Low-dose X-ray Imaging Research Articles

Organic Semiconductor Single Crystals for X-ray Imaging.

Low-dose, high-resolution X-ray imaging is vital for medical diagnostics and material/device analyses. Current X-ray imagers are dominated by expensive inorganic materials via high-temperature solid processes (up to 1700 °C, e.g., CsI:Tl) with heavy metal elements. It is essential to search for new materials as X-ray imagers with low growth temperature, low cost, high sensitivity, along with high chemical and environmental stability. Here, 9,10-diphenylanthracene (9,10-DPA) single crystals are used as a representative model, which are grown via low-temperature solution processes, exhibiting intense X-ray radioluminescence with ultrahigh photon-conversion efficiency, ultrafast response and high sensitivity. The resolution of devices based on organic crystals exceeds 20.00 lp mm-1 . Meanwhile the crystals exhibit high cycle performance under X-ray irradiation and environmental stability. This study demonstrates that organic semiconductors have potential use in low-cost, high-sensitivity and low-dose X-ray imaging systems.

Heterojunction structures for reduced noise in large-area and sensitive perovskite x-ray detectors.

Polycrystalline perovskites can be readily fabricated into large areas using solution depositions; however, they suffer from large dark currents that are tens to hundreds times higher than industrially relevant values, limiting their application in low-dose x-ray detection. Here, we show that the application of a heterojunction structure into polycrystalline films significantly reduces the dark current density by more than 200 times to subnanoampere per square centimeter without reducing the sensitivity of the detectors. The heterojunction perovskite films are formed by laminating several membrane films filled with perovskites of different bandgaps. A gradient bandgap is formed during annealing. The detectors have a lowest detectable dose rate of 13.8 ± 0.29 nGyair s−1 for 40-keV x-ray and can conduct dynamic x-ray imaging at a low-dose rate of 32.2 nGyair s−1. Simulation and experimental analysis show that the heterojunction is tolerant of halide diffusion and can be stable for over 15 years.

Fully vacuum-sealed addressable nanowire cold cathode flat-panel x-ray source

Addressable flat-panel x-ray source has important applications in next-generation low-dose x-ray imaging technology. In this study, a fully vacuum-sealed addressable flat-panel x-ray source has been fabricated using a ZnO nanowire cold cathode. The x-ray source was formed by assembling gated ZnO nanowire field emitters and a molybdenum thin film transmission anode target. Pulsed and addressable x-ray emission was achieved by applying voltages to the extraction gate. Radiation dose rates of 235 nGy/s were achieved when the device was operated at 21 kV anode voltage and 140 V gate voltage. Projection x-ray images of objects were obtained by the flat-panel x-ray source. Our results demonstrated the feasibility of a fully addressable cold cathode flat-panel x-ray source.

Physical bases for determination of scattering kernels from incomplete data in grid-less X-ray imaging

A mathematical model for the determination of X-ray scattering kernels’ shapes based on incomplete simulation or measurement data was introduced and tested using a mathematical phantom. The model is originally intended for low-dose X-ray imaging without anti-scatter grids. The proposed model fits different kinds of symmetrical and asymmetrical scattering kernels in different tissues well enough for practical applications. Kernels asymmetry is mostly caused by irradiation of the object near the boundaries of different tissues. The model describes a variety of asymmetrical kernels by proposed “sectoral” members. Application of the proposed model in scattering compensation procedure reduces resulting error up to 50 % for “wide” scattering kernels.

Theoretical derivation and clinical dose-response quantification of a unified multi-activation (UMA) model of cell survival from a logistic equation.

Objective:To theoretically derive a unified multiactivation (UMA) model of cell survival after ionising radiation that can accurately assess doses and responses in radiotherapy and X-ray imaging.Methods:A unified formula with only two parameters in fitting of a cell survival curve (CSC) is first derived from an assumption that radiation-activated cell death pathways compose the first- and second-order reaction kinetics. A logit linear regression of CSC data is used for precise determination of the two model parameters. Intrinsic radiosensitivity, biologically effective dose (BED), equivalent dose to the traditional 2 Gy fractions (EQD2), tumour control probability, normal-tissue complication probability, BED50 and steepness (Γ50) at 50% of tumour control probability (or normal-tissue complication probability) are analytical functions of the model and treatment (or imaging) parameters.Results:The UMA model has almost perfectly fit typical CSCs over the entire dose range with R2≥0.99. Estimated quantities for stereotactic body radiotherapy of early stage lung cancer and the skin reactions from X-ray imaging agree with clinical results.Conclusion:The proposed UMA model has theoretically resolved the catastrophes of the zero slope at zero dose for multiple target model and the bending curve at high dose for the linear quadratic model. More importantly, it analytically predicts dose–responses to various dose–fraction schemes in radiotherapy and to low dose X-ray imaging based on these preclinical CSCs.Advances in knowledge:The discovery of a unified formula of CSC over the entire dose range may reveal a common mechanism of the first- and second-order reaction kinetics among multiple CD pathways activated by ionising radiation at various dose levels.

Ultra-Dense Denoising Network: Application to Cardiac Catheter-Based X-Ray Procedures.

Reducing radiation dose in cardiac catheter-based X-ray procedures increases safety but also image noise and artifacts. Excessive noise and artifacts can compromise vital image information, which can affect clinical decision-making. Developing more effective X-ray denoising methodologies will be beneficial to both patients and healthcare professionals by allowing imaging at lower radiation dose without compromising image information. This paper proposes a framework based on a convolutional neural network (CNN), namely Ultra-Dense Denoising Network (UDDN), for low-dose X-ray image denoising. To promote feature extraction, we designed a novel residual block which establishes a solid correlation among multiple-path neural units via abundant cross connections in its representation enhancement section. Experiments on synthetic additive noise X-ray data show that the UDDN achieves statistically significant higher peak signal-to-noise ratio (PSNR) and structural similarity index measure (SSIM) than other comparative methods. We enhanced the clinical adaptability of our framework by training using normally-distributed noise and tested on clinical data taken from procedures at St. Thomas' hospital in London. The performance was assessed by using local SNR and by clinical voting using ten cardiologists. The results show that the UDDN outperforms the other comparative methods and is a promising solution to this challenging but clinically impactful task.

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.

Low-dose real-time X-ray imaging with nontoxic double perovskite scintillators

X-rays are widely used in probing inside information nondestructively, enabling broad applications in the medical radiography and electronic industries. X-ray imaging based on emerging lead halide perovskite scintillators has received extensive attention recently. However, the strong self-absorption, relatively low light yield and lead toxicity of these perovskites restrict their practical applications. Here, we report a series of nontoxic double-perovskite scintillators of Cs2Ag0.6Na0.4In1-yBiyCl6. By controlling the content of the heavy atom Bi3+, the X-ray absorption coefficient, radiative emission efficiency, light yield and light decay were manipulated to maximise the scintillator performance. A light yield of up to 39,000 ± 7000 photons/MeV for Cs2Ag0.6Na0.4In0.85Bi0.15Cl6 was obtained, which is much higher than that for the previously reported lead halide perovskite colloidal CsPbBr3 (21,000 photons/MeV). The large Stokes shift between the radioluminescence (RL) and absorption spectra benefiting from self-trapped excitons (STEs) led to a negligible self-absorption effect. Given the high light output and fast light decay of this scintillator, static X-ray imaging was attained under an extremely low dose of ∼1 μGyair, and dynamic X-ray imaging of finger bending without a ghosting effect was demonstrated under a low-dose rate of 47.2 μGyair s−1. After thermal treatment at 85 °C for 50 h followed by X-ray irradiation for 50 h in ambient air, the scintillator performance in terms of the RL intensity and X-ray image quality remained almost unchanged. Our results shed light on exploring highly competitive scintillators beyond the scope of lead halide perovskites, not only for avoiding toxicity but also for better performance.

Low-dose X-ray imaging may increase the risk of neurodegenerative diseases

The hypothesis presented here explores the possibility that X-ray imaging commonly used in dental practices may be a shared risk factor for sporadic dementias and motor-neuron diseases. As the evidence will suggest, the brain is ill-equipped to manage the intrusion of low-dose ionizing radiation (IR) beyond that which is naturally occurring. When the brain’s antioxidant defenses are overwhelmed by IR, it produces an abundance of reactive oxygen species (ROS) that can lead to oxidative stress, mitochondrial dysfunction, loss of synaptic plasticity, altered neuronal structure and microvascular impairment that have been identified as early signs of neurodegeneration in Alzheimer’s disease, Parkinson’s, amyotrophic lateral sclerosis, vascular dementia and other diseases that progressively damage the brain and central nervous system. Although genes play a role in all outcomes, the focus here will be on the non-genetic processes at work. Common assumptions regarding the risks of low-dose IR will be addressed, such as: 1) comparing rapid, repeated bursts of man-made IR sent exclusively into the head to equivalent amounts of head-to-toe background IR over longer periods of time; 2) whether epidemiological studies that dismiss concerns regarding low-dose IR due to lack of evidence it causes cancer, heritable mutations or shortened life spans also apply to neurodegeneration; and 3) why even radiation-resistant neurons can be severely impacted by IR exposure, due to IR-induced injury to the processes they need to function. Also considered will be the difficulty of distinguishing the effects of dental X-ray exposure from similarly low amounts of background IR and where to find the evidence that they may, in fact, be responsible for neurodegeneration. Finally, the long-standing belief that whatever risks are inherent in dental radiography must be undertaken for the sake of oral health is challenged on two counts: 1) while dentists continue to drape their patients in lead-lined aprons, the most effective IR safety precautions are often ignored; and 2) there is an alternative dental imaging technology that does not use IR. While the thrust of this article will be on dental radiation and will touch on how age, gender, X-ray equipment and protocols may increase risk, chiropractic radiographs also will be considered because they focus exclusively on the central nervous system. If X-ray imaging is found to be associated with neurodegeneration, the risk-versus-benefit must be reevaluated, every means of reducing exposure implemented and imaging protocols revised.

Chest X-ray enhancement to interpret pneumonia malformation based on fuzzy soft set and Dempster–Shafer theory of evidence

Image enhancement algorithms are commonly used to increase the contrast and visual quality of low-dose x-ray images. This paper proposes an automated enhancement method using soft fuzzy sets with a new decision-making scheme based on Dempster-Shafer theory of evidence for the visual interpretation of pneumonia malformation in low-dose x-ray images, called as XEFSDS. The XEFSDS model first generates an original source x-ray image into a complementary image, then each original and complement image is applied to the characterized image object and background areas of fuzzy space. The S-function is utilized to define fuzzy soft sets for the classification of gray level ambiguity in both images, and hence a decision criterion via Dempster-Shafer approach and fuzzy interval has been adapted to discriminate uncertainties on the pixel intensity and the spatial information. Modified membership grade operations have been performed on each object/background area, and Werner’s AND/OR operator (an aggregation operator) has been utilized to build a new membership function from two modified membership functions. Finally, an enhanced image is obtained from the new membership function via defuzzification. Experiments on different pneumonia X-ray images demonstrate that the XEFSDS scheme produces better results than the existing methods. To show the advantages of the XEFSDS scheme, we have executed a segmentation based examination on enhanced image for the detection of pneumonia malformation as well as abnormal lobe (lobar pneumonia) or bronchopneumonia.

Low-dose chest X-ray image super-resolution using generative adversarial nets with spectral normalization

Chest X-ray (CXR) imaging is one of the most widely-used and cost-effective technology for chest screening and diagnosis of Pulmonary diseases. An always concerned improvement about CXR is to reduce X-ray radiation while achieving ultra-high quality imaging with fine structural details since CXR involves ionizing radiation and tolerance of different populations. In this paper, we present a supervised generative adversarial nets approach to accurately recover high-resolution (HR) CXR images from low-resolution (LR) counterparts while keep pathological invariance. Specifically, the auxiliary label information is introduced to constrain the feature generation to attack the potential risk of pathological variance. Then, spectral normalization is designed to control the performance of discriminative network with the guarantee of theoretical demonstration in controlling Lipschitz bound of discriminator. Results from quantitative and qualitative evaluations demonstrate that our method delivers more authentic improvement for CXR super-resolution (SR) compared to recent state-of-the-art methods. The proposed method has outperformed average 13.0%, 12.2% in FSIM and 13.7%, 12.5% in MSIM on two datasets, respectively. Besides, the index of generative performance GAN-train and GAN-test have achieved average increment 9.3% and 10.5% on CXR2 dataset. Subjective evaluation on SR CXR has outperformed average score 0.425 and 0.525 in terms of pathological invariance and acceptability, respectively.

Relative position of sacral base in the pelvis and its correlation with spino-pelvic parameters.

To investigate the relationship between relative location of the sacral base and spinal alignment in standing healthy adult volunteers. One hundred seventy-two volunteers (men = 83, mean age = 39.3years [20-70], women = 89, mean age = 39.6years [20-62]) with no history of spinal disease were imaged using a low-dose biplanar slot-scanning 3D X-ray imaging system. A circle was drawn around three points: cranial vertex of the iliac crest (A), caudal vertex of the ischium (B), and anterior vertex of the pubis. Pelvic height (PH) was defined as the diameter (A-B). A tangent line perpendicular to PH (C) was drawn by passing through (A). Sacral height (SH) was defined as the distance between (C) and the center of the sacral base parallel to PH. Relative SH (rSH) was calculated as SH/PH × 100. Mean (SD) rSH was 18.3 ± 3.2 (men 20.0 ± 2.9, women 16.7 ± 2.6). rSH significantly positively correlated with thoracic kyphosis (r = 0.20, p < 0.05), lumbar lordosis (r = 0.28, p < 0.05), pelvic incidence (r = 0.28, p < 0.05), and sacral slope (r = 0.32, p < 0.0001), and significantly negatively correlated with pelvic thickness (r = - 0.66, p < 0.0001). rSH did not correlate with pelvic tilt. The center of the sacral base is normally located 3.8 ± 0.8cm caudal to the cranial vertex of the iliac crest. The sacral base was located more caudally in men than in women, regardless of age. The more caudal the sacral base, the angle of the spino-pelvic parameters (TK, LL, PI, SS) progressively increases along with a decrease in the sacro-acetabular distance (Pth). Pelvic tilt did not correlate with the location of the sacrum.

A statistical evaluation of eye-tracking data of screening mammography: Effects of expertise and experience on image reading

Screening mammography has been widely used over the last few decades to detect breast cancers at an early stage using low-dose X-ray imaging. Being able to detect breast lesions early is highly beneficial for the patients as it makes treatment more successful and increases the chance of recovery. However, lesion detection is prone to false negative and false positive results due to the subtlety or low prevalence of the lesions, which compromises patient safety. It is critical to understand human perception and interpretation of images and use such understanding to develop ways to help increase the accuracy of cancer detection. In this paper, we present a large-scale eye-tracking study where 1,568 eye movement trials of expert breast radiologists, trainee radiologists, and physicists were recorded while assessing a large number of mammogram images. First, we statistically analysed viewers’ gaze behaviour based on their mean fixation duration. Results demonstrated that there was no difference amongst experts, and that trainees’ mean fixation duration was significantly shorter than experts and physicists’ fixation duration was significantly longer than that of the experts. Second, we evaluated spatial gaze deployment of the three viewer groups and quantified the difference between groups. Results showed that trainees and physicists deviated from where expert radiologists looked in images, and trainees performed better than physicists in terms of focusing on regions of interest. Finally, we investigated the saccadic behaviour of viewers and illustrated the differences between viewer groups in terms of saccade amplitudes and orientations.

Low-dose X-ray imaging for paediatric trauma comes of age

Low-dose X-ray imaging for paediatric trauma comes of age

Artifact correction in low‐dose dental CT imaging using Wasserstein generative adversarial networks

In recent years, health risks concerning high-dose x-ray radiation have become a major concern in dental computed tomography (CT) examinations. Therefore, adopting low-dose computed tomography (LDCT) technology has become a major focus in the CT imaging field. One of these LDCT technologies is downsampling data acquisition during low-dose x-ray imaging processes. However, reducing the radiation dose can adversely affect CT image quality by introducing noise and artifacts in the resultant image that can compromise diagnostic information. In this paper, we propose an artifact correction method for downsampling CT reconstruction based on deep learning. We used clinical dental CT data with low-dose artifacts reconstructed by conventional filtered back projection (FBP) as inputs to a deep neural network and corresponding high-quality labeled normal-dose CT data during training. We trained a generative adversarial network (GAN) with Wasserstein distance (WGAN) and mean squared error (MSE) loss, called m-WGAN, to remove artifacts and obtain high-quality CT dental images in a clinical dental CT examination environment. The experimental results confirmed that the proposed algorithm effectively removes low-dose artifacts from dental CT scans. In addition, we showed that the proposed method is efficient for removing noise from low-dose CT scan images compared to existing approaches. We compared the performances of the general GAN, convolutional neural networks, and m-WGAN. Through quantitative and qualitative analysis of the results, we concluded that the proposed m-WGAN method resulted in better artifact correction performance preserving the texture in dental CT scanning. The image quality evaluation metrics indicated that the proposed method effectively improves image quality when used as a postprocessing technique for dental CT images. To the best of our knowledge, this work is the first deep learning architecture used with a commercial cone-beam dental CT scanner. The artifact correction performance was rigorously evaluated and demonstrated to be effective. Therefore, we believe that the proposed algorithm represents a new direction in the research area of low-dose dental CT artifact correction.

Development and evaluation of a new methodology for the fast generation of patient-specific Finite Element models of the buttock for sitting-acquired deep tissue injury prevention

The occurrence and management of Pressure Ulcers remain a major issue for patients with reduced mobility and neurosensory loss despite significant improvement in the prevention methods. These injuries are caused by biological cascades leading from a given mechanical loading state in tissues to irreversible tissue damage. Estimating the internal mechanical conditions within loaded soft tissues has the potential of improving the management and prevention of PU. Several Finite Element models of the buttock have therefore been proposed based on either MRI or CT-Scan data. However, because of the limited availability of MRI or CT-Scan systems and of the long segmentation time, all studies in the literature include the data of only one individual. Yet the inter-individual variability can’t be overlooked when dealing with patient specific estimation of internal tissue loading. As an alternative, this contribution focuses on the combined use of low-dose biplanar X-ray images, B-mode ultrasound images and optical scanner acquisitions in a non-weight-bearing sitting posture for the fast generation of patient-specific FE models of the buttock. Model calibration was performed based on Ischial Tuberosity sagging. Model evaluation was performed by comparing the simulated contact pressure with experimental observations on a population of 6 healthy subjects. Analysis of the models confirmed the high inter-individual variability of soft tissue response (maximum Green Lagrange shear strains of 213 ± 101% in the muscle). This methodology opens the way for investigating inter-individual factors influencing the soft tissue response during sitting and for providing tools to assess PU risk.

Poisson-Gaussian Noise Analysis and Estimation for Low-Dose X-ray Images in the NSCT Domain.

The noise distribution of images obtained by X-ray sensors in low-dosage situations can be analyzed using the Poisson and Gaussian mixture model. Multiscale conversion is one of the most popular noise reduction methods used in recent years. Estimation of the noise distribution of each subband in the multiscale domain is the most important factor in performing noise reduction, with non-subsampled contourlet transform (NSCT) representing an effective method for scale and direction decomposition. In this study, we use artificially generated noise to analyze and estimate the Poisson–Gaussian noise of low-dose X-ray images in the NSCT domain. The noise distribution of the subband coefficients is analyzed using the noiseless low-band coefficients and the variance of the noisy subband coefficients. The noise-after-transform also follows a Poisson–Gaussian distribution, and the relationship between the noise parameters of the subband and the full-band image is identified. We then analyze noise of actual images to validate the theoretical analysis. Comparison of the proposed noise estimation method with an existing noise reduction method confirms that the proposed method outperforms traditional methods.

Knee medial and lateral contact forces in a musculoskeletal model with subject-specific contact point trajectories

Contact point (CP) trajectory is a crucial parameter in estimating medial/lateral tibio-femoral contact forces from the musculoskeletal (MSK) models. The objective of the present study was to develop a method to incorporate the subject-specific CP trajectories into the MSK model. Ten healthy subjects performed 45 s treadmill gait trials. The subject-specific CP trajectories were constructed on the tibia and femur as a function of extension-flexion using low-dose bi-plane X-ray images during a quasi-static squat. At each extension-flexion position, the tibia and femur CPs were superimposed in the three directions on the medial side, and in the anterior-posterior and proximal-distal directions on the lateral side to form the five kinematic constraints of the knee joint. The Lagrange multipliers associated to these constraints directly yielded the medial/lateral contact forces. The results from the personalized CP trajectory model were compared against the linear CP trajectory and sphere-on-plane CP trajectory models which were adapted from the commonly used MSK models. Changing the CP trajectory had a remarkable impact on the knee kinematics and changed the medial and lateral contact forces by 1.03 BW and 0.65 BW respectively, in certain subjects. The direction and magnitude of the medial/lateral contact force were highly variable among the subjects and the medial-lateral shift of the CPs alone could not determine the increase/decrease pattern of the contact forces. The suggested kinematic constraints are adaptable to the CP trajectories derived from a variety of joint models and those experimentally measured from the 3D imaging techniques.

3D map of theranostic nanoparticles distribution in mice brain and liver by means of X-ray Phase Contrast Tomography

The word “theranostic” derives from the fusion of two terms: therapeutic and diagnostic. It is a promising research field that aims to develop innovative therapies with high target specificity by exploiting the therapeutic and diagnostic properties, in particular for metal-based nanoparticles (NPs) developed to erase cancer. In the framework of a combined research program on low dose X-ray imaging and theranostic nanoparticles (NPs), high resolution Phase-Contrast Tomography images of mice organs injected with gadolinium and gold-NPs were acquired at the European Synchrotron Radiation Facility (ESRF). Both compounds are good X-ray contrast agents due to their high attenuation coefficient with respect to biological tissues, especially immediately above K-edge energy. X-ray tomography is a powerful non-invasive technique to image the 3D vasculature network in order to detect abnormalities. Phase contrast methods provide more detailed anatomical information with higher discrimination among soft tissues. We present the images of mice liver and brain injected with gold and gadolinium NPs, respectively. We discuss different image processing methods used aiming at enhancing the accuracy on localizing nanoparticles.

Can portable tomosynthesis improve the diagnostic value of bedside chest X-ray in the intensive care unit? A proof of concept study

Portable bedside chest X-ray (CXR) is an important and frequently used tool in the intensive care unit (ICU). Unfortunately, the diagnostic value of portable CXR is often low due to technical limitations and suboptimal patient positioning. Additionally, abnormalities in the chest may be hidden on the projection image by overlapping anatomy and devices such as endotracheal tubes, lines and catheters. Digital tomosynthesis (DTS) can solve the problem of anatomical overlap. In DTS, several low-dose X-ray images from different angles are acquired and subsequently used by a reconstruction algorithm to compute section images along planes parallel to the detector. However, a portable device to be used for portable bedside chest DTS is not on the market yet. In this work, we discuss modifications to a portable X-ray device to enable portable DTS and illustrate the potential of portable DTS to improve the diagnostic value of bedside CXR in the ICU. A simulation, based on computed tomography scans, is presented. Our experiments comparing portable DTS with conventional bedside CXR showed a substantially improved detection of pneumothorax and other abnormalities.