Three-Dimensional Quantitative Color-Coding Analysis of Hepatic Arterial Flow Change during Chemoembolization of Hepatocellular Carcinoma.

PURPOSE: To evaluate feasibility of using three-dimensional (3D) quantitative color-coding analysis (QCA) to quantify substasis endpoints after transcatheter arterial chemoembolization of hepatocellular carcinoma (HCC). MATERIALS AND METHODS: This single-institution prospective study included 20 patients with HCC who had undergone segmental or subsegmental transcatheter arterial chemoembolization between December 2015 and March 2017. The chemoembolization endpoint was a sluggish anterograde tumor-feeding arterial flow without residual tumor stains. Contrast medium bolus arrival time (BAT) was used as an indicator of arterial flow. BAT of the proper hepatic artery was obtained as a reference point. BATs of the proximal right lobar artery, proximal left lobar artery, and segmental artery that received embolization were analyzed before and after chemoembolization. Wilcoxon signed rank test was used to evaluate the difference between BATs before and after chemoembolization. RESULTS: BATs before and after chemoembolization of the segmental artery that received embolization were 0.47 seconds (interquartile range [IQR], 0.31-0.70 s) and 1.04 seconds (IQR, 0.78-2.01 s; P < .001), respectively. BATs before and after chemoembolization of the proximal left lobar hepatic artery (0.35 s [IQR, 0.11-0.55] and 0.13 s [IQR, 0.05-0.32], P = .025) and right lobar hepatic artery (0.23 s [IQR, 0.13-0.65] and 0.22 s [IQR, 0.08-0.39], P = .027) exhibited no significant change. CONCLUSIONS: 3D QCA is a feasible method for quantifying sluggish segmental arterial flow after transcatheter arterial chemoembolization in patients with HCC.

Finding the optimal deconvolution algorithm for MR perfusion in carotid stenosis: Correlations with angiographic cerebral circulation time.

PURPOSE: The aim of our study is to explore the impacts of different deconvolution algorithms on correlations between CBF, MTT, CBV, TTP, Tmax from MR perfusion (MRP) and angiography cerebral circulation time (CCT). METHODS: Retrospectively, 30 patients with unilateral carotid stenosis, and available pre-stenting MRP and angiography were included for analysis. All MRPs were conducted in a 1.5-T MR scanner. Standard singular value decomposition, block-circulant, and two delay-corrected algorithms were used as the deconvolution methods. All angiographies were obtained in the same bi-plane flat-detector angiographic machine. A contrast bolus of 12mL was administrated via angiocatheter at a rate of 8mL/s. The acquisition protocols were the same for all cases. CCT was defined as the difference between time to peak from the cavernous ICA and the parietal vein in lateral view. Pearson correlations were calculated for CCT and CBF, MTT, CBV, TTP, Tmax. RESULTS: The correlation between CCT and MTT was highest with Tmax (r=0.65), followed by MTT (r=0.60), CBF (r=-0.57), and TTP (r=0.33) when standard singular value decomposition was used. No correlation with CBV was noted. CONCLUSIONS: MRP using a singular value decomposition algorithm confirmed the feasibility of quantifying cerebral blood flow deficit in steno-occlusive disease within the angio-room. This approach might further improve patient safety by providing immediate cerebral hemodynamics without extraradiation and iodine contrast.

Measurement of quantifiable parameters by time-density curves in the elastase-induced aneurysm model: first results in the comparison of a flow diverter and a conventional aneurysm stent.

BACKGROUND: Quantifiable parameters to evaluate the effectiveness of flow diverters (FDs) are desirable. We measured time-density curves (TDCs) and calculated quantifiable parameters in the rabbit elastase-induced aneurysm model after stent (Neuroform [NF]) and FD (Pipeline embolisation device [PED]) treatment. METHODS: Sixteen rabbit elastase-induced aneurysms were treated with FD (n = 9) or NF (n = 5). Angiography was performed before and after treatment and TDCs were created. The time to peak (TTP), the full width at half maximum (FWHM) and the average slope of the curve which represent the inflow (IF) and outflow (OF) were calculated. RESULTS: Mean values before treatment were TTP = 0.8 s, FWHM = 1.2 s, IF = 153.5 and OF = -54.9. After PED treatment, the TTP of 1.8 s and FWHM of 47.8 s were extended. The IF was 31.2 and the OF was -11.5 and therefore delayed. The values after NF treatment (TTP = 1.1 s, FWHM = 1.8 s, IF = 152.9, OF = -33.2) changed only slightly. CONCLUSION: It was feasible to create TDCs in the rabbit aneurysm model. Parameters describing the haemodynamic effect of PED and NF were calculated and were different according to the type of device used. These parameters could possibly serve as predictive markers for aneurysm occlusion.

Objective assessment of transcatheter arterial chemoembolization angiographic endpoints: preliminary study of quantitative digital subtraction angiography.

PURPOSE: To explore the significance of quantitative digital subtraction angiography (DSA; Q-DSA) in the assessment of chemoembolization endpoints. MATERIALS AND METHODS: Twenty patients with hepatocellular carcinoma treated with chemoembolization were included in the study. All DSA series before and after chemoembolization were postprocessed with Q-DSA. The maximal enhancement and time to peak (TTP) were measured for several homologous anatomic landmarks, including the origin and embolized site of the tumor-feeding artery, parenchyma of the tumor, and ostia of the pre- and postprocedure catheter. The TTP, tumor blood supply time, and maximal enhancement of the time density curve (TDC) were analyzed. RESULTS: Of the 20 DSA series collected, 18 were successfully postprocessed. The TTPs of the landmarks before and after treatment were 3.60 seconds±1.02 and 3.57 seconds±0.78 at the ostia of the catheter, 3.91 seconds±1.01 and 4.09 seconds±1.14 at the origin site of the tumor-feeding artery, and 4.07 seconds±1.02 and 5.60 seconds±1.56 s the embolized site of the main tumor-feeding artery, respectively. Statistical differences were detected between pre- and postprocedural TTP of the embolized site of the feeding artery (P<.01), as well as between pre- and postprocedural tumor blood supply time (P<.01). The mean maximal TDC enhancements of selected tumor spots were 3.01 units±1.04 and 0.81 units±0.35 before and after the procedure (P<.01), respectively. CONCLUSIONS: Q-DSA may provide a feasible quantitative measurement in the assessment of chemoembolization endpoints.

BOSS: Bones, organs and skin shape model.

A virtual anatomical model of a patient can be a valuable tool for enhancing clinical tasks such as workflow automation, patient-specific X-ray dose optimization, markerless tracking, positioning, and navigation assistance in image-guided interventions. For these tasks, it is highly desirable that the patient's surface and internal organs are of high quality for any pose and shape estimate. At present, the majority of statistical shape models (SSMs) are restricted to a small number of organs or bones or do not adequately represent the general population. To address this, we propose a deformable human shape and pose model that combines skin, internal organs, and bones, learned from CT images. By modeling the statistical variations in a pose-normalized space using probabilistic PCA while also preserving joint kinematics, our approach offers a holistic representation of the body that can be beneficial for automation in various medical applications. In an interventional setup, our model could, for example, facilitate automatic system/patient positioning, organ-specific iso-centering, automated collimation or collision prediction. We assessed our model's performance on a registered dataset, utilizing the unified shape space, and noted an average error of 3.6 mm for bones and 8.8 mm for organs. By utilizing solely skin surface data or patient metadata like height and weight, we find that the overall combined error for bone-organ measurement is 8.68 mm and 8.11 mm, respectively. To further verify our findings, we conducted additional tests on publicly available datasets with multi-part segmentations, which confirmed the effectiveness of our model. In the diverse TotalSegmentator dataset, the errors for bones and organs are observed to be 5.10mm and 8.72mm, respectively. Our work shows that anatomically parameterized statistical shape models can be created accurately and in a computationally efficient manner. The proposed approach enables the construction of shape models that can be directly integrated into to various medical applications.

Accuracy of Dose-Saving Artificial-Intelligence-Based 3D Angiography (3DA) for Grading of Intracranial Artery Stenoses: Preliminary Findings.

BACKGROUND AND PURPOSE: Based on artificial intelligence (AI), 3D angiography (3DA) is a novel postprocessing algorithm for "DSA-like" 3D imaging of cerebral vasculature. Because 3DA requires neither mask runs nor digital subtraction as the current standard 3D-DSA does, it has the potential to cut the patient dose by 50%. The object was to evaluate 3DA's diagnostic value for visualization of intracranial artery stenoses (IAS) compared to 3D-DSA. MATERIALS AND METHODS: 3D-DSA datasets of IAS (nIAS = 10) were postprocessed using conventional and prototype software (Siemens Healthineers AG, Erlangen, Germany). Matching reconstructions were assessed by two experienced neuroradiologists in consensus reading, considering image quality (IQ), vessel diameters (VD1/2), vessel-geometry index (VGI = VD1/VD2), and specific qualitative/quantitative parameters of IAS (e.g., location, visual IAS grading [low-/medium-/high-grade] and intra-/poststenotic diameters [dintra-/poststenotic in mm]). Using the NASCET criteria, the percentual degree of luminal restriction was calculated. RESULTS: In total, 20 angiographic 3D volumes (n3DA = 10; n3D-DSA = 10) were successfully reconstructed with equivalent IQ. Assessment of the vessel geometry in 3DA datasets did not differ significantly from 3D-DSA (VD1: r = 0.994, p = 0.0001; VD2:r = 0.994, p = 0.0001; VGI: r = 0.899, p = 0.0001). Qualitative analysis of IAS location (3DA/3D-DSA:nICA/C4 = 1, nICA/C7 = 1, nMCA/M1 = 4, nVA/V4 = 2, nBA = 2) and the visual IAS grading (3DA/3D-DSA:nlow-grade = 3, nmedium-grade = 5, nhigh-grade = 2) revealed identical results for 3DA and 3D-DSA, respectively. Quantitative IAS assessment showed a strong correlation regarding intra-/poststenotic diameters (rdintrastenotic = 0.995, pdintrastenotic = 0.0001; rdpoststenotic = 0.995, pdpoststenotic = 0.0001) and the percentual degree of luminal restriction (rNASCET 3DA = 0.981; pNASCET 3DA = 0.0001). CONCLUSIONS: The AI-based 3DA is a resilient algorithm for the visualization of IAS and shows comparable results to 3D-DSA. Hence, 3DA is a promising new method that allows a considerable patient-dose reduction, and its clinical implementation would be highly desirable.

Learning-based occupational x-ray scatter estimation.

Objective.During x-ray-guided interventional procedures, the medical staff is exposed to scattered ionizing radiation caused by the patient. To increase the staff's awareness of the invisible radiation and monitor dose online, computational scatter estimation methods are convenient. However, such methods are usually based on Monte Carlo (MC) simulations, which are inherently computationally expensive. Yet, in the interventional environment, immediate feedback to the personnel is desirable.Approach. In this work, we propose deep neural networks to mitigate the computational effort of MC simulations. Our learning-based models consider detailed models of the (outer) patient shape and (inner) anatomy, additional objects in the room, and the x-ray tube spectrum to cover imaging settings encountered in real interventional settings. We investigate two cases of scatter prediction. First, we employ network architectures to estimate the full three-dimensional (3D) scatter distribution. Second, we investigate the prediction of two-dimensional (2D) intensity projections that facilitate the intra-procedural visualization.Main results.Depending on the dimensionality of the estimated scatter distribution and the network architecture, the mean relative error of each network is in the range of 12% and 14% compared to MC simulations. However, 3D scatter distributions can be estimated within 60 ms and 2D distributions within 15 ms.Significance.Overall, our method is suitable to support the online assessment of scattered ionizing radiation in the interventional environment and can help to lower the occupational radiation risk.

Robust learning-based x-ray image denoising-potential pitfalls, their analysis and solutions.

Purpose:Since guidance based on x-ray imaging is an integral part of interventional procedures, continuous efforts are taken towards reducing the exposure of patients and clinical staff to ionizing radiation. Even though a reduction in the x-ray dose may lower associated radiation risks, it is likely to impair the quality of the acquired images, potentially making it more difficult for physicians to carry out their procedures.Method:We present a robust learning-based denoising strategy involving model-based simulations of low-dose x-ray images during the training phase. The method also utilizes a data-driven normalization step-based on an x-ray imaging model-to stabilize the mixed signal-dependent noise associated with x-ray images. We thoroughly analyze the method's sensitivity to a mismatch in dose levels used for training and application. We also study the impact of differing noise models used when training for low and very low-dose x-ray images on the denoising results.Results:A quantitative and qualitative analysis based on acquired phantom and clinical data has shown that the proposed learning-based strategy is stable across different dose levels and yields excellent denoising results, if an accurate noise model is applied. We also found that there can be severe artifacts when the noise characteristics of the training images are significantly different from those in the actual images to be processed. This problem can be especially acute at very low dose levels. During a thorough analysis of our experimental results, we further discovered that viewing the results from the perspective of denoising via thresholding of sub-band coefficients can be very beneficial to get a better understanding of the proposed learning-based denoising strategy.Conclusion:The proposed learning-based denoising strategy provides scope for significant x-ray dose reduction without the loss of important image information if the characteristics of noise is accurately accounted for during the training phase.

XDose: toward online cross-validation of experimental and computational X-ray dose estimation.

PURPOSE: As the spectrum of X-ray procedures has increased both for diagnostic and for interventional cases, more attention is paid to X-ray dose management. While the medical benefit to the patient outweighs the risk of radiation injuries in almost all cases, reproducible studies on organ dose values help to plan preventive measures helping both patient as well as staff. Dose studies are either carried out retrospectively, experimentally using anthropomorphic phantoms, or computationally. When performed experimentally, it is helpful to combine them with simulations validating the measurements. In this paper, we show how such a dose simulation method, carried out together with actual X-ray experiments, can be realized to obtain reliable organ dose values efficiently. METHODS: A Monte Carlo simulation technique was developed combining down-sampling and super-resolution techniques for accelerated processing accompanying X-ray dose measurements. The target volume is down-sampled using the statistical mode first. The estimated dose distribution is then up-sampled using guided filtering and the high-resolution target volume as guidance image. Second, we present a comparison of dose estimates calculated with our Monte Carlo code experimentally obtained values for an anthropomorphic phantom using metal oxide semiconductor field effect transistor dosimeters. RESULTS: We reconstructed high-resolution dose distributions from coarse ones (down-sampling factor 2 to 16) with error rates ranging from 1.62 % to 4.91 %. Using down-sampled target volumes further reduced the computation time by 30 % to 60 %. Comparison of measured results to simulated dose values demonstrated high agreement with an average percentage error of under [Formula: see text] for all measurement points. CONCLUSIONS: Our results indicate that Monte Carlo methods can be accelerated hardware-independently and still yield reliable results. This facilitates empirical dose studies that make use of online Monte Carlo simulations to easily cross-validate dose estimates on-site.

Deep action learning enables robust 3D segmentation of body organs in various CT and MRI images.

In this study, we propose a novel point cloud based 3D registration and segmentation framework using reinforcement learning. An artificial agent, implemented as a distinct actor based on value networks, is trained to predict the optimal piece-wise linear transformation of a point cloud for the joint tasks of registration and segmentation. The actor network estimates a set of plausible actions and the value network aims to select the optimal action for the current observation. Point-wise features that comprise spatial positions (and surface normal vectors in the case of structured meshes), and their corresponding image features, are used to encode the observation and represent the underlying 3D volume. The actor and value networks are applied iteratively to estimate a sequence of transformations that enable accurate delineation of object boundaries. The proposed approach was extensively evaluated in both segmentation and registration tasks using a variety of challenging clinical datasets. Our method has fewer trainable parameters and lower computational complexity compared to the 3D U-Net, and it is independent of the volume resolution. We show that the proposed method is applicable to mono- and multi-modal segmentation tasks, achieving significant improvements over the state-of-the-art for the latter. The flexibility of the proposed framework is further demonstrated for a multi-modal registration application. As we learn to predict actions rather than a target, the proposed method is more robust compared to the 3D U-Net when dealing with previously unseen datasets, acquired using different protocols or modalities. As a result, the proposed method provides a promising multi-purpose segmentation and registration framework, particular in the context of image-guided interventions.

X-Ray Scatter Estimation Using Deep Splines.

X-ray scatter compensation is a very desirable technique in flat-panel X-ray imaging and cone-beam computed tomography. State-of-the-art U-net based scatter removal approaches yielded promising results. However, as there are no physics' constraints applied to the output of the U-Net, it cannot be ruled out that it yields spurious results. Unfortunately, in the context of medical imaging, those may be misleading and could lead to wrong conclusions. To overcome this problem, we propose to embed B-splines as a known operator into neural networks. This inherently constrains their predictions to well-behaved and smooth functions. In a study using synthetic head and thorax data as well as real thorax phantom data, we found that our approach performed on par with U-net when comparing both algorithms based on quantitative performance metrics. However, our approach not only reduces runtime and parameter complexity, but we also found it much more robust to unseen noise levels. While the U-net responded with visible artifacts, the proposed approach preserved the X-ray signal's frequency characteristics.

Data-driven estimation of noise variance stabilization parameters for low-dose x-ray images.

PURPOSE: Denoising x-ray images corrupted by signal-dependent mixed noise is usually approached either by considering noise statistics directly or by using noise variance stabilization (NVS) techniques. An advantage of the latter is that the noise variance can be stabilized to a known constant throughout the image, facilitating the application of denoising algorithms designed for the removal of additive Gaussian noise. A well-performing NVS is the generalized Anscombe transform (GAT). To calculate the GAT, the system gain as well as the variance of electronic noise are required. Unfortunately, these parameters are difficult to predict from the x-ray tube settings in clinical practice, because the system gain observed at the detector depends on the beam hardening caused by the patient. MATERIALS AND METHODS: We propose a data-driven method for estimating the parameters required to carry out an NVS using the GAT. It utilizes the energy compaction property of the discrete cosine transform to obtain the NVS parameters using a robust regression approach relying on a linear Poisson-Gaussian model. The method has been experimentally validated with respect to beam hardening as well as denoising performance for different dose and scatter levels. RESULTS: Across a range of low-dose x-ray settings, the proposed robust regression approach has estimated both system gain and electronic noise level with an average error of only 4.2%. When used to perform a GAT followed by the denoising of low-dose x-ray images, performance gains of 5% for peak-signal-to-noise ratio and 4% for structural similarity index can be obtained. CONCLUSION: The parameters needed to calculate the GAT can be estimated efficiently and robustly using a data-driven approach. The improved parameter estimation method facilitates a more accurate GAT-based NVS and, hence, better denoising of low-dose x-ray images when algorithms designed for additive Gaussian noise are applied.

Appearance Learning for Image-Based Motion Estimation in Tomography.

In tomographic imaging, anatomical structures are reconstructed by applying a pseudo-inverse forward model to acquired signals. Geometric information within this process is usually depending on the system setting only, i.e., the scanner position or readout direction. Patient motion therefore corrupts the geometry alignment in the reconstruction process resulting in motion artifacts. We propose an appearance learning approach recognizing the structures of rigid motion independently from the scanned object. To this end, we train a siamese triplet network to predict the reprojection error (RPE) for the complete acquisition as well as an approximate distribution of the RPE along the single views from the reconstructed volume in a multi-task learning approach. The RPE measures the motion-induced geometric deviations independent of the object based on virtual marker positions, which are available during training. We train our network using 27 patients and deploy a 21-4-2 split for training, validation and testing. In average, we achieve a residual mean RPE of 0.013mm with an inter-patient standard deviation of 0.022mm. This is twice the accuracy compared to previously published results. In a motion estimation benchmark the proposed approach achieves superior results in comparison with two state-of-the-art measures in nine out of twelve experiments. The clinical applicability of the proposed method is demonstrated on a motion-affected clinical dataset.

Evaluation of an Artificial Intelligence-Based 3D-Angiography for Visualization of Cerebral Vasculature.

PURPOSE: The three-dimensional digital subtraction angiography (3D DSA) technique is the current standard and is based on both mask and fill runs to enable the subtraction technique. Artificial intelligence (AI)-based 3D angiography (3DA) was developed to reduce radiation dosage because only one contrast-enhanced run of the C­arm system is required for reconstruction of DSA-like 3D volumes. The aim was the evaluation of this algorithm regarding its diagnostic information. METHODS: 3D DSA datasets without pathologic findings were reconstructed both with subtraction technique and with the AI-based algorithm. Corresponding reconstructions were evaluated by 2 neuroradiologists with respect to image quality (IQ), visualization of major segments of the circle of Willis (ICA = C4-C7; OphA; ACA = A1-A2, MCA = M1-M2; VA = V4; BA; AICA; SUCA; PCA = P1-P2), identifiability of perforators (lenticulostriate/thalamoperforating arteries) and vessel diameters (ICA = C4; MCA = M1; BA; PCA = P1). RESULTS: In total 15 datasets were successfully reconstructed as 3D DSA and 3DA with diagnostic image quality. All major segments of the circle of Willis and perforators were comparably visualized with 3DA. Quantitative analysis of vessel diameters in 3D DSA and 3DA datasets was equivalent and did not show relevant differences (rICA = 0.901, p = 0.001; rM1 = 0.951, p = 0.001; rBA = 0.906, p = 0.001; rP1 = 0.991, p = 0.001). CONCLUSIONS: The use of 3DA demonstrated reliable visualization of cerebral vasculature with respect to quantitative and qualitative parameters. Therefore, 3DA is a promising method that might help to reduce patient radiation.

Simultaneous reconstruction of multiple stiff wires from a single X-ray projection for endovascular aortic repair.

PURPOSE: Endovascular repair of aortic aneurysms (EVAR) can be supported by fusing pre- and intraoperative data to allow for improved navigation and to reduce the amount of contrast agent needed during the intervention. However, stiff wires and delivery devices can deform the vasculature severely, which reduces the accuracy of the fusion. Knowledge about the 3D position of the inserted instruments can help to transfer these deformations to the preoperative information. METHOD: We propose a method to simultaneously reconstruct the stiff wires in both iliac arteries based on only a single monoplane acquisition, thereby avoiding interference with the clinical workflow. In the available X-ray projection, the 2D course of the wire is extracted. Then, a virtual second view of each wire orthogonal to the real projection is estimated using the preoperative vessel anatomy from a computed tomography angiography as prior information. Based on the real and virtual 2D wire courses, the wires can then be reconstructed in 3D using epipolar geometry. RESULTS: We achieve a mean modified Hausdorff distance of 4.2 mm between the estimated 3D position and the true wire course for the contralateral side and 4.5 mm for the ipsilateral side. CONCLUSION: The accuracy and speed of the proposed method allow for use in an intraoperative setting of deformation correction for EVAR.

A Semi-Automated Usability Evaluation Framework for Interactive Image Segmentation Systems.

For complex segmentation tasks, the achievable accuracy of fully automated systems is inherently limited. Specifically, when a precise segmentation result is desired for a small amount of given data sets, semi-automatic methods exhibit a clear benefit for the user. The optimization of human computer interaction (HCI) is an essential part of interactive image segmentation. Nevertheless, publications introducing novel interactive segmentation systems (ISS) often lack an objective comparison of HCI aspects. It is demonstrated that even when the underlying segmentation algorithm is the same throughout interactive prototypes, their user experience may vary substantially. As a result, users prefer simple interfaces as well as a considerable degree of freedom to control each iterative step of the segmentation. In this article, an objective method for the comparison of ISS is proposed, based on extensive user studies. A summative qualitative content analysis is conducted via abstraction of visual and verbal feedback given by the participants. A direct assessment of the segmentation system is executed by the users via the system usability scale (SUS) and AttrakDiff-2 questionnaires. Furthermore, an approximation of the findings regarding usability aspects in those studies is introduced, conducted solely from the system-measurable user actions during their usage of interactive segmentation prototypes. The prediction of all questionnaire results has an average relative error of 8.9%, which is close to the expected precision of the questionnaire results themselves. This automated evaluation scheme may significantly reduce the resources necessary to investigate each variation of a prototype's user interface (UI) features and segmentation methodologies.

Physics-driven learning of x-ray skin dose distribution in interventional procedures.

PURPOSE: Radiation doses accumulated during very complicated image-guided x-ray procedures have the potential to cause stochastic, but also deterministic effects, such as skin rashes or even hair loss. To monitor and reduce radiation-related risks to patients' skin, x-ray imaging devices are equipped with online air kerma monitoring components. Traditionally, such measurements have been used to estimate skin entrance dose by (a) estimating air kerma at the interventional reference point (IRP), (b) forward projecting the dose distribution, and (c) considering a backscatter factor among other correction factors. Unfortunately, the complicated interaction between incident x-ray photons, secondary electrons, and skin tissue cannot be properly accounted for by assuming a linear relationship between forward projected air kerma and a backscatter factor. Gold standard skin dose models are therefore determined using Monte Carlo (MC) techniques. However, MC simulations are computationally complex in general and possible acceleration mainly depends on the employed hardware and variance reduction techniques. To obtain reliable and fast dose estimates, we propose to combine MC-based simulations with learning-based methods. METHODS: The basic idea of our method is to approximate the radiation physics to calculate a first-order exposure estimate quickly. This initial estimate is then refined using prior knowledge derived from MC simulations. To this end, the primary photon propagation inside a voxelized patient model is estimated using a less accurate but fast photon ray casting (RC) simulation based on the Beer-Lambert law. The results of the RC simulation are then fed into a convolutional neural network (CNN), which maps the propagation of primary photons to the dose deposition inside the patient model. Additionally, the patient model itself including anatomy and material properties, such as mass density and mass energy-absorption coefficients, are fed into the CNN as well. The CNN is trained using smoothed results of MC simulations as output and RC simulations of identical imaging settings and patient models as input. RESULTS: In total, 163 MC and associated RC simulations are carried out for the head, thorax, abdomen, and pelvis in three different voxel phantoms. We used 10 8 or 10 9 primarily emitted photons sampled from a 125 kV peak voltage spectrum, respectively. Edge-preserving smoothing (EPS) is applied to reduce (a) general stochastic uncertainties and (b) stochastic uncertainty concerning MC simulations of less primary photons. The CNN is trained using seven imaging settings of the abdomen in a single phantom. Testing its performance on the remaining datasets, the CNN is capable of estimating skin dose with an error of below 10% for the majority of test cases. CONCLUSION: The combination of deep neural networks and MC simulation of particle physics has the potential to decrease the computational complexity of accurate skin dose estimation. The proposed approach can provide dose distributions in under one second when running on high-end hardware. On lower cost hardware, it took up to 2 min to arrive at the same result. This makes our approach applicable in high-end environments as well as in budget solutions. Furthermore, the number of primary photons only affects the training time, while the execution time is independent of the number of primary photons.

An analytical approach for the simulation of realistic low-dose fluoroscopic images.

PURPOSE: The quality of X-ray images plays an important role in computer-assisted interventions. Although learning-based denoising techniques have been shown to be successful in improving the image quality, they often rely on pairs of associated low- and high-dose X-ray images that are usually not possible to acquire at different dose levels in a clinical scenario. Moreover, since data variation is an important requirement for learning-based methods, the use of phantom data alone may not be sufficient. A possibility to address this issue is a realistic simulation of low-dose images from their related high-dose counterparts. METHOD: We introduce a novel noise simulation method based on an X-ray image formation model. The method makes use of the system parameters associated with low- and high-dose X-ray image acquisitions, such as system gain and electronic noise, to preserve the image noise characteristics of low-dose images. RESULTS: We have compared several corresponding regions of the associated real and simulated low-dose images-obtained from two different imaging systems-visually as well as statistically, using a two-sample Kolmogorov-Smirnov test at 5% significance. In addition to being visually similar, the hypothesis that the corresponding regions-from 80 pairs of real and simulated low-dose regions-belonging to the same distribution has been accepted in 81.43% of the cases. CONCLUSION: The results suggest that the simulated low-dose images obtained using the proposed method are almost indistinguishable from real low-dose images. Since extensive calibration procedures required in previous methods can be avoided using the proposed approach, it allows an easy adaptation to different X-ray imaging systems. This in turn leads to an increased diversity of the training data for potential learning-based methods.

A machine learning pipeline for internal anatomical landmark embedding based on a patient surface model.

PURPOSE: With the recent introduction of fully assisting scanner technologies by Siemens Healthineers (Erlangen, Germany), a patient surface model was introduced to the diagnostic imaging device market. Such a patient representation can be used to automate and accelerate the clinical imaging workflow, manage patient dose, and provide navigation assistance for computed tomography diagnostic imaging. In addition to diagnostic imaging, a patient surface model has also tremendous potential to simplify interventional imaging. For example, if the anatomy of a patient was known, a robotic angiography system could be automatically positioned such that the organ of interest is positioned in the system's iso-center offering a good and flexible view on the underlying patient anatomy quickly and without any additional X-ray dose. METHOD: To enable such functionality in a clinical context with sufficiently high accuracy, we present an extension of our previous patient surface model by adding internal anatomical landmarks associated with certain (main) bones of the human skeleton, in particular the spine. We also investigate different approaches to positioning of these landmarks employing CT datasets with annotated internal landmarks as training data. The general pipeline of our proposed method comprises the following steps: First, we train an active shape model using an existing avatar database and segmented CT surfaces. This stage also includes a gravity correction procedure, which accounts for shape changes due to the fact that the avatar models were obtained in standing position, while the CT data were acquired with patients in supine position. Second, we match the gravity-corrected avatar patient surface models to surfaces segmented from the CT datasets. In the last step, we derive the spatial relationships between the patient surface model and internal anatomical landmarks. RESULT: We trained and evaluated our method using cross-validation using 20 datasets, each containing 50 internal landmarks. We further compared the performance of four different generalized linear models' setups to describe the positioning of the internal landmarks relative to the patient surface. The best mean estimation error over all the landmarks was achieved using lasso regression with a mean error of [Formula: see text]. CONCLUSION: Considering that interventional X-ray imaging systems can have detectors covering an area of about [Formula: see text] ([Formula: see text]) at iso-center, this accuracy is sufficient to facilitate automatic positioning of the X-ray system.

Symmetry prior for epipolar consistency.

PURPOSE: For a perfectly plane symmetric object, we can find two views-mirrored at the plane of symmetry-that will yield the exact same image of that object. In consequence, having one image of a plane symmetric object and a calibrated camera, we automatically have a second, virtual image of that object if the 3-D location of the symmetry plane is known. METHODS: We propose a method for estimating the symmetry plane from a set of projection images as the solution of a consistency maximization based on epipolar consistency. With the known symmetry plane, we can exploit symmetry to estimate in-plane motion by introducing the X-trajectory that can be acquired with a conventional short-scan trajectory by simply tilting the acquisition plane relative to the plane of symmetry. RESULTS: We inspect the symmetry plane estimation on a real scan of an anthropomorphic human head phantom and show the robustness using a synthetic dataset. Further, we demonstrate the advantage of the proposed method for estimating in-plane motion using the acquired projection data. CONCLUSION: Symmetry breakers in the human body are widely used for the detection of tumors or strokes. We provide a fast estimation of the symmetry plane, robust to outliers, by computing it directly from a set of projections. Further, by coupling the symmetry prior with epipolar consistency, we overcome inherent limitations in the estimation of in-plane motion.