Search

The method of surgical treatment of pleomorph adenomas of parotid salivary gland with use skin-muscular autotransplantante Is developed. The given method is the most comprehensible as reduces risk of occurrence of relapses, Lucy Frej’s syndrome, and also promotes replacement of forming defect of fabrics, that provides good therapeutic and cosmetic effect.

Super resolution for satellite imagery possesses specific challenges due to its unique feature geometry compared to classic computer vision. Specifically, satellite images contain a high amount of relatively small, distributed high-frequency features that are hard to preserve when sampling up to a higher resolution. General adversarial networks (GANs) are suitable for super resolution tasks but need special attention concerning the geometric and radiometric accuracy of the results. We propose a network for versatile satellite imagery super resolution (VSISR) that focuses on high-frequency detail preservation and radiometric consistency. As a novelty, it is able to utilize a reference high resolution image during inference to lower hallucinations without altering the source image’s radiometric characteristics.A GAN that is already handling well high frequencies in standard computer vision cases is adapted for satellite imagery and supplemented with a mixed pixel approach for data augmentation. Training on a diverse RGB dataset from four satellite missions results in a versatile super resolution model that is optimized to preserve radiometric features and minimize hallucinations. Compared to other neural networks for satellite image super resolution in their respective datasets, VSISR performs in the middle regarding mean PSNR (25.30 dB) and SSIM (0.8098). Nevertheless, it is the first time that, using mixed pixel training on a comparably small dataset, this versatility concerning the satellite data source is achieved while maintaining their unique radiometric traits.

In the fields of photogrammetry, computer vision and computer graphics, the task of neural 3D scene reconstruction has led to the exploration of various techniques. Among these, 3D Gaussian Splatting stands out for its explicit representation of scenes using 3D Gaussians, making it appealing for tasks like 3D point cloud extraction and surface reconstruction. Motivated by its potential, we address the domain of 3D scene reconstruction, aiming to leverage the capabilities of the Microsoft HoloLens 2 for instant 3D Gaussian Splatting. We present HoloGS, a novel workflow utilizing HoloLens sensor data, which bypasses the need for pre-processing steps like Structure from Motion by instantly accessing the required input data i.e. the images, camera poses and the point cloud from depth sensing. We provide comprehensive investigations, including the training process and the rendering quality, assessed through the Peak Signal-to-Noise Ratio, and the geometric 3D accuracy of the densified point cloud from Gaussian centers, measured by Chamfer Distance. We evaluate our approach on two self-captured scenes: An outdoor scene of a cultural heritage statue and an indoor scene of a fine-structured plant. Our results show that the HoloLens data, including RGB images, corresponding camera poses, and depth sensing based point clouds to initialize the Gaussians, are suitable as input for 3D Gaussian Splatting.

3D reconstruction is a long-standing research topic in the photogrammetric and computer vision communities; although a plethora of open-source and commercial solutions for 3D reconstruction have been released in the last few years, several open challenges and limitations still exist. Undoubtedly, deep learning algorithms have demonstrated great potential in several remote sensing tasks, including image-based 3D reconstruction. State-of-the-art monocular and stereo algorithms leverage deep learning techniques and achieve increased performance in depth estimation and 3D reconstruction. However, one of the limitations of such methods is that they highly rely on large training sets that are often tedious to obtain; even when available, they typically refer to indoor, close-range scenarios and low-resolution images. Especially while considering UAV (Unmanned Aerial Vehicle) scenarios, such data are not available and domain adaptation is not a trivial challenge. To fill this gap, the UAV-based multi-sensor dataset for geospatial research (UseGeo - https://usegeo.fbk.eu/home) is introduced in this paper. It contains both image and LiDAR data and aims to support relevant research in photogrammetry and computer vision with a useful training set for both stereo and monocular 3D reconstruction algorithms. In this regard, the dataset provides ground truth data for both point clouds and depth maps. In addition, UseGeo can be also a valuable dataset for other tasks such as feature extraction and matching, aerial triangulation, or image and LiDAR co-registration. The paper introduces the UseGeo dataset and validates some state-of-the-art algorithms to assess their usability for both monocular and multi-view 3D reconstruction.

Depth estimation and 3D model reconstruction from aerial imagery is an important task in photogrammetry, remote sensing, and computer vision. To compare the performance of different image-based approaches, this study presents a benchmark for UAV-based aerial imagery using the UseGeo dataset. The contributions include the release of various evaluation routines on GitHub, as well as a comprehensive comparison of baseline approaches, such as methods for offline multi-view 3D reconstruction resulting in point clouds and triangle meshes, online multi-view depth estimation, as well as single-image depth estimation using self-supervised deep learning. With the release of our evaluation routines, we aim to provide a universal protocol for the evaluation of depth estimation and 3D reconstruction methods on the UseGeo dataset. The conducted experiments and analyses show that each method excels in a different category: the depth estimation from COLMAP outperforms that of the other approaches, ACMMP achieves the lowest error and highest completeness for point clouds, while OpenMVS produces triangle meshes with the lowest error. Among the online methods for depth estimation, the approach from the Plane-Sweep Library outperforms the FaSS-MVS approach, while the latter achieves the lowest processing time. And even though the particularly challenging nature of the dataset and the small amount of training data leads to a significantly higher error in the results of the self-supervised single-image depth estimation approach, it outperforms all other approaches in terms of processing time and frame rate. In our evaluation, we have also considered modern learning-based approaches that can be used for image-based 3D reconstruction, such as NeRFs. However, due to the significantly lower quality of the resulting 3D models, we have only included a qualitative comparison between NeRF-based and conventional approaches in the scope of this work.

Generating geometric 3D reconstructions from Neural Radiance Fields (NeRFs) is of great interest. However, accurate and complete reconstructions based on the density values are challenging. The network output depends on input data, NeRF network configuration and hyperparameter. As a result, the direct usage of density values, e.g. via filtering with global density thresholds, usually requires empirical investigations. Under the assumption that the density increases from non-object to object area, the utilization of density gradients from relative values is evident. As the density represents a position-dependent parameter it can be handled anisotropically, therefore processing of the voxelized 3D density field is justified. In this regard, we address geometric 3D reconstructions based on density gradients, whereas the gradients result from 3D edge detection filters of the first and second derivatives, namely Sobel, Canny and Laplacian of Gaussian. The gradients rely on relative neighboring density values in all directions, thus are independent from absolute magnitudes. Consequently, gradient filters are able to extract edges along a wide density range, almost independent from assumptions and empirical investigations. Our approach demonstrates the capability to achieve geometric 3D reconstructions with high geometric accuracy on object surfaces and remarkable object completeness. Notably, Canny filter effectively eliminates gaps, delivers a uniform point density, and strikes a favorable balance between correctness and completeness across the scenes.

Neural Radiance Fields (NeRFs) use a set of camera poses with associated images to represent a scene through a position-dependent density and radiance at given spatial location. Generating a geometric representation in form of a point cloud is gained by ray tracing and sampling 3D points with density and color along the rays. In this contribution we evaluate object reconstruction by NeRFs in 3D metric space against Terrestrial Laser Scanning (TLS) using ground truth data in form of a Structured Light Imaging (SLI) mesh and investigate the influence of the density to the reconstruction’s accuracy. We extend the accuracy assessment from 2D to 3D space and perform high resolution investigations on NeRFs by using camera images with 36MP resolution as well as comparison among point clouds of more than 20 million points against a 0.1mm ground truth mesh. TLS achieves the highest geometric accuracy results with a standard deviation of 1.68mm, while NeRFδt=300 diverges 18.61mm from the ground truth. All NeRF reconstructions contain 3D points inside the object which have the highest displacements from the ground truth, thus contribute the most to the accuracy results. NeRFs accuracy improves with increasing the density threshold as a consequence of completeness, since beside noise and outliers the object points are also being removed.

Abstract The nuclear shell model is one of the prime many-body methods to study the structure of atomic nuclei, but it is hampered by an exponential scaling on the basis size as the number of particles increases. We present a shell-model quantum circuit design strategy to find nuclear ground states by exploiting an adaptive variational quantum eigensolver algorithm. Our circuit implementation is in excellent agreement with classical shell-model simulations for a dozen of light and medium-mass nuclei, including neon and calcium isotopes. We quantify the circuit depth, width and number of gates to encode realistic shell-model wavefunctions. Our strategy also addresses explicitly energy measurements and the required number of circuits to perform them. Our simulated circuits approach the benchmark results exponentially with a polynomial scaling in quantum resources for each nucleus. This work paves the way for quantum computing shell-model studies across the nuclear chart and our quantum resource quantification may be used in configuration-interaction calculations of other fermionic systems.

An inverse measurement procedure for the determination of a full set of piezoelectric material parameters using a single sample is presented. The basis for the measurement procedure is a measurement of the frequency-dependent impedance of the sample. To yield sufficient sensitivity of this measurement with respect to all material parameters (mechanical, electrical, and coupling parameters), an optimal electrode configuration for the sample is determined before the inverse measurement procedure is realised using a novel topology optimisation approach. After initial estimates for the material parameters are provided by analytical expressions, a sensitivity-based, staged, local optimisation procedure yields material parameters for the sample by fitting the impedance of a finite element simulation model to the measured electrical impedance. Results for different absorption models as well as for different piezoelectric materials (hard, soft, and lead-free piezoceramics) are included.

Neural Radiance Fields (NeRFs) are trained using a set of camera poses and associated images as input to estimate density and color values for each position. The position-dependent density learning is of particular interest for photogrammetry, enabling 3D reconstruction by querying and filtering the NeRF coordinate system based on the object density. While traditional methods like Structure from Motion are commonly used for camera pose calculation in pre-processing for NeRFs, the HoloLens offers an interesting interface for extracting the required input data directly. We present a workflow for high-resolution 3D reconstructions almost directly from HoloLens data using NeRFs. Thereby, different investigations are considered: Internal camera poses from the HoloLens trajectory via a server application, and external camera poses from Structure from Motion, both with an enhanced variant applied through pose refinement. Results show that the internal camera poses lead to NeRF convergence with a PSNR of 25 dB with a simple rotation around the x-axis and enable a 3D reconstruction. Pose refinement enables comparable quality compared to external camera poses, resulting in improved training process with a PSNR of 27 dB and a better 3D reconstruction. Overall, NeRF reconstructions outperform the conventional photogrammetric dense reconstruction using Multi-View Stereo in terms of completeness and level of detail.

This work represents a large step into modern ways of fast 3D reconstruction based on RGB camera images. Utilizing a Microsoft HoloLens 2 as a multisensor platform that includes an RGB camera and an inertial measurement unit for SLAM-based camera-pose determination, we train a Neural Radiance Field (NeRF) as a neural scene representation in real-time with the acquired data from the HoloLens. The HoloLens is connected via Wifi to a high-performance PC that is responsible for the training and 3D reconstruction. After the data stream ends, the training is stopped and the 3D reconstruction is initiated, which extracts a point cloud of the scene. With our specialized inference algorithm, five million scene points can be extracted within 1 second. In addition, the point cloud also includes radiometry per point. Our method of 3D reconstruction outperforms grid point sampling with NeRFs by multiple orders of magnitude and can be regarded as a complete real-time 3D reconstruction method in a mobile mapping setup.

Abstract Background Machine-learning models are susceptible to external influences which can result in performance deterioration. The aim of our study was to elucidate the impact of a sudden shift in covariates, like the one caused by the Covid-19 pandemic, on model performance. Methods After ethical approval and registration in Clinical Trials (NCT04092933, initial release 17/09/2019), we developed different models for the prediction of perioperative mortality based on preoperative data: one for the pre-pandemic data period until March 2020, one including data before the pandemic and from the first wave until May 2020, and one that covers the complete period before and during the pandemic until October 2021. We applied XGBoost as well as a Deep Learning neural network (DL). Performance metrics of each model during the different pandemic phases were determined, and XGBoost models were analysed for changes in feature importance. Results XGBoost and DL provided similar performance on the pre-pandemic data with respect to area under receiver operating characteristic (AUROC, 0.951 vs. 0.942) and area under precision-recall curve (AUPR, 0.144 vs. 0.187). Validation in patient cohorts of the different pandemic waves showed high fluctuations in performance from both AUROC and AUPR for DL, whereas the XGBoost models seemed more stable. Change in variable frequencies with onset of the pandemic were visible in age, ASA score, and the higher proportion of emergency operations, among others. Age consistently showed the highest information gain. Models based on pre-pandemic data performed worse during the first pandemic wave (AUROC 0.914 for XGBoost and DL) whereas models augmented with data from the first wave lacked performance after the first wave (AUROC 0.907 for XGBoost and 0.747 for DL). The deterioration was also visible in AUPR, which worsened by over 50% in both XGBoost and DL in the first phase after re-training. Conclusions A sudden shift in data impacts model performance. Re-training the model with updated data may cause degradation in predictive accuracy if the changes are only transient. Too early re-training should therefore be avoided, and close model surveillance is necessary.

Abstract Inattention can negatively impact several aspects of a child’s life, including at home and school. Cognitive and physical interventions are two promising non-pharmaceutical approaches used to enhance attention abilities, with combined approaches often being marketed to teachers, therapists, and parents typically without research validation. Here, we assessed the feasibility of incorporating an integrated, cognitive-physical, closed-loop video game (body-brain trainer or ‘BBT’) as an after-school program, and also evaluated if there were attention benefits following its use. Twenty-two children (7–12 years of age) with a range of attention abilities were recruited to participate in this proof of concept, single-arm, longitudinal study (24 sessions over 8 weeks, ~30 min/day). We interrogated attention abilities through a parent survey of their child’s behaviors, in addition to objective performance-based and neural measures of attention. Here we observed 95% compliance as well as, significant improvements on the parent-based reports of inattention and on cognitive tests and neural measures of attention that were comparable in scale to previous work. Exploratory measures of other cognitive control abilities and physical fitness also showed similar improvement, with exploratory evaluation of retained benefits on the primary attention-related outcomes being present 1-year later. Lastly, there was no correlation between the baseline parent-rated inattention score and the improvement on the primary task-based measures of attention, suggesting that intervention-based benefits were not solely attained by those who stood the most to gain. These pilot findings warrant future research to replicate and extend these findings.

Climate warming and related drivers of soil thermal change in the Arctic are expected to modify the distribution and dynamics of carbon contained in perennially frozen grounds. Thawing of permafrost in the Mackenzie River watershed of northwestern Canada, coupled with increases in river discharge and coastal erosion, triggers the release of terrestrial organic matter (OMt) from the largest Arctic drainage basin in North America into the Arctic Ocean. While this process is ongoing and its rate is accelerating, the fate of the newly mobilized organic matter as it transits from the watershed through the delta and into the marine system remains poorly understood. In the framework of the European Horizon 2020 Nunataryuk programme, and as part of the Work Package 4 (WP4) Coastal Waters theme, four field expeditions were conducted in the Mackenzie Delta region and southern Beaufort Sea from April to September 2019. The temporal sampling design allowed the survey of ambient conditions in the coastal waters under full ice cover prior to the spring freshet, during ice breakup in summer, and anterior to the freeze-up period in fall. To capture the fluvial–marine transition zone, and with distinct challenges related to shallow waters and changing seasonal and meteorological conditions, the field sampling was conducted in close partnership with members of the communities of Aklavik, Inuvik and Tuktoyaktuk, using several platforms, namely helicopters, snowmobiles, and small boats. Water column profiles of physical and optical variables were measured in situ, while surface water, groundwater, and sediment samples were collected and preserved for the determination of the composition and sources of OMt, including particulate and dissolved organic carbon (POC and DOC), and colored dissolved organic matter (CDOM), as well as a suite of physical, chemical, and biological variables. Here we present an overview of the standardized datasets, including hydrographic profiles, remote sensing reflectance, temperature and salinity, particle absorption, nutrients, dissolved organic carbon, particulate organic carbon, particulate organic nitrogen, CDOM absorption, fluorescent dissolved organic matter intensity, suspended particulate matter, total particulate carbon, total particulate nitrogen, stable water isotopes, radon in water, bacterial abundance, and a string of phytoplankton pigments including total chlorophyll. Datasets and related metadata can be found in Juhls et al. (2021) (https://doi.org/10.1594/PANGAEA.937587).

Arctic river deltas define the interface between the terrestrial Arctic and the Arctic Ocean. They are the site of sediment, nutrient, and soil organic carbon discharge to the Arctic Ocean. Arctic deltas are unique globally because they are underlain by permafrost and acted on by river and sea ice, and many are surrounded by a broad shallow ramp. Such ramps may buffer the delta from waves, but as the climate warms and permafrost thaws, the evolution of Arctic deltas will likely take a different course, with implications for both the local scale and the wider Arctic Ocean. One important way to understand and predict the evolution of Arctic deltas is through numerical models. Here we present ArcDelRCM.jl, an improved reduced-complexity model (RCM) of arctic delta evolution based on the DeltaRCM-Arctic model (Lauzon et al., 2019), which we have reconstructed in Julia language using published information. Unlike previous models, ArcDelRCM.jl is able to replicate the ramp around the delta. We have found that the delayed breakup of the so-called “bottom-fast ice” (i.e. ice that is in direct contact with the bed of the channel or the sea, also known as “bed-fast ice”) on and around the deltas is ultimately responsible for the appearance of the ramp feature in our models. However, changes made to the modelling of permafrost erosion and the protective effects of bottom-fast ice are also important contributors. Graph analyses of the delta network performed on ensemble runs show that deltas produced by ArcDelRCM.jl have more interconnected channels and contain less abandoned subnetworks. This may suggest a more even feeding of sediments to all sections of the delta shoreline, supporting ramp growth. Moreover, we showed that the morphodynamic processes during the summer months remain active enough to contribute significant sediment input to the growth and evolution of Arctic deltas and thus should not be neglected in simulations gauging the multi-year evolution of delta features. Finally, we tested a strong climate-warming scenario on the simulated deltas of ArcDelRCM.jl, with temperature, discharge, and ice conditions consistent with RCP7–8.5. We found that the ramp features degrade on the timescale of centuries and effectively disappear in under 1 millennium. Ocean processes, which are not included in these models, may further shorten the timescale. With the degradation of the ramps, any dissipative effects on wave energy they offered would also decrease. This could expose the sub-aerial parts of the deltas to increased coastal erosion, thus impacting permafrost degradation, nutrients, and carbon releases.

Rapid Arctic warming accelerates permafrost thaw, causing an additional release of terrestrial organic matter (OM) into rivers and, ultimately, after transport via deltas and estuaries, to the Arctic Ocean nearshore. The majority of our understanding of nearshore OM dynamics and fate has been developed from freshwater rivers despite the likely impact of highly dynamic estuarine and deltaic environments on the transformation, storage, and age of OM delivered to coastal waters. Here, we studied particulate organic carbon (POC) dynamics in the Lena River delta and compared them with POC dynamics in the Lena River main stem along a ∼ 1600 km long transect from Yakutsk, downstream to the delta. We measured POC, total suspended matter (TSM), and carbon isotopes (δ13C and Δ14C) in POC to compare riverine and deltaic OM composition and changes in OM source and fate during transport offshore. We found that TSM and POC concentrations decreased by 70 % during transit from the main stem to the delta and Arctic Ocean. We found deltaic POC to be strongly depleted in 13C relative to fluvial POC. Dual-carbon (Δ14C and δ13C) isotope mixing model analyses indicated a significant phytoplankton contribution to deltaic POC (∼ 68 ± 6 %) and suggested an additional input of permafrost-derived OM into deltaic waters (∼ 18 ± 4 % of deltaic POC originates from Pleistocene deposits vs. ∼ 5 ± 4 % in the river main stem). Despite the lower concentration of POC in the delta than in the main stem (0.41 ± 0.10 vs. 0.79 ± 0.30 mg L−1, respectively), the amount of POC derived from Yedoma deposits in deltaic waters was almost twice as large as the amount of POC of Yedoma origin in the main stem (0.07 ± 0.02 and 0.04 ± 0.02 mg L−1, respectively). We assert that estuarine and deltaic processes require consideration in order to correctly understand OM dynamics throughout Arctic nearshore coastal zones and how these processes may evolve under future climate-driven change.