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.

Comparison of methods for sensitivity correction in Talbot-Lau computed tomography.

PURPOSE: In Talbot-Lau X-ray phase contrast imaging, the measured phase value depends on the position of the object in the measurement setup. When imaging large objects, this may lead to inhomogeneous phase contributions within the object. These inhomogeneities introduce artifacts in tomographic reconstructions of the object. METHODS: In this work, we compare recently proposed approaches to correct such reconstruction artifacts. We compare an iterative reconstruction algorithm, a known operator network and a U-net. The methods are qualitatively and quantitatively compared on the Shepp-Logan phantom and on the anatomy of a human abdomen. We also perform a dedicated experiment on the noise behavior of the methods. RESULTS: All methods were able to reduce the specific artifacts in the reconstructions for the simulated and virtual real anatomy data. The results show method-specific residual errors that are indicative for the inherently different correction approaches. While all methods were able to correct the artifacts, we report a different noise behavior. CONCLUSION: The iterative reconstruction performs very well, but at the cost of a high runtime. The known operator network shows consistently a very competitive performance. The U-net performs slightly worse, but has the benefit that it is a general-purpose network that does not require special application knowledge.

Cephalogram synthesis and landmark detection in dental cone-beam CT systems.

Due to the lack of a standardized 3D cephalometric analysis methodology, 2D cephalograms synthesized from 3D cone-beam computed tomography (CBCT) volumes are widely used for cephalometric analysis in dental CBCT systems. However, compared with conventional X-ray film based cephalograms, such synthetic cephalograms lack image contrast and resolution, which impairs cephalometric landmark identification. In addition, the increased radiation dose applied to acquire the scan for 3D reconstruction causes potential health risks. In this work, we propose a sigmoid-based intensity transform that uses the nonlinear optical property of X-ray films to increase image contrast of synthetic cephalograms from 3D volumes. To improve image resolution, super resolution deep learning techniques are investigated. For low dose purpose, the pixel-to-pixel generative adversarial network (pix2pixGAN) is proposed for 2D cephalogram synthesis directly from two cone-beam projections. For landmark detection in the synthetic cephalograms, an efficient automatic landmark detection method using the combination of LeNet-5 and ResNet50 is proposed. Our experiments demonstrate the efficacy of pix2pixGAN in 2D cephalogram synthesis, achieving an average peak signal-to-noise ratio (PSNR) value of 33.8 with reference to the cephalograms synthesized from 3D CBCT volumes. Pix2pixGAN also achieves the best performance in super resolution, achieving an average PSNR value of 32.5 without the introduction of checkerboard or jagging artifacts. Our proposed automatic landmark detection method achieves 86.7% successful detection rate in the 2 mm clinical acceptable range on the ISBI Test1 data, which is comparable to the state-of-the-art methods. The method trained on conventional cephalograms can be directly applied to landmark detection in the synthetic cephalograms, achieving 93.0% and 80.7% successful detection rate in 4 mm precision range for synthetic cephalograms from 3D volumes and 2D projections, respectively.

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 Learning-Based ECG-Free Cardiac Navigation for Multi-Dimensional and Motion-Resolved Continuous Magnetic Resonance Imaging.

For the clinical assessment of cardiac vitality, time-continuous tomographic imaging of the heart is used. To further detect e.g., pathological tissue, multiple imaging contrasts enable a thorough diagnosis using magnetic resonance imaging (MRI). For this purpose, time-continous and multi-contrast imaging protocols were proposed. The acquired signals are binned using navigation approaches for a motion-resolved reconstruction. Mostly, external sensors such as electrocardiograms (ECG) are used for navigation, leading to additional workflow efforts. Recent sensor-free approaches are based on pipelines requiring prior knowledge, e.g., typical heart rates. We present a sensor-free, deep learning-based navigation that diminishes the need for manual feature engineering or the necessity of prior knowledge compared to previous works. A classifier is trained to estimate the R-wave timepoints in the scan directly from the imaging data. Our approach is evaluated on 3-D protocols for continuous cardiac MRI, acquired in-vivo and free-breathing with single or multiple imaging contrasts. We achieve an accuracy of > 98% on previously unseen subjects, and a well comparable image quality with the state-of-the-art ECG-based reconstruction. Our method enables an ECG-free workflow for continuous cardiac scans with simultaneous anatomic and functional imaging with multiple contrasts. It can be potentially integrated without adapting the sampling scheme to other continuous sequences by using the imaging data for navigation and reconstruction.

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.

Known Operator Learning Enables Constrained Projection Geometry Conversion: Parallel to Cone-Beam for Hybrid MR/X-Ray Imaging.

X-ray imaging is a wide-spread real-time imaging technique. Magnetic Resonance Imaging (MRI) offers a multitude of contrasts that offer improved guidance to interventionalists. As such simultaneous real-time acquisition and overlay would be highly favorable for image-guided interventions, e.g., in stroke therapy. One major obstacle in this setting is the fundamentally different acquisition geometry. MRI k -space sampling is associated with parallel projection geometry, while the X-ray acquisition results in perspective distorted projections. The classical rebinning methods to overcome this limitation inherently suffers from a loss of resolution. To counter this problem, we present a novel rebinning algorithm for parallel to cone-beam conversion. We derive a rebinning formula that is then used to find an appropriate deep neural network architecture. Following the known operator learning paradigm, the novel algorithm is mapped to a neural network with differentiable projection operators enabling data-driven learning of the remaining unknown operators. The evaluation aims in two directions: First, we give a profound analysis of the different hypotheses to the unknown operator and investigate the influence of numerical training data. Second, we evaluate the performance of the proposed method against the classical rebinning approach. We demonstrate that the derived network achieves better results than the baseline method and that such operators can be trained with simulated data without losing their generality making them applicable to real data without the need for retraining or transfer learning.

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.

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.

Pitfalls in interventional X-ray organ dose assessment-combined experimental and computational phantom study: application to prostatic artery embolization.

PURPOSE: With X-ray radiation protection and dose management constantly gaining interest in interventional radiology, novel procedures often undergo prospective dose studies using anthropomorphic phantoms to determine expected reference organ-equivalent dose values. Due to inherent uncertainties, such as impact of exact patient positioning, generalized geometry of the phantoms, limited dosimeter positioning options, and composition of tissue-equivalent materials, these dose values might not allow for patient-specific risk assessment. Therefore, first the aim of this study is to quantify the influence of these parameters on local X-ray dose to evaluate their relevance in the assessment of patient-specific organ doses. Second, this knowledge further enables validating a simulation approach, which allows employing physiological material models and patient-specific geometries. METHODS: Phantom dosimetry experiments using MOSFET dosimeters were conducted reproducing imaging scenarios in prostatic arterial embolization (PAE). Associated organ-equivalent dose of prostate, bladder, colon, and skin was determined. Dose deviation induced by possible small displacements of the patient was reproduced by moving the X-ray source. Dose deviation induced by geometric and material differences was investigated by analyzing two different commonly used phantoms. We reconstructed the experiments using Monte Carlo (MC) simulations, a reference male geometry, and different material properties to validate simulations and experiments against each other. RESULTS: Overall, MC-simulated organ dose values are in accordance with the measured ones for the majority of cases. Marginal displacements of X-ray source relative to the phantoms lead to deviations of 6-135% in organ dose values, while skin dose remains relatively constant. Regarding the impact of phantom material composition, underestimation of internal organ dose values by 12-20% is prevalent in all simulated phantoms. Skin dose, however, can be estimated with low deviation of 1-8% at least for two materials. CONCLUSIONS: Prospective reference dose studies might not extend to precise patient-specific dose assessment. Therefore, online organ dose assessment tools, based on advanced patient modeling and MC methods, are desirable.