Physical Phantom Research Articles

X-ray Cherenkov-luminescence tomography reconstruction with a three-component deep learning algorithm: Swin transformer, convolutional neural network, and locality module.

X-ray Cherenkov-luminescence tomography (XCLT) produces fast emission data from megavoltage (MV) x-ray scanning, in which the excitation location of molecules within tissue is reconstructed. However standard filtered backprojection (FBP) algorithms for XCLT sinogram reconstruction can suffer from insufficient data due to dose limitations, so there are limits in the reconstruction quality with some artifacts. We report a deep learning algorithm for XCLT with high image quality and improved quantitative accuracy. To directly reconstruct the distribution of emission quantum yield for x-ray Cherenkov-luminescence tomography, we proposed a three-component deep learning algorithm that includes a Swin transformer, convolution neural network, and locality module model. A data-to-image model x-ray Cherenkov-luminescence tomography is developed based on a Swin transformer, which is used to extract pixel-level prior information from the sinogram domain. Meanwhile, a convolutional neural network structure is deployed to transform the extracted pixel information from the sinogram domain to the image domain. Finally, a locality module is designed between the encoder and decoder connection structures for delivering features. Its performance was validated with simulation, physical phantom, and in vivo experiments. This approach can better deal with the limits to data than conventional FBP methods. The method was validated with numerical and physical phantom experiments, with results showing that it improved the reconstruction performance mean square error ( ), peak signal-to-noise ratio ( ), and Pearson correlation ( ) compared with the FBP algorithm. The Swin-CNN also achieved a 32.1% improvement in PSNR over the deep learning method AUTOMAP. This study shows that the three-component deep learning algorithm provides an effective reconstruction method for x-ray Cherenkov-luminescence tomography.

Robust reconstruction of fluorescence molecular tomography based on adaptive adversarial learning strategy

Objective. Fluorescence molecular tomography (FMT) is a promising molecular imaging modality for quantifying the three-dimensional (3D) distribution of tumor probes in small animals. However, traditional deep learning reconstruction methods that aim to minimize the mean squared error (MSE) and iterative regularization algorithms that rely on optimal parameters are typically influenced by strong noise, resulting in poor FMT reconstruction robustness. Approach. In this letter, we propose an adaptive adversarial learning strategy (3D-UR-WGAN) to achieve robust FMT reconstructions. Unlike the pixel-based MSE criterion in traditional CNNs or the regularization strategy in iterative solving schemes, the reconstruction strategy can greatly facilitate the performance of the network models through alternating loop training of the generator and the discriminator. Second, the reconstruction strategy combines the adversarial loss in GANs with the L1 loss to significantly enhance the robustness and preserve image details and textual information. Main results. Both numerical simulations and physical phantom experiments demonstrate that the 3D-UR-WGAN method can adaptively eliminate the effects of different noise levels on the reconstruction results, resulting in robust reconstructed images with reduced artifacts and enhanced image contrast. Compared with the state-of-the-art methods, the proposed method achieves better reconstruction performance in terms of target shape recovery and localization accuracy. Significance. This adaptive adversarial learning reconstruction strategy can provide a possible paradigm for robust reconstruction in complex environments, and also has great potential to provide an alternative solution for solving the problem of poor robustness encountered in other optical imaging modalities such as diffuse optical tomography, bioluminescence imaging, and Cherenkov luminescence imaging.

Image-spectral decomposition extended-learning assisted by sparsity for multi-energy computed tomography reconstruction.

Multi-energy computed tomography (CT) provides multiple channel-wise reconstructed images, and they can be used for material identification and k-edge imaging. Nonetheless, the projection datasets are frequently corrupted by various noises (e.g., electronic, Poisson) in the acquisition process, resulting in lower signal-noise-ratio (SNR) measurements. Multi-energy CT images have local sparsity, nonlocal self-similarity in spatial dimension, and correlation in spectral dimension. In this paper, we propose an image-spectral decomposition extended-learning assisted by sparsity (IDEAS) method to fully exploit these intrinsic priors for multi-energy CT image reconstruction. Particularly, a nonlocal low-rank Tucker decomposition (TD) is employed to utilize the correlation and nonlocal self-similarity priors. Moreover, considering the advantages of multi-task tensor dictionary learning (TDL) in sparse representation, an adaptive spatial dictionary and an adaptive spectral dictionary are trained during the iterative reconstruction process. Furthermore, a weighted total variation (TV) regularization term is employed to encourage local sparsity. Numerical simulation, physical phantom, and preclinical mouse experiments are performed to validate the proposed IDEAS algorithm. Specifically, in the simulation experiments, the proposed IDEAS reconstructed high-quality images that are very close to the references. For example, the root mean square error (RMSE) of IDEAS image in energy bin 1 is as low as 0.0672, while the RMSE of other methods are higher than 0.0843. Besides, the structural similarity (SSIM) of IDEAS reconstructed image in energy bin 1 is greater than 0.98. For material decomposition, the RMSE of IDEAS bone component is as low as 0.0152, and other methods are higher than 0.0199. In addition, the computational cost of IDEAS is as low as 98.8 s for one iteration, and the competing tensor decomposition method is higher than 327 s. To further improve the quality of the reconstructed multi-energy CT images, multiple prior regularizations are introduced to the multi-energy CT reconstructed model, leading to an IDEAS method. Both qualitative and quantitative evaluation of our results confirm the outstanding performance of the proposed algorithm compared to the state-of-the-arts.

Non-invasive cardiovascular magnetic resonance assessment of pressure recovery distance after aortic valve stenosis

BackgroundDecisions in the management of aortic stenosis are based on the peak pressure drop, captured by Doppler echocardiography, whereas gold standard catheterization measurements assess the net pressure drop but are limited by associated risks. The relationship between these two measurements, peak and net pressure drop, is dictated by the pressure recovery along the ascending aorta which is mainly caused by turbulence energy dissipation. Currently, pressure recovery is considered to occur within the first 40–50 mm distally from the aortic valve, albeit there is inconsistency across interventionist centers on where/how to position the catheter to capture the net pressure drop. MethodsWe developed a non-invasive method to assess the pressure recovery distance based on blood flow momentum via 4D Flow cardiovascular magnetic resonance (CMR). Multi-center acquisitions included physical flow phantoms with different stenotic valve configurations to validate this method, first against reference measurements and then against turbulent energy dissipation (respectively n = 8 and n = 28 acquisitions) and to investigate the relationship between peak and net pressure drops. Finally, we explored the potential errors of cardiac catheterisation pressure recordings as a result of neglecting the pressure recovery distance in a clinical bicuspid aortic valve (BAV) cohort of n = 32 patients. ResultsIn-vitro assessment of pressure recovery distance based on flow momentum achieved an average error of 1.8 ± 8.4 mm when compared to reference pressure sensors in the first phantom workbench. The momentum pressure recovery distance and the turbulent energy dissipation distance showed no statistical difference (mean difference of 2.8 ± 5.4 mm, R2 = 0.93) in the second phantom workbench. A linear correlation was observed between peak and net pressure drops, however, with strong dependences on the valvular morphology. Finally, in the BAV cohort the pressure recovery distance was 78.8 ± 34.3 mm from vena contracta, which is significantly longer than currently accepted in clinical practise (40–50 mm), and 37.5% of patients displayed a pressure recovery distance beyond the end of the ascending aorta. ConclusionThe non-invasive assessment of the distance to pressure recovery is possible by tracking momentum via 4D Flow CMR. Recovery is not always complete at the ascending aorta, and catheterised recordings will overestimate the net pressure drop in those situations. There is a need to re-evaluate the methods that characterise the haemodynamic burden caused by aortic stenosis as currently clinically accepted pressure recovery distance is an underestimation.

Selfrec-Net: self-supervised deep learning approach for the reconstruction of Cherenkov-excited luminescence scanned tomography.

As an emerging imaging technique, Cherenkov-excited luminescence scanned tomography (CELST) can recover a high-resolution 3D distribution of quantum emission fields within tissue using X-ray excitation for deep penetrance. However, its reconstruction is an ill-posed and under-conditioned inverse problem because of the diffuse optical emission signal. Deep learning based image reconstruction has shown very good potential for solving these types of problems, however they suffer from a lack of ground-truth image data to confirm when used with experimental data. To overcome this, a self-supervised network cascaded by a 3D reconstruction network and the forward model, termed Selfrec-Net, was proposed to perform CELST reconstruction. Under this framework, the boundary measurements are input to the network to reconstruct the distribution of the quantum field and the predicted measurements are subsequently obtained by feeding the reconstructed result to the forward model. The network was trained by minimizing the loss between the input measurements and the predicted measurements rather than the reconstructed distributions and the corresponding ground truths. Comparative experiments were carried out on both numerical simulations and physical phantoms. For singular luminescent targets, the results demonstrate the effectiveness and robustness of the proposed network, and comparable performance can be attained to a state-of-the-art deep supervised learning algorithm, where the accuracy of the emission yield and localization of the objects was far superior to iterative reconstruction methods. Reconstruction of multiple objects is still reasonable with high localization accuracy, although with limits to the emission yield accuracy as the distribution becomes more complex. Overall though the reconstruction of Selfrec-Net provides a self-supervised way to recover the location and emission yield of molecular distributions in murine model tissues.

Optimized three-dimensional ultrashort echo time: Magnetic resonance fingerprinting for myelin tissue fraction mapping.

Quantitative assessment of brain myelination has gained attention for both research and diagnosis of neurological diseases. However, conventional pulse sequences cannot directly acquire the myelin-proton signals due to its extremely short T2 and T2* values. To obtain the myelin-proton signals, dedicated short T2 acquisition techniques, such as ultrashort echo time (UTE) imaging, have been introduced. However, it remains challenging to isolate the myelin-proton signals from tissues with longer T2. In this article, we extended our previous two-dimensional ultrashort echo time magnetic resonance fingerprinting (UTE-MRF) with dual-echo acquisition to three dimensional (3D). Given a relatively low proton density (PD) of myelin-proton, we utilized Cramér-Rao Lower Bound to encode myelin-proton with the maximal SNR efficiency for optimizing the MR fingerprinting design, in order to improve the sensitivity of the sequence to myelin-proton. In addition, with a second echo of approximately 3ms, myelin-water component can be also captured. A myelin-tissue (myelin-proton and myelin-water) fraction mapping can be thus calculated. The optimized 3D UTE-MRF with dual-echo acquisition is tested in simulations, physical phantom and in vivo studies of both healthy subjects and multiple sclerosis patients. The results suggest that the rapidly decayed myelin-proton and myelin-water signal can be depicted with UTE signals of our method at clinically relevant resolution (1.8 mm isotropic) in 15 min. With its good sensitivity to myelin loss in multiple sclerosis patients demonstrated, our method for the whole brain myelin-tissue fraction mapping in clinical friendly scan time has the potential for routine clinical imaging.

Dosimetric impact of deformable image registration using radiophotoluminescent glass dosimeters with a physical geometric phantom.

To study the dosimetry impact of deformable image registration (DIR) using radiophotoluminescent glass dosimeter (RPLD) and custom developed phantom with various inserts. The phantom was developed to facilitate simultaneous evaluation of geometric and dosimetric accuracy of DIR. Four computed tomography (CT) images of the phantom were acquired with four different configurations. Four volumetric modulated arc therapy (VMAT) plans were computed for different phantom. Two different patterns were applied to combination of four phantom configurations. RPLD dose measurement was combined between corresponding two phantom configurations. DIR-based dose accumulation was calculated between corresponding two CT images with two commercial DIR software and various DIR parameter settings, and an open source software. Accumulated dose calculated using DIR was then compared with measured dose using RPLD. The mean±standard deviation (SD) of dose difference was 2.71±0.23% (range, 2.22%-3.01%) for tumor-proxy and 3.74±0.79% (range, 1.56%-4.83%) for rectum-proxy. The mean±SD of target registration error (TRE) was 1.66±1.36mm (range, 0.03-4.43mm) for tumor-proxy and 6.87±5.49mm (range, 0.54-17.47mm) for rectum-proxy. These results suggested that DIR accuracy had wide range among DIR parameter setting. The dose difference observed in our study was 3% for tumor-proxy and within 5% for rectum-proxy. The custom developed physical phantom with inserts showed potential for accurate evaluation of DIR-based dose accumulation. The prospect of simultaneous evaluation of geometric and dosimetric DIR accuracy in a single phantom may be useful for validation of DIR for clinical use.

Spectral CT Image-Domain Material Decomposition via Sparsity Residual Prior and Dictionary Learning

The spectral computed tomography (CT) system based on a photon-counting detector (PCD) can quantitatively analyze the material composition of the inspected object by material decomposition. Nonetheless, the raw projection of spectral CT is frequently disturbed by noise and artifacts, resulting in poor quality material decomposition images. Recently, a generalized dictionary learning based image-domain material decomposition (GDLIMD) to obtain high-quality material images. DL has great advantages in noise suppression and artifacts, while its protection of the fine structure and edge information is insufficient. To address this limitation, we proposed a sparsity residual prior and dictionary learning (SRPDL) algorithm for spectral CT image-domain material decomposition. The SRPDL method retains the noise-resistance performance of dictionary learning (DL) while introducing the pixel-value-based <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$l_{0} $ </tex-math></inline-formula> norm constraint to guide the material decomposition process by using the structural redundancy information between the prior image and the material images, which further improves structure protection and reduces material misclassification. We conducted numerical simulations, physical phantom, and preclinical experiments to validate and evaluate the SRPDL method. The results demonstrate that the proposed SRPDL method obtained better material decomposition accuracy than the state-of-the-art methods in noise reduction and edge protection.

Subjective Contrast of Pulmonary Tissues in CsI-FPD Chest Radiography Using Model Phantom by Monte Carlo Simulation

Subject contrast of pulmonary tissues was investigated for five X-ray beams (70 kV without filter, 90 kV with 0.15 mm Cu filter, 90 kV with 0.2 mm Cu filter, 120 kV without filter, and 120 kV with 0.2 mm Cu filter) in CsI-FPD chest radiography using two types of model phantoms by Monte Carlo simulation. A total of 72 million photons were entered to the model lung phantom (width, 300 mm; length, 300 mm; thickness, 200 mm; air space, 120 mm) and model mediastinum phantom (width, 300 mm; length, 300 mm; thickness, 200 mm; air space, 40 mm). Individual primary and secondary photon's process (absorption, scattering, and penetration) in the phantom and CsI-detector was recorded by Monte Carlo simulation. Subject contrast was calculated by entered and absorbed photon's number in the CsI-detector. Subject contrast pulmonary tissues were high to low energy X-ray beam; however, the ones of soft tissue and soft tissue overlaying bone had few differences for beam quality except 70 kV without filter. Moreover, the subject contrast by absorbed photons was higher compared to the one by entered photons in CsI. It was shown that the subject contrast study by Monte Carlo calculation can be replaced by the way of physical chest phantom, and that the subject contrast by absorbed photons and by injected photons in CsI was different. Furthermore, be verified that the subject contrast of soft tissue and soft tissue overlaying bone differs hardly.

Evaluation of silicon and 10% bismuth shield with variable thickness compared with constant thickness on the dose reduction and image quality during chest CT examination

One of the best methods to reduce the breast radiation dose in chest CT scan exam is using the bismuth shield. Due to the fact that, in CT scan tests, the breast dose is higher in the central axis of body than other sides, so the bismuth shield with variable thickness (outer half 1 mm thick and inner half 2 mm thick) is designed in this study. The objective of this study is to investigate in effectiveness of silicon and 10% bismuth composite shied with variable thickness on reducing radiation dose and image quality parameters in chest CT scan test compare to 1 mm and 2 mm of fixed thickness shield. Physical chest phantom underwent chest CT scan without or with bismuth shields with thickness of 1 mm, 2 mm, and variable in 90, 120 and 140 kVp in inactive TCM mode. Dosimetry was performed by using TLD, and image quality was evaluated quantitatively (by drawing the ROI in the identical images like image j, calculation of noise, CT number, SNR and CNR) and qualitatively (by two expert radiologists). Designed bismuth shield with unsteady thickness in inner and outer side compare to 1 mm and 2 mm thickness shields is presented in 120 kVp, has a significant difference in the amount of breast dose reduction (19% reduction), as well as in 140 kVp, all these three bismuth shields resulted in a significant dose reduction almost in the same similarity. At 120 kVp, the bismuth shield with variable thickness lead to a significant change in CT numbers in the heart and lungs, but it does not have a remarkable effect on other image quality parameters. The bismuth shield with variable thickness can lead to a better effectiveness in reducing breast dose without any side effect on image quality at 120 kVp, which requires further studies in this field.

Персонализированная дозиметрия внутреннего облучения опухолевых образований и органов риска пациентов: разработка и реализация методического базиса для дозиметрического обеспечения клинических исследований терапевтических радиофармпрепаратов

The cluster of calculational and instrumental methods for estimation of personalized internal radiation doses to foci and organs at risk among patients undergoing to therapy by radiopharmaceuticals has been developed. The developed set of methods was used for dosimetrical support of clinical trials of therapeutic radiopharmaceuticals: a) Lu-177-DOTA-PSMA (or "Lutaprost") targeted for radioligand therapy of metastatic castrate-resistant prostate cancer; b) Re-188 albumin microspheres 5-10 μm (or "Artroren") targeted for radiosynovectomy in the local treatment of chronic inflammatory diseases of the joints and c) 188Re albumin microspheres 20-40 μm (or "Gepatoren") targeted for intra-arterial radionuclide embolization in the treatment of inoperable liver cancer. The results of estimations absolute activities of radiopharmaceuticals and their dynamics during SPECT/CT scanning of radiopharmaceuticals in the body of patients were verified by measurements using physical phantoms of humans with different body weights and various standard activities of radionuclides distributed inside the phantoms. The developed cluster of programs (for calculating the absorbed fraction of energy in biostructures), relevant databases and instrumental methods were used as the basis for estimations personalized internal radiation doses in foci and organs at risk among patients included in clinical trials of the studied radiopharmaceuticals. Individual internal radiation doses in foci among 39 patients, included in the first phase of clinical trials of the three studied radiopharmaceuticals, were estimated. The analysis of the distribution of estimated doses shows that absorbed doses in foci are very differing not only in different patients, but also in different foci in the same patient. Irradiation doses in foci increase with an increasing in the administered activities of radiopharmaceuticals and, depending on the volume of foci, are within the following ranges: 1) in a case of “Lutaprost” – from 1.4 to 32 Gy (planned administered activity – 5 GBq), from 5.1 to 59 Gy (planned activity – 7.5 GBq), from 13 to 94 Gy (planned activity – 10 GBq); 2) in a case of "Arthroren" – from 17.5 to 74 Gy (planned administered activities – from 0.37 to 0.925 GBq); 3) in a case of “Gepatoren” – from 10.7 to 43 Gy (planned administered activities – from 1 to 3 GBq). Individual absorbed doses in organs at risk were estimated as well. Dose values in organs at risk also vary greatly between different patients and between different critical organs. These doses ranged from 0.01 to 7.4 Gy (39 patients), which is many times less than “commonly applied dose constraints” in radiotherapy.

Feasibility Study of Early Pressure Ulcer Detection Using Electrical Impedance Tomography

Pressure ulcers are chronic skin wounds with a particularly high morbidity and a relatively low recovery rate. At present, there is still a lack of a portable and visual real-time monitoring system for early pressure ulcers. Electrical impedance tomography (EIT) is a kind of functional imaging technique that can diagnose the healthy condition in the measured tissues by visualizing the distribution of bioelectrical impedance parameters. According to the principle of electrical characteristic of pressure ulcers, i.e., changes in physiological and pathological information of tissues caused by pressure lead to changes in electrical impedance of tissues, a non-invasive imaging system of pressure ulcers based on EIT with novel flexible electrode array is proposed. Two kinds of 3D finite element simulation models are constructed to discuss the measurement effects of the flexible array on simulated pressure ulcers at different locations, depths, and with different sizes. The effects of different imaging algorithms on imaging results are also compared and analyzed. Physical phantom model experiments and preliminary experiments on human skin are conducted. Both simulations and physical experiments indicate that EIT technique is excepted to realize a safer, portable, low-cost and visual real-time monitoring system for early pressure ulcers.

Extending arterial stiffness assessment along the circumference using beam-steered ARFI and wave-tracking: A proof-of-principle study in phantoms and ex vivo

BackgroundTo fully quantify arterial wall and plaque stiffness, acoustic radiation force impulse (ARFI)-induced wave-tracking along the entire vessel circumference is desired. However, attenuation and guided wave behavior in thin vessel walls limits wave-tracking to short trajectories. This study investigated the potential of beam-steered ARFI and wave-tracking to extend group velocity estimation over a larger proportion of the circumference compared to conventional 0° ARFI-induced wave-tracking. MethodsSeven vessel-mimicking polyvinyl alcohol cryogel phantoms with various dimensions and compositions and an ex vivo human carotid artery were imaged in a dedicated setup. For every 20⁰ phantom rotation, transverse group wave velocity measurements were performed with an Aixplorer Ultimate system and SL18–5 transducer using 0⁰/20⁰/−20⁰-angled ultrasound pushes. Transmural angular wave velocities were derived along 60⁰-trajectories. A 360⁰-angular velocity map was composed from the top-wall 60⁰-trajectories 0°-data, averaged over all physical phantom rotations (reference). For each phantom rotation, 360⁰-angular velocity maps were composed using 0°-data (0⁰-approach) or data from all angles (beam-steered approach). Percentages of rotations with visible waves and relative angular velocity errors compared to the reference map as function of the circumferential angle were determined for both approaches. ResultsReference 360°-angular velocity maps could be derived for all samples, representing their stiffness. Beam-steering decreased the proportion of the circumference where waves were untraceable by 20% in phantoms and 10% ex vivo, mainly at 0° push locations. Relative errors were similar for both approaches (phantoms: 10–15%, ex vivo: 15–35%). ConclusionBeam-steering enables wave-tracking along a higher proportion of the wall circumference than 0⁰ ARFI-induced wave-tracking.

ASSESSMENT OF PCXMC MONTE CARLO SIMULATIONS IN SLOT-SCANNING-BASED EXAMINATIONS: COMPARISON WITH IN-PHANTOM THERMOLUMINESCENT DOSIMETRY.

Slot-scanning technology is nowadays a valid solution for the follow-up of chronic musculoskeletal disorders on children and adolescent patients, but there is no commercial software designed for simulating this X-ray beam geometry. PC Program for X-ray Monte Carlo (PCXMC) is a widespread Monte Carlo software developed for dose computation in projection radiography. In this study, experimental measurements were performed to evaluate its applicability in examinations with a slit-beam device. Physical phantoms corresponding to an adult and a 5-y-old child with calibrated thermoluminescent dosemeters were used for experiments. Different simulation approaches were investigated. Differences between measured and calculated organ doses ranged from -95 to 67% and were statistically significant for almost all organs. For both patients, PCXMC underestimated the effective dose of about 25%. This study suggests that PCXMC is not suited for organ dose evaluation in examinations with slot-scanning devices. It is still a useful tool for effective dose estimation when a proper correction factor is applied.

A High-Fidelity Artificial Urological System for the Quantitative Assessment of Endoscopic Skills.

Minimally-invasive surgery is rapidly growing and has become a standard approach for many operations. However, it requires intensive practice to achieve competency. The current training often relies on animal organ models or physical organ phantoms, which do not offer realistic surgical scenes or useful real-time feedback for surgeons to improve their skills. Furthermore, the objective quantitative assessment of endoscopic skills is also lacking. Here, we report a high-fidelity artificial urological system that allows realistic simulation of endourological procedures and offers a quantitative assessment of the surgical performance. The physical organ model was fabricated by 3D printing and two-step polymer molding with the use of human CT data. The system resembles the human upper urinary tract with a high-resolution anatomical shape and vascular patterns. During surgical simulation, endoscopic videos are acquired and analyzed to quantitatively evaluate performance skills by a customized computer algorithm. Experimental results show significant differences in the performance between professional surgeons and trainees. The surgical simulator offers a unique chance to train endourological procedures in a realistic and safe environment, and it may also lead to a quantitative standard to evaluate endoscopic skills.

Learning CT projection denoising from adjacent views.

Many learning-based low-dose (LD) computed tomography (CT) imaging methods require large paired full- and low-dose datasets for training, which are usually unavailable in clinic. Whereas models trained on simulated data often face the generalization problem on real clinical data. To develop an unsupervised learning technique to acquire clean CT projection from its adjacent LD projections. Given a sequential LD projection set, the method extracts out the middle projection as the target and treats the rest ones as the input. The model is trained with the mean absolute error with proposed inter-view gradient constraint term, which helps to suppress outliers and preserve edges in the denoised projection. The simulated low-dose CT grand challenge dataset and a real physical torso phantom dataset were employed for experiment. The peak signal to noise ratio (PSNR) and structural similarity index measure (SSIM) were calculated for quantitative evaluation. In experiments with both the simulated and real datasets, visual comparisons reveal that the proposed method obtained images superior to unsupervised and supervised methods working in both image and projection domain. For numerical comparison, our method obtains larger SSIMs than other unsupervised methods at quarter and eighth dose levels. As for PSNR, our method obtains larger value at eighth dose whereas smaller value at quarter dose. The supervised models obtain better numerical results than all unsupervised models on simulated datasets. The proposed method can reduce the noise in CT projections effectively, making it an attractive tool for practical LDCT pre-processing.

Emphysema Quantifications With CT Scan: Assessing the Effects of Acquisition Protocols and Imaging Parameters Using Virtual Imaging Trials

CT scan has notable potential to quantify the severity and progression of emphysema in patients. Such quantification should ideally reflect the true attributes and pathologic conditions of subjects, not scanner parameters. To achieve such an objective, the effects of the scanner conditions need to be understood so the influence can be mitigated. How do CT scan imaging parameters affect the accuracy of emphysema-based quantifications and biomarkers? Twenty anthropomorphic digital phantoms were developed with diverse anatomic attributes and emphysema abnormalities informed by a real COPD cohort. The phantoms were input to a validated CT scan simulator (DukeSim), modeling a commercial scanner (Siemens Flash). Virtual images were acquired under various clinical conditions of dose levels, tube current modulations (TCM), and reconstruction techniques and kernels. The images were analyzed to evaluate the effects of imaging parameters on the accuracy of density-based quantifications (percent of lung voxels with HU<-950 [LAA-950] and 15th percentile of lung histogram HU [Perc15]) across varied subjects. Paired t tests were performed to explore statistical differences between any two imaging conditions. The most accurate imaging condition corresponded to the highest acquired dose (100 mAs) and iterative reconstruction (SAFIRE) with the smooth kernel of I31, where the measurement errors (difference between measurement and ground truth) were 35 ± 3 Hounsfield Units (HU),-4%± 5%, and 26 ± 10 HU (average ± SD), for the mean lung HU, LAA-950, and Perc15, respectively. Without TCM and at the I31 kernel, increase of dose (20 to 100 mAs) improved the lung mean absolute error (MAE) by 4.2 ± 2.3 HU (average ± SD). TCM did not contribute to a systematic improvement of lung MAE. The results highlight that although CT scan quantification is possible, its reliability is impacted by the choice of imaging parameters. The developed virtual imaging trial platform in this study enables comprehensive evaluation of CT scan methods in reliable quantifications, an effort that cannot be readily made with patient images or simplistic physical phantoms.

Clinical Efficacy of an Electronic Portal Imaging Device versus a Physical Phantom Tool for Patient-Specific Quality Assurance.

Pre-treatment patient-specific quality assurance (QA) is critical to prevent radiation accidents. The electronic portal imaging device (EPID) is a dose measurement tool with good resolution and a low volume-averaging effect. EPIbeam—an EPID-based portal dosimetry software—has been newly installed in three institutions in Korea. This study evaluated the efficacy of the EPID-based patient-specific QA tool versus the PTW729 detector (a previously used QA tool) based on gamma criteria and planning target volume (PTV). A significant difference was confirmed through the R statistical analysis software. The average gamma passing rates of PTW729 and EPIbeam were 98.73% and 99.60% on 3 mm/3% (local), 96.66% and 97.91% on 2 mm/2% (local), and 88.41% and 74.87% on 1 mm/1% (local), respectively. The p-values between them were 0.015 (3 mm/3%, local), 0.084 (2 mm/2%, local), and less than 0.01 (1 mm/1%, local). Further, the average gamma passing rates of PTW 729 and EPIbeam according to PTV size were 99.55% and 99.91% (PTV < 150 cm3) and 97.91% and 99.28% (PTV > 150 cm3), respectively. The p-values between them were 0.087 (PTV < 150 cm3) and 0.036 (PTV > 150 cm3). These results confirm that EPIbeam can be an effective patient-specific QA tool.

Engineering functional and anthropomorphic models for surgical training in interventional radiology: A state-of-the-art review.

Training medical students in surgical procedures and evaluating their performance are both necessary steps to ensure the safety and efficacy of surgeries. Traditionally, trainees practiced on live patients, cadavers or animals under the supervision of skilled physicians, but realistic anatomical phantom models have provided a low-cost alternative because of the advance of material technology that mimics multi-layer tissue structures. This setup provides safer and more efficient training. Many research prototypes of phantom models allow rapid in-house prototyping for specific geometries and tissue properties. The gel-based method and 3D printing-based method are two major methods for developing phantom prototypes. This study excluded virtual reality based technologies and focused on physical phantoms, total 189 works published between 2015 and 2020 on anatomical phantom prototypes made for interventional radiology were reviewed in terms of their functions and applications. The phantom prototypes were first categorized based on fabrication methods and then subcategorized based on the organ or body part they simulated; the paper is organized accordingly. Engineering specifications and applications were analyzed and summarized for each study. Finally, current challenges in the development of phantom models and directions for future work were discussed.

Geometry calibration for a dental cone-beam CT system with an offset detector

In this work, we propose a robust offline geometry calibration method for an offset-detector-based dental cone-beam computed tomography (CBCT) system. Although a host of geometric calibration methods for CT system have been developed, the calibration of dental CBCT systems remains to be one of the challenging cases due to the presence of an offset detector and due to the smaller system dimensions. We highlight the technical difficulties encountered in such dental CBCT geometric calibration, and provide an efficient and robust recipe for the accurate calibration. Through a particular design of the phantom, a heuristic estimation of the isocenter, and an enforcement of smooth change of the extracted geometric parameters, we show that the aforementioned challenges can be overcome. Fabrication errors that may exist in the geometry calibration phantom are also considered. We scanned two physical phantoms with a commercial dental CBCT for evaluating the proposed geometry calibration method. The results confirm that the proposed method can achieve an accurate and efficient geometry calibration in the offset-detector-based CBCT systems.