173102 research outputs found
Sort by
Streak artefact removal in x-ray dark-field computed tomography using a convolutional neural network.
No full textBACKGROUND: Computed tomography (CT) relies on the attenuation of x-rays, and is, hence, of limited use for weakly attenuating organs of the body, such as the lung. X-ray dark-field (DF) imaging is a recently developed technology that utilizes x-ray optical gratings to enable small-angle scattering as an alternative contrast mechanism. The DF signal provides structural information about the micromorphology of an object, complementary to the conventional attenuation signal. A first human-scale x-ray DF CT has been developed by our group. Despite specialized processing algorithms, reconstructed images remain affected by streaking artifacts, which often hinder image interpretation. In recent years, convolutional neural networks have gained popularity in the field of CT reconstruction, amongst others for streak artefact removal.
PURPOSE: Reducing streak artifacts is essential for the optimization of image quality in DF CT, and artefact free images are a prerequisite for potential future clinical application. The purpose of this paper is to demonstrate the feasibility of CNN post-processing for artefact reduction in x-ray DF CT and how multi-rotation scans can serve as a pathway for training data.
METHODS: We employed a supervised deep-learning approach using a three-dimensional dual-frame UNet in order to remove streak artifacts. Required training data were obtained from the experimental x-ray DF CT prototype at our institute. Two different operating modes were used to generate input and corresponding ground truth data sets. Clinically relevant scans at dose-compatible radiation levels were used as input data, and extended scans with substantially fewer artifacts were used as ground truth data. The latter is neither dose-, nor time-compatible and, therefore, unfeasible for clinical imaging of patients.
RESULTS: The trained CNN was able to greatly reduce streak artifacts in DF CT images. The network was tested against images with entirely different, previously unseen image characteristics. In all cases, CNN processing substantially increased the image quality, which was quantitatively confirmed by increased image quality metrics. Fine details are preserved during processing, despite the output images appearing smoother than the ground truth images.
CONCLUSIONS: Our results showcase the potential of a neural network to reduce streak artifacts in x-ray DF CT. The image quality is successfully enhanced in dose-compatible x-ray DF CT, which plays an essential role for the adoption of x-ray DF CT into modern clinical radiology
FBXL6 is a vulnerability in AML and unmasks proteolytic cleavage as a major experimental pitfall in myeloid cells
No full textToward a disruptive, minimally invasive small finger joint implant concept: Cellular and molecular interactions with materials in vivo
No full textOsteoarthritis (OA) poses significant therapeutic challenges, particularly OA that affects the hand. Currently available treatment strategies are often limited in terms of their efficacy in managing pain, regulating invasiveness, and restoring joint function. The APRICOTⓇ implant system developed by Aurora Medical Ltd (Chichester, UK) introduces a minimally invasive, bone-conserving approach for treating hand OA (https://apricot-project.eu/). By utilizing polycarbonate urethane (PCU), this implant incorporates a caterpillar track-inspired design to promote the restoration of natural movement to the joint. Surface modifications of PCU have been proposed for the biological fixation of the implant. This study investigated the biocompatibility of PCU alone or in combination with two surface modifications, namely dopamine-carboxymethylcellulose (dCMC) and calcium-phosphate (CaP) coatings. In a rat soft tissue model, native and CaP-coated PCU foils did not increase cellular migration or cytotoxicity at the implant–soft tissue interface after 3 d, showing gene expression of proinflammatory cytokines similar to that in non-implanted sham sites. However, dCMC induced an amplified initial inflammatory response that was characterized by increased chemotaxis and cytotoxicity, as well as pronounced gene activation of proinflammatory macrophages and neoangiogenesis. By 21 d, inflammation subsided in all the groups, allowing for implant encapsulation. In a rat bone model, 6 d and 28 d after release of the periosteum, all implant types were adapted to the bone surface with a surrounding fibrous capsule and no protracted inflammatory response was observed. These findings demonstrated the biocompatibility of native and CaP-coated PCU foils as components of APRICOTⓇ implants. Statement of significance: Hand osteoarthritis treatments require materials that minimize irritation of the delicate finger joints. Differing from existing treatments, the APRICOTⓇ implant leverages polycarbonate urethane (PCU) for minimally invasive joint replacement. This interdisciplinary, preclinical study investigated the biocompatibility of thin polycarbonate urethane (PCU) foils and their surface modifications with calcium-phosphate (CaP) or dopamine-carboxymethylcellulose (dCMC). Cellular and morphological analyses revealed that both native and Ca-P coated PCU elicit transient inflammation, similar to sham sites, and a thin fibrous encapsulation in soft tissues and on bone surfaces. However, dCMC surface modification amplified initial chemotaxis and cytotoxicity, with pronounced activation of proinflammatory and neoangiogenesis genes. Therefore, native and CaP-coated PCU possess sought-for biocompatible properties, crucial for patient safety and performance of APRICOTⓇ implant
Sugar or milk: Tribological study on the sensation of coffee beverages
No full textCoffee is a beverage with a long history of global popularity. In this study, the physicochemical characteristics, rheological and tribological properties, and oral tribological simulations of six different coffee formulations were investigated, and 12 characteristic attributes were quantitatively evaluated. The results showed that the coefficient of friction (COF) was directly related to the important sensory characteristics of coffee beverages and that changes in both sugar and milk content manifest as changes in the COF, which affect the stress and strain distribution at the tongue-palate interface. Our study demonstrated that tribology can, in conjunction with the characterization of other physicochemical properties, be used as a tool for the fast and easy prediction of important taste parameters of food
Methodenentwicklung zur Quantifizierung von Urban Mining Potentialen auf Quartiersebene
No full textThe anthropogenic reservoir is growing worldwide due to the growth of the built environment. In 2010,
the size of this stockpile in Germany amounted to around 51.7 billion [t] [1]. Of this, 90 % is made up
of the built environment [2]. Dynamic construction activity means that buildings, most of which consist of
mineral building materials, are demolished, renovated and/or newly built. In Germany, a large proportion
of the resulting construction waste is used for road construction, backfilling or landfill construction [3]. High
technical recycling rates for mineral building materials are already possible today [4]. Urban mining can be
used to make targeted recycling possible. This refers to the ”integral management of the anthropogenic
stockpile “[5]. This work deals with the question of how to determine the quantity and material composition
for a specific neighborhood on the basis of GIS data. This has been implemented by combining several
databases. Synthetic components can be generated from the information, which reflect the material com-
position and material quantity of the neighborhood. It is also possible to calculate the times at which the
building materials identified in the district will be converted back into construction waste based on certain
assumptions, for example through renovation. By knowing the composition and quantity, targeted urban
mining measures can then be taken. Construction waste can be recycled to produce construction materials
for building construction. This work also determines whether the material cycle within the neighborhood
can be closed by reusing the recovered building materials. During the investigation of a specific district,
potentials were identified in particular for the building material groups concrete and steel. By looking at the
amount of material within a neighborhood, it is also possible to make statements about how much ”grey
emissions“ have been emitted for the production of the building materials assumed in the neighborhood.
In addition to critically examining the potential of urban mining, the results generated also led to specific
recommendations for action for stakeholders along the construction value chain
Self-supervised Vessel Segmentation from X-ray Images using Digitally Reconstructed Radiographs
No full textImproving OpenMP Loop Scheduling in AutoPas
No full textParticle simulators commonly loop over particles and compute their interactions pairwise. Indeed, running the computations in parallel threads can significantly speed up the simulations. However, developers should look out for load imbalance: the uneven distribution of tasks between threads, such that some finish early and wait instead of taking up more work. For example, this may occur when tasks differ in complexity. To maintain optimal performance, simulators can thus benefit from load balancing techniques when scheduling their loop iterations to the threads. This thesis improves parallel loop scheduling in AutoPas: a particle simulation library. AutoPas employs OpenMP: a library that provides loop scheduling features via compiler directives. We integrate Auto4OMP, which is an OpenMP extension with advanced load balancing capabilities. To test the new setup, we ran different simulations with various parameters. In general, Auto4OMP performed close to the optimal standard OpenMP schedule, but the impact of its advanced scheduling techniques on non-homogeneous simulations was weaker than expected. That said, it showed promising results with homogeneous simulations
After effects of historical grassland on soil organic carbon content and plant growth in croplands in southern Germany determined using satellite data
No full textNumerous studies have reported that grasslands harbor higher soil organic carbon (SOC) stocks compared with arable land; however, the relevant carbon dynamics and sink persistence remain unclear. Herein, arable fields characterized by historical grassland zones (h_GL; grassland use decades ago) and permanent arable land zones (h_CL) were examined. The h_GL zones were determined using historical maps. The change in land use from grassland to cropland occurred 30–50 years ago. In eight arable fields, SOC and total nitrogen (TN) stocks in the topsoil were analyzed at a high spatial resolution. Additionally, remote sensing via satellites was employed to determine the biomass yield at a high spatial resolution using the normalized difference vegetation index (NDVI). In all the fields, the mean SOC content of the h_GL zones (1.81 %, n = 97 measuring points) was higher than the mean SOC content of the h_CL zones (1.52 %, n = 220). Furthermore, the mean relative NDVI was higher in the h_GL zones than in the h_CL zones. SOC and NDVI were positively correlated (up to r = 0.79), as well as TN and NDVI (up to r = 0.72). To evaluate the first dataset, zonal soil samples were collected from the h_GL and h_CL zones from 14 arable fields to determine the SOC and TN content. The mean SOC content of the h_GL zones was 1.92 % and that of the h_CL zones was 1.39 %—a difference of absolute SOC stocks in the topsoil of 23.8 t ha−1 (bulk density: 1.5 g cm−3). The work combines the knowledge of historical soil maps, remote sensing applications and georeferenced soil sampling and shows that SOC stocks in grassland have a high persistence and can have positive impact on yields even decades after a land use change. Historical land use proved to be a major factor for spatial SOC variability at the study site. © 2024 Technische Universität Münche
Increased asphericity of the femoral head-neck junction in professional breakers compared to hobby athletes - a retrospective case-control study.
No full textOBJECTIVE: Breaking has gained public attention as a form of sports activity. The associated intense movements of the hip joints are possibly linked to the development of femoroacetabular impingement (FAI). Therefore, this study aimed to assess clinical and radiographic FAI measures in professional breakers compared to hobby athletes.
METHODS: The study cohort consisted of professional breakers with persisting hip pain who were 1:1 matched to a cohort of FAI patients without professional sports careers from our outpatient clinic. The primary endpoint assessed on standardized plain radiographs was the alpha angle (AA). Further measures were the acetabular index (AI), lateral center-edge angle (LCEA), crossover sign, ischial spine sign, and femoral head extrusion index (FHEI). The modified Harris Hip Score (mHHS) and the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) score were used to obtain patient-reported measures.
RESULTS: We recruited ten professional breakers and matched them to ten hobby athletes. The median AA was significantly higher in the breakers compared with the hobby athletes (73° [IQR 66.5°, 84.2°]) vs. 61.8° [IQR 59.5°, 64.8°], p = 0.0004). There was a significant reduction in weekly training hours in breakers after diagnosis (13.0 hours [interquartile range [IQR] 9.5, 32.4] to 1.5 hours [IQR 0, 4.8], p = 0.0039). There were no inter-group differences regarding mHHS, WOMAC, and additional radiographic measurements.
CONCLUSION: Breakers have higher AA in cam-type FAI compared to nonprofessional athletes. The corresponding hip pain significantly reduced training hours and caused the end of their breaking career. The potentially high prevalence of FAI in breakers and the corresponding consequences need to be considered early when athletes present with hip pain