Severe Artifacts Research Articles

Using the Navier-Cauchy equation for motion estimation in dynamic imaging

<p style='text-indent:20px;'>Tomographic image reconstruction is well understood if the specimen being studied is stationary during data acquisition. However, if this specimen changes its position during the measuring process, standard reconstruction techniques can lead to severe motion artefacts in the computed images. Solving a dynamic reconstruction problem therefore requires to model and incorporate suitable information on the dynamics in the reconstruction step to compensate for the motion.</p><p style='text-indent:20px;'>Many dynamic processes can be described by partial differential equations which thus could serve as additional information for the purpose of motion compensation. In this article, we consider the Navier-Cauchy equation which characterizes small elastic deformations and serves, for instance, as a simplified model for respiratory motion. Our goal is to provide a proof-of-concept that by incorporating the deformation fields provided by this PDE, one can reduce the respective motion artefacts in the reconstructed image. To this end, we solve the Navier-Cauchy equation prior to the image reconstruction step using suitable initial and boundary data. Then, the thus computed deformation fields are incorporated into an analytic dynamic reconstruction method to compute an image of the unknown interior structure. The feasibility is illustrated with numerical examples from computerized tomography.</p>

Aliasing Artifacts Suppression in MIMO Millimeter-Wave Imaging Systems Based on K-Space Analysis

This paper focuses on dealing with aliasing artifacts in millimeter-wave imaging systems. To address the challenges and complexities of designing multi-static arrays, sparse arrangements are utilized in practical imaging systems. Depending on the arrangement of antennas, such arrays may present severe aliasing artifacts. Some techniques have already been introduced to tackle such artifacts and enhance the image quality. These techniques may either distort the target image by causing some loss in parts of the image, impose high computational/system costs or suffer from the lack of generalizability. One of these techniques uses non-uniform arrays. Such arrays are designed by optimizing the point spread function (PSF). The PSF is a suitable representative of the system’s performance provided that the system is shift-invariant. Since non-uniform arrays are shift-variant, these techniques are unreliable, especially in the case of off-center targets. As a consequence, there are no frameworks to design non-uniform arrays on a reliable basis to suppress aliasing artifacts in the whole region of interest (ROI). To address this issue, based on the k-space analysis of aliasing in multi-static arrays and non-uniform Fourier transform concept, we propose a small random displacement to be applied to antennas’ location in sparse periodic arrays. This displacement prevents the aliasing components from coherently accumulating. Nevertheless, due to the imposed non-uniformity, it is impossible to use FFT-based algorithms in this case. A subsequent resampling or the use of non-uniform FFT can address this problem. The provided simulations and experimental measurements show that the proposed method can efficiently lower the peak level of aliasing artifacts by around 3dB compared to sparse periodic arrays. So, the proposed method can be used as an effective way to suppress aliasing artifacts in real-time multi-static millimeter-wave imaging systems.

DuDoDR-Net: Dual-domain data consistent recurrent network for simultaneous sparse view and metal artifact reduction in computed tomography.

Sparse-view computed tomography (SVCT) aims to reconstruct a cross-sectional image using a reduced number of x-ray projections. While SVCT can efficiently reduce the radiation dose, the reconstruction suffers from severe streak artifacts, and the artifacts are further amplified with the presence of metallic implants, which could adversely impact the medical diagnosis and other downstream applications. Previous methods have extensively explored either SVCT reconstruction without metallic implants, or full-view CT metal artifact reduction (MAR). The issue of simultaneous sparse-view and metal artifact reduction (SVMAR) remains under-explored, and it is infeasible to directly apply previous SVCT and MAR methods to SVMAR which may yield non-ideal reconstruction quality. In this work, we propose a dual-domain data consistent recurrent network, called DuDoDR-Net, for SVMAR. Our DuDoDR-Net aims to reconstruct an artifact-free image by recurrent image domain and sinogram domain restorations. To ensure the metal-free part of acquired projection data is preserved, we also develop the image data consistent layer (iDCL) and sinogram data consistent layer (sDCL) that are interleaved in our recurrent framework. Our experimental results demonstrate that our DuDoDR-Net is able to produce superior artifact-reduced results while preserving the anatomical structures, that outperforming previous SVCT and SVMAR methods, under different sparse-view acquisition settings.

The effect of metallic dental restorations and implants in causing patient discomfort and artefacts during magnetic resonance imaging of the head and neck.

Patients with metallic dental objects in their oral cavity are often associated with discomfort or artefacts during magnetic resonance imaging (MRI) of the head and neck. This study was conducted to evaluate the effects of metallic dental objects in causing patient discomfort and artefacts during MRI of the head and neck region. This is an observational study including 538 participants from various specialized MRI centres in Kozhikode. MRI records of 538 participants who underwent head and neck MRI as a part of their medical treatment were collected. With the help of questionnaire and clinical examination, the type of metallic dental object in the patient's oral cavity was identified. After examining the MRI images, four categories of artefacts were established. In MRI brain, artefacts in three different sequences were also studied. The significance of the difference between proportions was analysed by the Chi-square test. Artefacts were found in 65 cases with metallic dental objects but none of the participants had experienced any kind of discomfort. The artefact formation significantly depended on the type of metallic dental restoration in the patient's oral cavity. Orthodontic braces and titanium bone plate created severe artefacts in head and neck MRI making interpretation almost impossible. In the MRI brain, diffusion-weighted imaging is most sensitive to artefact followed by fluid-attenuated inversion recovery. Artefacts were found in 65 cases with metallic dental restorations but none of the participants had experienced any kind of discomfort.

AS-Net: Fast Photoacoustic Reconstruction With Multi-Feature Fusion From Sparse Data

Photoacoustic (PA) imaging is a biomedical imaging modality capable of acquiring high-contrast images of optical absorption at depths much greater than traditional optical imaging techniques. However, practical instrumentation and geometry limit the number of available acoustic sensors surrounding the imaging target, which results in the sparsity of sensor data. Conventional PA image reconstruction methods give severe artifacts when they are applied directly to the sparse PA data. In this paper, we firstly propose to employ a novel signal processing method to make sparse PA raw data more suitable for the neural network, concurrently speeding up image reconstruction. Then we propose Attention Steered Network (AS-Net) for PA reconstruction with multi-feature fusion. AS-Net is validated on different datasets, including simulated photoacoustic data from fundus vasculature phantoms and experimental data from <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">in vivo</i> fish and mice. Notably, the method is also able to eliminate some artifacts present in the ground truth for in vivo data. Results demonstrated that our method provides superior reconstructions at a faster speed.

End-to-End Deep Learning CT Image Reconstruction for Metal Artifact Reduction

Metal artifacts are common in CT-guided interventions due to the presence of metallic instruments. These artifacts often obscure clinically relevant structures, which can complicate the intervention. In this work, we present a deep learning CT reconstruction called iCTU-Net for the reduction of metal artifacts. The network emulates the filtering and back projection steps of the classical filtered back projection (FBP). A U-Net is used as post-processing to refine the back projected image. The reconstruction is trained end-to-end, i.e., the inputs of the iCTU-Net are sinograms and the outputs are reconstructed images. The network does not require a predefined back projection operator or the exact X-ray beam geometry. Supervised training is performed on simulated interventional data of the abdomen. For projection data exhibiting severe artifacts, the iCTU-Net achieved reconstructions with SSIM = 0.970±0.009 and PSNR = 40.7±1.6. The best reference method, an image based post-processing network, only achieved SSIM = 0.944±0.024 and PSNR = 39.8±1.9. Since the whole reconstruction process is learned, the network was able to fully utilize the raw data, which benefited from the removal of metal artifacts. The proposed method was the only studied method that could eliminate the metal streak artifacts.

Bowel Peristalsis Artifact on Dual-Energy CT: In Vitro Study on the Influence of Different Dual-Energy CT Platforms and Enteric Contrast Agents.

BACKGROUND. The value of dual-energy CT (DECT) for bowel wall assessment is increasingly recognized. Although technical improvements reduce peristalsis artifact in conventional CT, the effects of peristalsis on DECT image reconstructions remain poorly studied. OBJECTIVE. The purpose of this study was to evaluate the influence of different DECT scanners and enteric contrast agents on the severity of bowel peristalsis artifact in vitro. METHODS. To simulate bowel peristalsis, a 3-cm-diameter corrugated hollow tube representing the bowel was oscillated constantly in the z-axis within a larger water-filled cylinder. The bowel was serially filled with air, water, and iodinated or experimental dark contrast material and scanned on four different DECT platforms (spectral detector, rapid peak kilovoltage switching, split filter, and dual source) to reconstruct 120-kVp-like and iodine images. Two readers rated each image reconstruction for artifact severity from 0 (none) to 3 (severe) and recorded the degree to which iodine images depicted bowel wall hyperattenuation on 120-kVp-like images as artifactual. Artifact severity scores were compared by ANOVA with Bonferroni correction. RESULTS. Interrater agreement on artifact scores was excellent (intraclass correlation coefficient, 0.82 [95% CI, 0.79-0.84]). For 120-kVp-like images, mean peristalsis artifact scores were lower (all p < .001) for split-filter (1.47) and dual-source (1.86) scanners than for spectral-detector (2.58) and rapid-kilovoltage-switching (2.74) scanners. Compared with those on 120-kVp images, peristalsis artifacts on iodine images were less severe for spectral-detector (score, 1.03; p < .001) and rapid-kilovoltage-switching (2.09; p < .001) systems but more severe for dual-source (2.77; p < .001) and split-filter (2.62; p < .001) systems. Peristalsis artifact was rated less severe with experimental dark bowel contrast medium (score, 1.79) than with other bowel contrast agents (all p < .001). Iodine images helped identify bowel wall hyperattenuation as artifactual in 94.7% of reviewed cases for spectral-detector and 40.7% of cases for rapid-kilovoltage-switching scanners. CONCLUSION. For spectral-detector and rapid-kilovoltage-switching DECT, iodine images minimize peristalsis artifact, but for dual-source and split-filter DECT, mixed 120-kVp-like images are preferred. Compared with iodinated contrast material and water, experimental dark bowel contrast material reduces peristalsis artifact. CLINICAL IMPACT. Knowledge of the preferred images for reducing peristalsis artifact can lessen the effect of peristalsis on clinical DECT interpretation. Dark enteric contrast agents, when they become clinically available, may further reduce the effects of peristalsis.

Virtual unrolling technology based on terahertz computed tomography

Due to the shape diversity and the occlusion problem, non-invasive inspection of cultural relic with special-shaped structures has always been the bottleneck of the existing imaging technology. We propose a virtual unrolling technique that can extract content from structures similar to folded paper and scroll, exploiting the three-dimensional imaging capability provided by terahertz computed tomography (THz CT). The method performs THz CT imaging on the object and obtains several slice images of its cross section. Considering the morphological properties of the slice image, the cross-sectional profile of the object is obtained by an improved single-pixel ridgeline extraction algorithm after suppressing the severe metal artifact on the slice images with our segmentation and inpainting-based scheme. And then a match-graft approach based on dynamic exemplar (MGDE) is developed to stitch lines of different layers into the final image. To illustrate, the occluding text content is successfully extracted from a mimic oil painting scroll samples with a thickness of sub-wavelength level. The method makes it possible for THz non-destructive inspection of similar structure that widely exist in industry, security, and especially heritage conservation.

3D ring artifacts removal algorithm combined low-rank tensor decomposition with spatial-sequential total variation regularization and its application in phase-contrast microtomography.

High-resolution synchrotron radiation X-ray phase contrast microtomography (PC-μCT) images often suffer from severe ring artifacts, which are mainly caused by undesirable responses of detector elements. In the medical imaging field, the existence of ring artifacts can lead to degraded visual quality and can directly affect diagnosis accuracy. Thus, removing or at least effectively reducing ring artifacts is indispensable. The existing ring artifacts removal algorithms mainly focus on two-dimensional (matrix-based) priors, and these algorithms fail to consider correlations hidden in sequential computed tomography (CT) images. This paper proposed a novel three-dimensional (tensor-based) ring artifacts removal algorithm for synchrotron radiation X-ray PC-μCT images. In the sinogram domain, ring artifacts manifest as vertical stripe artifacts. From an image decomposition perspective, a degraded sinogram can be decomposed into a stripe artifacts component and an underlying clean sinogram component. The proposed algorithm is designed to detect and remove stripe artifacts from a degraded sinogram by fully identifying the characteristics of the two components. Specifically, for the stripe artifacts component, tensor Tucker decomposition is used to describe its low-rank character. For the underlying clean sinogram component, spatial-sequential total variation regularization is adopted to enhance the piecewise smoothness. Moreover, the Frobenius norm term is further used to model Gaussian noise. An efficient augmented Lagrange multiplier method is designed to solve the proposed optimization model. The proposed method is evaluated utilizing both simulations and real data containing different ring artifacts patterns. In the simulations, the human chest CT images are used for evaluating the proposed method. We compare the peak signal-to-noise ratio (PSNR), structural similarity (SSIM), and mean absolute error (MAE) results of our algorithm with the Naghia's method, the RRRTV method, the wavelet-FFT method, and the SDRSD-GIF method. The proposed method was also evaluated on real data from rat liver samples and rat tooth samples. Our proposed method outperforms the competing methods in terms of both qualitative and quantitative evaluation results. Additionally, the 3D visualization results were presented to make the ring artifacts removal effect more intuitive. The experimental results on simulations and real data clearly demonstrated that the proposed algorithm can significantly improve the quality of PC-μCT images compared with the existing popular algorithms, and it has great potential to promote the application of high-resolution imaging for visualizing biological tissues.

Technical evaluation of the cone‐beam computed tomography imaging performance of a novel, mobile, gantry‐based X‐ray system for brachytherapy

PurposeA novel, mobile cone‐beam computed tomography (CBCT) system for image‐guided adaptive brachytherapy was recently deployed at our hospital as worldwide first site. Prior to the device's clinical operation, a profound characterization of its imaging performance was conducted. This was essential to optimize both the imaging workflow and image quality for achieving the best possible clinical outcomes. We present the results of our investigations.MethodsThe novel CBCT‐system features a ring gantry with 121 cm clearance as well as a 43.2 × 43.2 cm2 flat‐panel detector, and is controlled via a tablet‐personal computer (PC). For evaluating its imaging performance, the geometric reproducibility as well as imaging fidelity, computed tomography (CT)‐number accuracy, uniformity, contrast‐noise‐ratio (CNR), noise characteristics, and spatial resolution as fundamental image quality parameters were assessed. As dose metric the weighted cone‐beam dose index (CBDIw) was measured. Image quality was evaluated using standard quality assurance (QA) as well as anthropomorphic upper torso and breast phantoms. Both in‐house and manufacturer protocols for abdomen, pelvis, and breast imaging were examined.ResultsUsing the in‐house protocols, the QA phantom scans showed altogether a high image quality, with high CT‐number accuracy (R 2 > 0.97) and uniformity (<12 Hounsfield Unit (HU) cupping), reasonable noise and imaging fidelity, and good CNR at bone–tissue transitions of up to 28:1. Spatial resolution was strongly limited by geometric instabilities of the device. The breast phantom scans fulfilled clinical requirements, whereas the abdomen and pelvis scans showed severe artifacts, particularly at air/bone–tissue transitions.ConclusionWith the novel CBCT‐system, achieving a high image quality appears possible in principle. However, adaptations of the standard protocols, performance enhancements in image reconstruction referring to artifact reductions, as well as the extinction of geometric instabilities are imperative.

Model-based reconstruction for enhanced x-ray CT of dense tri-structural isotropic particles.

Tri-structural isotropic (TRISO) fuel particles are a key component of next generation nuclear fuels. Using x-ray computed tomography (CT) to characterize TRISO particles is challenging because of the strong attenuation of the x-ray beam by the uranium core, leading to severe photon starvation in a substantial fraction of the measurements. Furthermore, the overall acquisition time for a high-resolution CT scan can be very long when using conventional laboratory-based x-ray systems and reconstruction algorithms. Specifically, when analytic methods such as the Feldkamp-Davis-Kress (FDK) algorithm are used for reconstruction, it results in severe streak artifacts and noise in the corresponding 3D volume, which makes subsequent analysis of the particles challenging. In this paper, we develop and apply model-based image reconstruction (MBIR) algorithms to improve the quality of CT reconstructions for TRISO particles to facilitate better characterization. We demonstrate that the proposed MBIR algorithms can significantly suppress artifacts with minimal pre-processing compared to conventional approaches. We also demonstrate that the proposed MBIR approach can obtain high-quality reconstruction compared to the FDK approach even when using a fraction of the typically acquired measurements, thereby enabling dramatically faster measurement times for TRISO particles.

Intracranial Lesion Detection and Artifact Characterization: Comparative Study of Susceptibility and T2*-Weighted Imaging in Dogs and Cats.

Susceptibility-weighted imaging (SWI), an MRI sequence for the detection of hemorrhage, allows differentiation of paramagnetic and diamagnetic substances based on tissue magnetic susceptibility differences. The three aims of this retrospective study included a comparison of the number of areas of signal void (ASV) between SWI and T2*-weighted imaging (T2*WI), differentiation of hemorrhage and calcification, and investigation of image deterioration by artifacts. Two hundred twelve brain MRIs, 160 dogs and 52 cats, were included. The sequences were randomized and evaluated for presence/absence and numbers of ASV and extent of artifacts causing image deterioration by a single, blinded observer. In cases with a CT scan differentiation of paramagnetic (hemorrhagic) and diamagnetic (calcification) lesions was made, SWI was performed to test correct assignment using the Hounsfield Units. Non-parametric tests were performed to compare both sequences regarding detection of ASV and the effect of artifacts on image quality. The presence of ASV was found in 37 SWI sequences and 34 T2*WI sequences with a significant increase in ASV only in dogs >5 and ≤ 15 kg in SWI. The remaining weight categories showed no significance. CT examination was available in 11 cases in which 81 ASV were found. With the use of phase images, 77 were classified as paramagnetic and none as diamagnetic. A classification was not possible in four cases. At the level of the frontal sinus, significantly more severe artifacts occurred in cats and dogs (dogs, p < 0.001; cats, p = 0.001) in SWI. The frontal sinus artifact was significantly less severe in brachycephalic than non-brachycephalic dogs in both sequences (SWI, p < 0.001; T2*WI, p < 0.001). In conclusion, with the advantages of better detection of ASV in SWI compared with T2*WI and the opportunity to differentiate between paramagnetic and diamagnetic origin in most cases, SWI is generally recommended for dogs. Frontal sinus conformation appears to be a limiting factor in image interpretation.

A GMRT Narrowband vs. Wideband Analysis of the ACT−CL J0034.4+0225 Field Selected from the ACTPol Cluster Sample

Low frequency radio observations of galaxy clusters are a useful probe of the non-thermal intracluster medium (ICM), through observations of diffuse radio emission such as radio halos and relics. Current formation theories cannot fully account for some of the observed properties of this emission. In this study, we focus on the development of interferometric techniques for extracting extended, faint diffuse emissions in the presence of bright, compact sources in wide-field and broadband continuum imaging data. We aim to apply these techniques to the study of radio halos, relics and radio mini-halos using a uniformly selected and complete sample of galaxy clusters selected via the Sunyaev-Zel’dovich (SZ) effect by the Atacama Cosmology Telescope (ACT) project, and its polarimetric extension (ACTPol). We use the upgraded Giant Metrewave Radio Telescope (uGMRT) for targeted radio observations of a sample of 40 clusters. We present an overview of our sample, confirm the detection of a radio halo in ACT−CL J0034.4+0225, and compare the narrowband and wideband analysis results for this cluster. Due to the complexity of the ACT−CL J0034.4+0225 field, we use three pipelines to process the wideband data. We conclude that the experimental spam wideband pipeline produces the best results for this particular field. However, due to the severe artefacts in the field, further analysis is required to improve the image quality.

Automated detection and reacquisition of motion-degraded images in fetal HASTE imaging at 3 T.

Fetal brain Magnetic Resonance Imaging suffers from unpredictable and unconstrained fetal motion that causes severe image artifacts even with half-Fourier single-shot fast spin echo (HASTE) readouts. This work presents the implementation of a closed-loop pipeline that automatically detects and reacquires HASTE images that were degraded by fetal motion without any human interaction. A convolutional neural network that performs automatic image quality assessment (IQA) was run on an external GPU-equipped computer that was connected to the internal network of the MRI scanner. The modified HASTE pulse sequence sent each image to the external computer, where the IQA convolutional neural network evaluated it, and then the IQA score was sent back to the sequence. At the end of the HASTE stack, the IQA scores from all the slices were sorted, and only slices with the lowest scores (corresponding to the slices with worst image quality) were reacquired. The closed-loop HASTE acquisition framework was tested on 10 pregnant mothers, for a total of 73 acquisitions of our modified HASTE sequence. The IQA convolutional neural network, which was successfully employed by our modified sequence in real time, achieved an accuracy of 85.2% and area under the receiver operator characteristic of 0.899. The proposed acquisition/reconstruction pipeline was shown to successfully identify and automatically reacquire only the motion degraded fetal brain HASTE slices in the prescribed stack. This minimizes the overall time spent on HASTE acquisitions by avoiding the need to repeat the entire stack if only few slices in the stack are motion-degraded.

Prior image-guided cone-beam computed tomography augmentation from under-sampled projections using a convolutional neural network.

Acquiring sparse-view cone-beam computed tomography (CBCT) is an effective way to reduce the imaging dose. However, images reconstructed by the conventional filtered back-projection method suffer from severe streak artifacts due to the projection under-sampling. Existing deep learning models have demonstrated feasibilities in restoring volumetric structures from the highly under-sampled images. However, because of the inter-patient variabilities, they failed to restore the patient-specific details with the common restoring pattern learned from the group data. Although the patient-specific models have been developed by training models using the intra-patient data and have shown effectiveness in restoring the patient-specific details, the models have to be retrained to be exclusive for each patient. It is highly desirable to develop a generalized model that can utilize the patient-specific information for the under-sampled image augmentation. In this study, we proposed a merging-encoder convolutional neural network (MeCNN) to realize the prior image-guided under-sampled CBCT augmentation. Instead of learning the patient-specific structures, the proposed model learns a generalized pattern of utilizing the patient-specific information in the prior images to facilitate the under-sampled image enhancement. Specifically, the MeCNN consists of a merging-encoder and a decoder. The merging-encoder extracts image features from both the prior CT images and the under-sampled CBCT images, and merges the features at multi-scale levels via deep convolutions. The merged features are then connected to the decoders via shortcuts to yield high-quality CBCT images. The proposed model was tested on both the simulated CBCTs and the clinical CBCTs. The predicted CBCT images were evaluated qualitatively and quantitatively in terms of image quality and tumor localization accuracy. Mann-Whitney U test was conducted for the statistical analysis. P<0.05 was considered statistically significant. The proposed model yields CT-like high-quality CBCT images from only 36 half-fan projections. Compared to other methods, CBCT images augmented by the proposed model have significantly lower intensity errors, significantly higher peak signal-to-noise ratio, and significantly higher structural similarity with respect to the ground truth images. Besides, the proposed method significantly reduced the 3D distance of the CBCT-based tumor localization errors. In addition, the CBCT augmentation is nearly real-time. With the prior-image guidance, the proposed method is effective in reconstructing high-quality CBCT images from the highly under-sampled projections, considerably reducing the imaging dose and improving the clinical utility of the CBCT.

EyelashNet

Eyelashes play a crucial part in the human facial structure and largely affect the facial attractiveness in modern cosmetic design. However, the appearance and structure of eyelashes can easily induce severe artifacts in high-fidelity multi-view 3D face reconstruction. Unfortunately it is highly challenging to remove eyelashes from portrait images using both traditional and learning-based matting methods due to the delicate nature of eyelashes and the lack of eyelash matting dataset. To this end, we present EyelashNet, the first eyelash matting dataset which contains 5,400 high-quality eyelash matting data captured from real world and 5,272 virtual eyelash matting data created by rendering avatars. Our work consists of a capture stage and an inference stage to automatically capture and annotate eyelashes instead of tedious manual efforts. The capture is based on a specifically-designed fluorescent labeling system. By coloring the eyelashes with a safe and invisible fluorescent substance, our system takes paired photos with colored and normal eyelashes by turning the equipped ultraviolet (UVA) flash on and off. We further correct the alignment between each pair of photos and use a novel alpha matte inference network to extract the eyelash alpha matte. As there is no prior eyelash dataset, we propose a progressive training strategy that progressively fuses captured eyelash data with virtual eyelash data to learn the latent semantics of real eyelashes. As a result, our method can accurately extract eyelash alpha mattes from fuzzy and self-shadow regions such as pupils, which is almost impossible by manual annotations. To validate the advantage of EyelashNet, we present a baseline method based on deep learning that achieves state-of-the-art eyelash matting performance with RGB portrait images as input. We also demonstrate that our work can largely benefit important real applications including high-fidelity personalized avatar and cosmetic design.

Abstract 14407: Effect of Initial Defibrillation on Chest Compression Artifact in the ECG

Introduction: Chest compressions (CCs) during CPR cause electrical artifacts in the ECG. Prior work has found that the severity of CC artifact, quantified by the signal-to-noise ratio (SNR), affects the diagnostic sensitivity of defibrillator algorithms designed to detect shockable rhythms during CCs. Whether SNR is altered by defibrillation is unknown. We therefore compared SNR before and after defibrillation shocks. Methods: We evaluated patients with out-of-hospital cardiac arrest who received at least 1 defibrillation shock, had subsequent ventricular fibrillation (VF), and had a calculable SNR before and after initial shock. We measured the CC artifact during VF before and after the initial shock (and up to 3 subsequent shocks) using CC amplitude and SNR. CC amplitude was defined as the median peak-to-peak voltage of the ECG during CCs. SNR was calculated as the log ratio of the power of the CC-free VF signal to the power of the estimated noise caused by CC artifact (Figure). Differences in medians before and after the first 4 shocks were evaluated using Wilcoxon signed-rank test with Bonferroni correction (alpha = 0.0125). Results: A total of 192 patients had a calculable SNR during VF before and after initial shock. Of these, the median CC amplitude decreased after the initial shock (0.93 vs. 0.75 mV, p&lt;0.001), and SNR improved (-2.30 vs. -1.07 dB, p=0.004). In contrast to the initial shock, both CC amplitude and SNR did not differ significantly before and after shock 2 (n=107), shock 3 (n=54), or shock 4 (n=32). Conclusion: Measures of CC artifact in the ECG were greater before initial shock than afterward. This could potentially be due to changes in CC characteristics, variations in physical perturbation of the defibrillator electrodes, degradation of VF over time, or effects of tissue electroporation on paddle conductivity and noise. These findings may have implications for selection of decision thresholds in algorithms to detect shockable rhythm during CCs.

Abstract 12792: Enhancing the Accuracy of Cardiac Rhythm Analysis in Automated External Defibrillators During Ongoing Cardiopulmonary Resuscitation by Applying a Deep Encoder-Decoder Filtering Model

Survival from out-of-hospital cardiac arrests depends on an accurate defibrillatory shock decision during cardiopulmonary resuscitation (CPR). Since chest compressions induce severe motion artifact in the electrocardiogram (ECG), current automatic external defibrillators (AEDs) do not perform CPR during the rhythm analysis period. However, performing continuous CPR is vital and dramatically increases the chance of survival. Hence, we demonstrate a novel application of a deep convolutional neural network encoder-decoder (CNNED) method in suppressing CPR artifact in near real-time using only ECG data. The encoder portion of the CNNED uses the frequency and phase contents derived via time-varying spectral analysis to learn distinct features that are representative of both the ECG signal and CPR artifact. The decoder portion takes the results from the encoder and reconstructs what is perceived as the motion artifact removed ECG data. These procedures are done via multitude of training of CNNED using many different arrhythmia contaminated with CPR. In this study, CPR-contaminated ECGs were generated by combining clean ECG with CPR artifacts from 52 different performers. ECG data from CUDB, VFDB, and SDDB datasets which belong to the Physionet’s Physiobank archive were used to create the training set containing 89,984 14-second ECG segments. The performance of the proposed CNNED was evaluated on a separate test set comprising of 23,816 CPR-contaminated 14-second ECG segments from 458 subjects. The results were evaluated by two metrics: signal-to-noise ratio (SNR), and Defibtech’s AED rhythm analysis algorithm. CNNED resulted in the increase of the mean SNR value from -3 dB to 5.63 dB and 6.3 dB for shockable and non-shockable rhythms, respectively. Comparing Defibtech’s AED rhythm classifier before and after applying CNNED on the CPR-contaminated ECG, the specificity improved from 96.57% to 99.31% for normal sinus rhythm, and from 91.5% to 96.57% for other non-shockable rhythms. The sensitivity of shockable detection also increased from 67.68% to 87.76% for ventricular fibrillation, and from 62.71% to 81.27% for ventricular tachycardia. These results indicate continuous and accurate AED rhythm analysis without stoppage of CPR using only ECG data.

Diagnosing and mitigating method-based avidity artifacts that confound polyubiquitin-binding assays

Polyubiquitination is a complex form of posttranslational modification responsible for the control of numerous cellular processes. Many ubiquitin-binding proteins recognize distinct polyubiquitin chain types, and these associations help drive ubiquitin-signaling pathways. There is considerable interest in understanding the specificity of ubiquitin-binding proteins; however, because of the multivalent nature of polyubiquitin, affinity measurements of these interactions that rely on affixing ubiquitin-binding proteins to a surface can display artifactual, method-dependent avidity, or “bridging.” This artifact, which is distinct from biologically relevant, avid interactions with polyubiquitin, is commonplace in such polyubiquitin-binding measurements and can lead to dramatic overestimations of binding affinities for particular chain types, and thus, incorrect conclusions about specificity. Here, we use surface-based measurements of ubiquitin binding in three model systems to illustrate bridging and lay out practical ways of identifying and mitigating it. Specifically, we describe a simple fitting model that enables researchers to diagnose the severity of bridging artifacts, determine whether they can be minimized, and more accurately evaluate polyubiquitin-binding specificity.

Quantitative assessment of the parotid gland using computed tomography texture analysis to detect parotid sialadenitis

We aimed to quantitatively assess the parotid gland by using computed tomography (CT) texture analysis to detect parotid sialadenitis (PS). This retrospective case-control study included 43 patients with PS who underwent CT and magnetic resonance imaging (MRI). Parotid glands with an abnormal signal (STIR: High) on MRI were identified as showing PS. Patients with parotid gland tumors, bilateral PS, marked fatty degeneration, and severe artifacts on CT were excluded. The texture features of parotid glands with PS and the contralateral normal parotid glands were analyzed using the open-access software LIFEx. The regions of interest were manually placed by tracing contours of both parotid glands on CT images. The results were tested with the paired t-test (or Wilcoxon rank-sum test when appropriate). Receiver operating characteristic (ROC) curve analysis was performed to assess the ability of texture features to predict PS. Six gray level run length matrix features, 2 neighborhood gray level difference matrix features, and 5 gray level zone length matrix features displayed significant differences between PS and normal glands (P ≤ .047). ROC curve analysis showed acceptable accuracy in 4 texture features. CT texture analysis allowed quantitative assessment of parotid glands and may have the potential to detect PS.