images for each patient were selected according to sex and similarity of the axial length and water equivalent diameter of the virtual regional CT images. Automatic image stitching was performed by transformation initialization and iteration, while the stitched CT images and organ atlas were used in GPU-based Geant4 Monte Carlo simulations to generate a radiation dose map and absorbed organ dose. To evaluate the performance of the stitching model in radiation dosimetry, organ mass differences and Jaccard's coefficient of stitched and rescaled anchor images were calculated, and the radiation doses were compared among the corresponding values from the VirtualDose®, original whole-body CT, stitching model, regional CT, registration-based rescaling method, and WED-based rescaling method. RESULTS: The anatomical accuracy of stitched images was significantly improved. For organs partially inside the image volume, organ dose estimation from the stitching model could be more accurate than that reported in previous studies. The absolute differences in effective dose from the stitched images were 6.55% and 4.81% for chest and abdominopelvic CT scans, respectively. CONCLUSION: The proposed automatic stitching model partially complements organs inside or outside the CT scan range and provides more accurate anatomical representations for radiation dosimetry than traditional phantom rescaling methods.

Construction of a digital fetus library for radiation dosimetry.

PURPOSE: Accurate estimations of fetal absorbed dose and radiation risks are crucial for radiation protection and important for radiological imaging research owing to the high radiosensitivity of the fetus. Computational anthropomorphic models have been widely used in patient-specific radiation dosimetry calculations. In this work, we aim to build the first digital fetal library for more reliable and accurate radiation dosimetry studies. ACQUISITION AND VALIDATION METHODS: Computed tomography (CT) images of abdominal and pelvic regions of 46 pregnant females were segmented by experienced medical physicists. The segmented tissues/organs include the body contour, skeleton, uterus, liver, kidney, intestine, stomach, lung, bladder, gall bladder, spleen, and pancreas for maternal body, and placenta, amniotic fluid, fetal body, fetal brain, and fetal skeleton. Nonuniform rational B-spline (NURBS) surfaces of each identified region was constructed manually using 3D modeling software. The Hounsfield unit values of each identified organs were gathered from CT images of pregnant patients and converted to tissue density. Organ volumes were further adjusted according to reference measurements for the developing fetus recommended by the World Health Organization (WHO) and International Commission on Radiological Protection. A series of anatomical parameters, including femur length, humerus length, biparietal diameter, abdominal circumference (FAC), and head circumference, were measured and compared with WHO recommendations. DATA FORMAT AND USAGE NOTES: The first fetal patient-specific model library was developed with the anatomical characteristics of each model derived from the corresponding patient whose gestational age varies between 8 and 35 weeks. Voxelized models are represented in the form of MCNP matrix input files representing the three-dimensional model of the fetus. The size distributions of each model are also provided in text files. All data are stored on Zenodo and are publicly accessible on the following link: https://zenodo.org/record/6471884. POTENTIAL APPLICATIONS: The constructed fetal models and maternal anatomical characteristics are consistent with the corresponding patients. The resulting computational fetus could be used in radiation dosimetry studies to improve the reliability of fetal dosimetry and radiation risks assessment. The advantages of NURBS surfaces in terms of adapting fetal postures and positions enable us to adequately assess their impact on radiation dosimetry calculations.

Triple-source saddle-curve cone-beam photon counting CT image reconstruction: A simulation study.

PURPOSE: The most common detector material in the PC CT system, cannot achieve the best performance at a relatively higher photon flux rate. In the reconstruction view, the most commonly used filtered back projection, is not able to provide sufficient reconstructed image quality in spectral computed tomography (CT). Developing a triple-source saddle-curve cone-beam photon counting CT image reconstruction method can improve the temporal resolution. METHODS: Triple-source saddle-curve cone-beam trajectory was rearranged into four trajectory sets for simulation and reconstruction. Projection images in different energy bins were simulated by forward projection and photon counting CT respond model simulation. After simulation, the object was reconstructed using Katsevich's theory after photon counts correction using the pseudo inverse of photon counting CT response matrix. The material decomposition can be performed based on images in different energy bins. RESULTS: Root mean square error (RMSE) and structural similarity index (SSIM) are calculated to quantify the image quality of reconstruction images. Compared with FDK images, the RMSE for the triple-source image was improved by 27%, 21%, 14%, 8%, and 6% for the reconstrued image of 20-33, 33-47, 47-58, 58-69, 69-80 keV energy bin. The SSIM was improved by 1.031%, 0.665%, 0.396%, 0.235%, 0.174% for corresponding energy bin. The decomposition image based on corrected images shows improved RMSE and SSIM, each by 33.861% and 0.345%. SSIM of corrected decomposition image of iodine reaches 99.415% of the original image. CONCLUSIONS: A new Triple-source saddle-curve cone-beam PC CT image reconstruction method was developed in this work. The exact reconstruction of the triple-source saddle-curve improved both the image quality and temporal resolution.

Tracers progress for positron emission tomography imaging of glial-related disease.

Glial cells play an essential part in the neuron system. They can not only serve as structural blocks in the human brain but also participate in many biological processes. Extensive studies have shown that astrocytes and microglia play an important role in neurodegenerative diseases, such as Alzheimer's disease, Parkinson's disease, Huntington's disease, as well as glioma, epilepsy, ischemic stroke, and infections. Positron emission tomography is a functional imaging technique providing molecular-level information before anatomic changes are visible and has been widely used in many above-mentioned diseases. In this review, we focus on the positron emission tomography tracers used in pathologies related to glial cells, such as glioma, Alzheimer's disease, and neuroinflammation.

Internal dosimetry in F-18 FDG PET examinations based on long-time-measured organ activities using total-body PET/CT: does it make any difference from a short-time measurement?

PURPOSE: A 2-m axial field-of-view, total-body PET/CT scanner (uEXPLORER) has been recently developed to provide total-body coverage and ultra-high sensitivity, which together, enables opportunities for in vivo time-activity curve (TAC) measurement of all investigated organs simultaneously with high temporal resolution. This study aims at quantifying the cumulated activity and patient dose of 2-[F-18]fluoro-2-deoxy-D-glucose (F-18 FDG ) imaging by using delayed time-activity curves (TACs), measured out to 8-h post-injection, for different organs so that the comparison between quantifying approaches using short-time method (up to 75 min post-injection) or long-time method (up to 8 h post-injection) could be performed. METHODS: Organ TACs of 10 healthy volunteers were collected using total-body PET/CT in 4 periods after the intravenous injection of F-18 FDG. The 8-h post-injection TACs of 6 source organs were fitted using a spline method (based on Origin (version 8.1)). To compare with cumulated activity estimated from spline-fitted curves, the cumulated activity estimated from multi-exponential curve was also calculated. Exponential curve was fitted with shorter series of data consistent with clinical procedure and previous dosimetry works. An 8-h dynamic bladder wall dose model considering 2 voiding were employed to illustrate the differences in bladder wall dose caused by the different measurement durations. Organ absorbed doses were further estimated using Medical Internal Radiation Dose (MIRD) method and voxel phantoms. RESULTS: A short-time measurement could lead to significant bias in estimated cumulated activity for liver compared with long-time-measured spline fitted method, and the differences of cumulated activity were 18.38% on average. For the myocardium, the estimated cumulated activity difference was not statistically significant due to large variation in metabolism among individuals. The average residence time differences of brain, heart, kidney, liver, and lungs were 8.38%, 15.13%, 25.02%, 23.94%, and 16.50% between short-time and long-time methods. Regarding effective dose, the maximum differences of residence time between long-time-measured spline fitted curve and short-time-measured multi-exponential fitted curve was 9.93%. When using spline method, the bladder revealed the most difference in the effective dose among all the investigated organs with a bias up to 21.18%. The bladder wall dose calculated using a long-time dynamic model was 13.79% larger than the two-voiding dynamic model, and at least 50.17% lower than previous studies based on fixed bladder content volume. CONCLUSIONS: Long-time measurement of multi-organ TACs with high temporal resolution enabled by a total-body PET/CT demonstrated that the clinical procedure with 20 min PET scan at 1 h after injection could be used for retrospective dosimetry analysis in most organs. As the bladder content contributed the most to the effective dose, a long-time dynamic model was recommended for the bladder wall dose estimation.

Multi-scale Monte Carlo simulations of gold nanoparticle-induced DNA damages for kilovoltage X-ray irradiation in a xenograft mouse model using TOPAS-nBio.

Background: Gold nanoparticles (AuNPs) are considered as promising agents to increase the radiosensitivity of tumor cells. However, the biological mechanisms of radiation enhancement effects of AuNPs are still not well understood. We present a multi-scale Monte Carlo simulation framework within TOPAS-nBio to investigate the increase of DNA damage due to the presence of AuNPs in mouse tumor models. Methods: A tumor was placed inside a voxel mouse model and irradiated with either 100 kVp or 200 kVp x-ray beams. Phase spaces were employed to transfer particles from the macroscopic (voxel) scale to the microscopic scale, which consists of a cell geometry including a detailed mouse DNA model. Radiosensitizing effects were calculated in the presence and absence of hybrid nanoparticles with a Fe2O3 core surrounded by a gold layer (AuFeNPs). To simulate DNA damage even for very small energy tracks, Geant4-DNA physics and chemistry models were used on microscopic scale. Results: An AuFeNP induced enhancement of both dose and DNA strand breaks has been established for different scenarios. Produced chemical radicals including hydroxyl molecules, which were assumed to be responsible for DNA damage through chemical reactions, were found to be significantly increased. We further observed a dependency of the results on the location of the cells within the tumor for 200 kVp x-ray beams. Conclusions: Our multi-scale approach allows to study irradiation induced physical and chemical effects on cells. We showed a potential increase in cell radiosensitization caused by relatively small concentrations of AuFeNPs. Our new methodology allows the individual adjustment of parameters in each simulation step and therefore can be used for other studies investigating the radiosensitizing effects of AuFeNPs or AuNPs in living cells.

Age-dependent dose calculations for common PET radionuclides and brain radiotracers in nonhuman primate computational models.

PURPOSE: The combination of nonhuman primates (NHPs) with the state-of-the-art molecular imaging technologies allows for within-subject longitudinal research aiming at gaining new insights into human normal and disease conditions and provides an ideal foundation for future translational studies of new diagnostic tools, medical interventions, and therapies. However, radiation dose estimations for nonhuman primates from molecular imaging probes are lacking and are difficult to perform experimentally. The aim of this work is to construct age-dependent NHP computational model series to estimate the absorbed dose to NHP specimens in common molecular imaging procedures. MATERIALS AND METHODS: A series of NHP models from baby to adult were constructed based on nonuniform rational B-spline surface (NURBS) representations. Particle transport was simulated using Monte Carlo calculations to estimate S-values from nine positron-emitting radionuclides and absorbed doses from PET radiotracers. RESULTS: Realistic age-dependent NHP computational model series were developed. For most source-target pairs in computational NHP models, differences between C-11 S-values were between -13.4% and -8.8%/kg difference in body weight while differences between F-18 S-values were between -12.9% and -8.0%/kg difference in body weight. The absorbed doses of 11 C-labeled brain receptor substances, 18 F-labeled brain receptor substances, and 18 F-FDG in the brain ranged within 0.047-0.32 mGy/MBq, 0.25-1.63 mGy/MBq, and 0.32-2.12 mGy/MBq, respectively. CONCLUSION: The absorbed doses to organs are significantly higher in the baby NHP model than in the adult model. These results can be used in translational longitudinal studies to estimate the cumulated absorbed organ doses in NHPs at various ages.

Assessment of uncertainties associated with Monte Carlo-based personalized dosimetry in clinical CT examinations.

The clinical value of x-ray computed tomography (CT) has skyrocketed in the last decade while at the same time being the main source of medical exposure to the population. Concerns regarding the potential health hazards associated with the use of ionizing radiation were raised and an appropriate estimation of absorbed dose to patients is highly desired. In this work, we aim to validate our developed Monte Carlo CT simulator using in-phantom dose measurements and further assess the impact of personalized scan-related parameters on dosimetric calculations. We developed a Monte Carlo-based CT simulator for personalized organ level dose calculations, in which the CT source model, patient-specific computational model and personalized scanning protocol were integrated. The CT simulator was benchmarked using an ionization chamber and standard CT Dose Index phantom while the dosimetry methodology was validated through experimental measurements using thermoluminescent dosimeters (TLDs) embedded within an anthropomorphic phantom. Patient-specific scan protocols extracted from CT raw data and DICOM image metadata, respectively, were fed as input into the CT simulator to calculate individualized dose profiles. Thereby, the dosimetric uncertainties associated with using different protocol-related parameters were investigated. The absolute absorbed dose difference between measurements and simulations using the ionization chamber was less than 3%. In the case of the anthropomorphic phantom, the absolute absorbed dose difference between simulations and TLD measurements ranged from -8.3% to 22%, with a mean absolute difference of 14% while the uncertainties of protocol-related input parameters introduced an extra absolute error of 15% to the simulated results compared with TLD measurements. The developed methodology can be employed for accurate estimation of organ level dose from clinical CT examinations. The validated methodology can be further developed to produce an accurate MC simulation model with a reduced computational burden.

Development of a nonhuman primate computational phantom for radiation dosimetry.

PURPOSE: The nonhuman primate (NHP) is an important animal model for evaluating the response of the human body to radiation exposure owing to similarities between its organ structure, genome, life span, and metabolism. However, there is a lack of radiation dosimetry estimations for NHPs. The aim of this work is to construct a computational phantom of NHPs and estimate absorbed fractions and specific absorbed fractions for internal radiation dosimetry. MATERIALS AND METHODS: A female rhesus monkey was frozen and sectioned using a cryomacrotome. The transaxial sectioned images were imported into Adobe Photoshop for segmentation of internal organs. A total of 31 organs/tissues were identified and labeled. The segmented voxel phantom was then converted to a boundary representation (BREP) phantom to enable easy alteration of the phantom to mimic monkeys of different stature. The BREP model was then voxelized and imported into the MCNPX Monte Carlo radiation transport code for electron and photon dosimetry calculations. To estimate the appropriateness of using human phantoms as surrogate models for NHPs, absorbed fractions (AFs) and specific absorbed fractions (SAFs) of monoenergetic electrons and photons were calculated and compared with the ICRP reference newborn female phantom and a 1-yr-old female phantom. RESULTS: Considerable differences were observed for both self-absorbed and cross-absorbed doses for some organs between the NHP phantom and newborn phantom. For example, the ratios of the self-absorbed SAF(stomach wall) Monkey to SAF(stomach wall) Newborn ranged from 0.06 at 10 keV to 0.29 at 3 MeV for photons while the corresponding ratios to cross-absorbed SAF(liver→kidney) Monkey to SAF(liver→kidney) Newborn ranged from 0.78 at 50 keV to 5.78 at 10 keV for photons. Conversely, values of self-absorbed SAF were much higher (ratios of 2.39-4.19) for the brain and much lower for the uterus (0.51-0.61) in the monkey model compared to the newborn phantom. These dose differences can be attributed to the disparities between organ masses, shapes, and positions between the two phantoms. CONCLUSIONS: The developed NHP model can be exploited for the assessment of radiation dose to NHPs in preclinical radiation dosimetry studies and radiation therapy research.

Estimation of the radiation dose in pregnancy: an automated patient-specific model using convolutional neural networks.

OBJECTIVES: The conceptus dose during diagnostic imaging procedures for pregnant patients raises health concerns owing to the high radiosensitivity of the developing embryo/fetus. The aim of this work is to develop a methodology for automated construction of patient-specific computational phantoms based on actual patient CT images to enable accurate estimation of conceptus dose. METHODS: We developed a 3D deep convolutional network algorithm for automated segmentation of CT images to build realistic computational phantoms. The neural network architecture consists of analysis and synthesis paths with four resolution levels each, trained on manually labeled CT scans of six identified anatomical structures. Thirty-two CT exams were augmented to 128 datasets and randomly split into 80%/20% for training/testing. The absorbed doses for six segmented organs/tissues from abdominal CT scans were estimated using Monte Carlo calculations. The resulting radiation doses were then compared between the computational models generated using automated segmentation and manual segmentation, serving as reference. RESULTS: The Dice similarity coefficient for identified internal organs between manual segmentation and automated segmentation results varies from 0.92 to 0.98 while the mean Hausdorff distance for the uterus is 16.1 mm. The mean absorbed dose for the uterus is 2.9 mGy whereas the mean organ dose differences between manual and automated segmentation techniques are 0.07%, - 0.45%, - 1.55%, - 0.48%, - 0.12%, and 0.28% for the kidney, liver, lung, skeleton, uterus, and total body, respectively. CONCLUSION: The proposed methodology allows automated construction of realistic computational models that can be exploited to estimate patient-specific organ radiation doses from radiological imaging procedures. KEY POINTS: • The conceptus dose during diagnostic radiology and nuclear medicine imaging procedures for pregnant patients raises health concerns owing to the high radiosensitivity of the developing embryo/fetus. • The proposed methodology allows automated construction of realistic computational models that can be exploited to estimate patient-specific organ radiation doses from radiological imaging procedures. • The dosimetric results can be used for the risk-benefit analysis of radiation hazards to conceptus from diagnostic imaging procedures, thus guiding the decision-making process.

Construction of patient-specific computational models for organ dose estimation in radiological imaging.

PURPOSE: Diagnostic imaging procedures require optimization depending on the medical task at hand, the apparatus being used, and patient physical and anatomical characteristics. The assessment of the radiation dose and associated risks plays a key role in safety and quality management for radiation protection purposes. In this work, we aim at developing a methodology for personalized organ-level dose assessment in x-ray computed tomography (CT) imaging. METHODS: Regional voxel models representing reference patient-specific computational phantoms were generated through image segmentation of CT images for four patients. The best-fitting anthropomorphic phantoms were selected from a previously developed comprehensive phantom library according to patient's anthropometric parameters, then registered to the anatomical masks (skeleton, lung, and body contour) of patients to produce a patient-specific whole-body phantom. Well-established image registration metrics including Jaccard's coefficients for each organ, organ mass, body perimeter, organ-surface distance, and effective diameter are compared between the reference patient model, registered model, and anchor phantoms. A previously validated Monte Carlo code is utilized to calculate the absorbed dose in target organs along with the effective dose delivered to patients. The calculated absorbed doses from the reference patient models are then compared with the produced personalized model, anchor phantom, and those reported by commercial dose monitoring systems. RESULTS: The evaluated organ-surface distance and body effective diameter metrics show a mean absolute difference between patient regional voxel models, serving as reference, and patient-specific models around 4.4% and 4.5%, respectively. Organ-level radiation doses of patient-specific models are in good agreement with those of the corresponding patient regional voxel models with a mean absolute difference of 9.1%. The mean absolute difference of organ doses for the best-fitting model extracted from the phantom library and Radimetrics™ commercial dose tracking software are 15.5% and 41.1%, respectively. CONCLUSION: The results suggest that the proposed methodology improves the accuracy of organ-level dose estimation in CT, especially for extreme cases [high body mass index (BMI) and large skeleton]. Patient-specific radiation dose calculation and risk assessment can be performed using the proposed methodology for both monitoring of cumulative radiation exposure of patients and epidemiological studies. Further validation using a larger database is warranted.

Patient-Specific Computational Model and Dosimetry Calculations for PET/CT of a Patient Pregnant with Twins.

The radiation dose delivered to pregnant patients during radiologic imaging procedures raises health concerns because the developing embryo and fetus are considered to be highly radiosensitive. To appropriately weigh the diagnostic benefits against the radiation risks, the radiologist needs reasonably accurate and detailed estimates of the fetal dose. Expanding our previously developed series of computational phantoms for pregnant women, we here describe a personalized model for twin pregnancy, based on an actual clinical scan. Methods: The model is based on a standardized hybrid pregnant female and fetus phantom and on a clinical case of a patient who underwent an 18F-FDG PET/CT scan while expecting twins at 25 weeks' gestation. This model enabled us to produce a realistic physical representation of the pregnant patient and to estimate the maternal and fetal organ doses from the 18F-FDG and CT components. The Monte Carlo N-Particle Extended general-purpose code was used for radiation transport simulation. Results: The 18F-FDG doses for the 2 fetuses were 3.78 and 3.99 mGy, and the CT doses were 0.76 and 0.70 mGy, respectively. Therefore, the relative contribution of 18F-FDG and CT to the total dose to the fetuses was about 84% and 16%, respectively. Meanwhile, for 18F-FDG, the calculated personalized absorbed dose was about 40%-50% higher than the doses reported by other dosimetry computer software tools. Conclusion: Our approach to constructing personalized computational models allows estimation of a patient-specific radiation dose, even in cases with unusual anatomic features such as a twin pregnancy. Our results also show that, even in twins, the fetal organ doses from both 18F-FDG and CT present a certain variability linked to the anatomic characteristics. The CT fetal dose is smaller than the 18F-FDG PET dose.