Fast and scalable annotation-free LV-centered ROI localisation in stress perfusion cardiac MRI.

Accurate placement of the endotracheal tube (ETT) is critical for ensuring optimal care for patients requiring mechanical ventilation and preventing potential complications. ETT positioning can be assessed using several methods, with chest X-ray (CXR) being the most precise. Radiologists evaluate whether the ETT requires adjustment by measuring the distance between the distal tip of the ETT and the tracheal carina. This study presents the development of a machine learning model to detect and measure ETT position on adult CXRs and evaluates its performance. Six physicians annotated ETT and trachea locations on a dataset of 3856 CXRs. The U-Net-based model was then trained to generate trachea and ETT segmentations. After post-processing steps, an estimate of the distance between the distal tip of the ETT and the tracheal carina was found. It was demonstrated that the trained model is capable of estimating the position of the ETT and calculating the distance from the tube tip to the tracheal carina. The Dice index for the segmentations on the external validation subset for the trachea and ETT was 89.2% ± 9.0% and 87.8% ± 16.9%, respectively. The estimated absolute error on the external validation subset was 4.72mm. This model represents a promising tool to support clinicians, particularly in Intensive Care Units, where correct intubation and effective ventilation are critical. It may also be integrated into clinical workflows to facilitate patient management and enhance patient safety.

Calculating excited states in chemistry is crucial to providing insight into photoinduced molecular behavior beyond the ground state, enabling innovations in spectroscopy, material sciences, and drug design. While several approaches have been developed to compute excited-state properties, finding the best ratio between the computational cost and accuracy remains challenging. The advent of quantum computers brings new perspectives with the development of quantum algorithms that promise an advantage over classical ones. Most of these new algorithms are inspired by previous classical ones but with different pros and cons. In this work, we focus on the choice of the weights within the ensemble variational quantum eigensolver (VQE) based on the generalization of the variational principle for many-body excited states. We compare the performance of the equi-ensemble and weighted ensemble VQE on two types of problems that cover different correlation regimes: (1) the many-body electronic structure problem of formaldimine using the generalized unitary coupled cluster ansatz and (2) the one-body noninteracting Kohn-Sham problem of hydrogen chains using the RYCNOT hardware-efficient ansatz. While the equi-ensemble objective is to perform a block-diagonalization only, the weighted ensemble has the additional complexity to reach the targeted eigenstates and eigenvalues. In this work, we show that such a difference leads to several significant consequences, such as the increase in circuit depth together with optimization issues. We conclude that one should always favor the use of equi-weights, even if it requires an additional postprocessing step to extract the eigenvalues of the problem.

Portfolio optimization is a cornerstone of financial decision-making, traditionally relying on classical algorithms to balance risk and return. Recent advances in quantum computing offer a promising alternative, leveraging quantum algorithms to efficiently explore complex solution spaces and potentially outperform classical methods in high-dimensional settings. However, conventional quantum approaches typically assume a one-to-one correspondence between qubits and variables (e.g. financial assets), which severely limits the applicability of gate-based quantum systems due to current hardware constraints. As a result, only quantum annealing-like methods have been used in realistic scenarios. In this work, we show how a gate-based variational quantum algorithm can be applied to a real-world portfolio optimization problem by assigning multiple variables per qubit, using the Pauli Correlation Encoding algorithm. Specifically, we address a problem involving over 250 variables, where the market graph representing a real stock market is iteratively partitioned into sub-portfolios of highly correlated assets. A classical postprocessing step is run to build a reduced, optimized assets portfolio. This approach enables improved scalability compared to traditional variational methods and opens new possibilities for quantum-enhanced financial applications.

Trends in Micro- and Nano-Lithography required for future development of large area applications ranging from high-packing-density electronics to solar cells are surveyed and outlined. Strategies to use direct laser writing to define etch masks over large areas by: (i) fixed beam moving stage and (ii) moving beam moving stage approaches are presented. The extension of planar 2D and stacked 2D (or 2.5D) fabrication methods into 3D micro- and nano-fabrication is discussed. One of the essential future characteristics of 3D nanolithography is real-time feedback capability. This can be realised via inherent 3D-capable holography, which bridges lithographic exposure control, wavefront sensing, and adaptive feedback, providing a pathway to stitch-free, large-area 3D patterning. The future of micro-fabrication is expected to evolve via highly specialised 3D architecture design and reduction in post-processing steps.

Quantum voting, inspired by quantum game theory, provides a framework in which the quantum majority rule (QMR) constitution of Bao and Yunger Halpern [Phys. Rev. A 95, 062306 (2017)] violates the quantum analogue of Arrow's impossibility theorem. We evaluate this QMR constitution analytically on classical profile data and implement its final measurement stage as a quantum circuit, running on both noiseless simulators and noisy IBM quantum hardware to map how realistic noise deforms the resulting societal ranking distribution. For the dephased profiles studied here, the post-processing steps act diagonally in the preference basis, so the ideal distributions underlying QMR's violation of the quantum analogue of Arrow's impossibility theorem can be computed classically before any optional quantum circuit sampling. Moderate readout and device noise generally preserve the qualitative behavior of QMR, whereas strong noise can shift the distribution toward different dominant winners or larger top-cycle structures, depending on the profile geometry. We quantify this behavior using winner-agreement rates, Condorcet-winner flip rates, and Jensen-Shannon divergence between societal ranking distributions. In addition to two representative hand-crafted profiles, we perform further robustness checks on randomized Dirichlet-sampled electorates and on cyclic and almost-cyclic near-threshold profiles, showing that the main QMR trends persist beyond the original examples while also revealing a sharp dependence on proximity to cycle-dominated majority structures. Unlike the two benchmark hand-crafted profiles, the randomized Dirichlet ensembles do not in general exhibit near-perfect robustness at very low noise, revealing substantial profile-to-profile sensitivity already near the noiseless limit. In a second, complementary component, we demonstrate an explicitly entanglement-based variant of the QMR constitution that serves as a testbed for multi-voter quantum correlations under noise, which we refer to as the QMR2-inspired variant. There, GHZ-type blocks and separable superpositions over opposite rankings have the same single-voter marginals, while their different correlation structure changes draw rates and variance in small mini-rounds; this effect is fragile under local noise and is washed out in the large-population limited-entanglement setting. Taken together, these two components connect the abstract QMR constitution to concrete implementations on noisy intermediate-scale quantum (NISQ) devices and highlight design considerations for future quantum and quantum-inspired voting protocols.

Three-dimensional (3D) printing is increasingly used for pediatric crown restorations; however, post-processing procedures may critically influence their mechanical performance. This in vitro study evaluated the effects of washing and post-curing durations on the fracture resistance and fracture pattern distribution of 3D-printed pediatric crowns after thermal cycling. Ninety crowns fabricated using a digital light processing (DLP) printer were assigned to nine groups according to washing (0, 3, and 10min of ultrasonic cleaning) and post-curing (0, 10, and 20min) durations. After cementation onto standardized dies and 10,000 thermal cycles, fracture resistance was tested and fracture patterns were classified. Two-way analysis of variance revealed significant effects of washing and post-curing (p < 0.001), with post-curing demonstrating the strongest individual influence and a significant interaction between factors. The highest fracture resistance was observed with 10min of ultrasonic washing followed by 10min of post-curing, whereas prolonged post-curing reduced strength. Fracture patterns also differed significantly among groups. These findings indicate that mechanical performance depends on the coordinated interaction of post-processing steps. For the tested resin system, 10min of ultrasonic washing followed by 10min of post-curing appears to represent a clinically applicable protocol to maximize fracture resistance.

Weed detection in complex agricultural environments faces significant challenges due to drastic illumination variations, high visual similarity between crops and weeds, and strong background interference, which place strict requirements on both detection accuracy and real-time performance. Most existing object detection methods rely on non-maximum suppression (NMS) as a post-processing step. However, this mechanism suffers from threshold sensitivity, limited cross-scene adaptability, and accumulated inference latency in practical applications. To address these limitations, this paper proposes a lightweight adaptive transformer-based model, termed CEAM-DETR, under an end-to-end detection paradigm to achieve efficient and stable weed detection without relying on NMS. A cross-stage efficient attention backbone is first constructed by integrating cross-stage partial connections with single-head self-attention, where feature splitting and cross-stage shortcuts preserve lightweight information and gradient propagation paths, while attention is applied only to partial channels to reduce computational and memory overhead and enhance fine-grained representations of small-scale targets. On this basis, an adaptive sparse feature interaction module (ASFI) is introduced to dynamically fuse sparse and dense attention branches, thereby improving the concentration of discriminative information under complex backgrounds. Furthermore, a multi-scale dilated re-parameterization block (MSDRB) is designed to extract features using parallel convolutions with different dilation rates and to equivalently merge them into a single convolution layer during inference, which expands the receptive field without increasing computational burden and supplements multi-scale contextual information. Experimental results on public datasets demonstrate that CEAM-DETR outperforms state-of-the-art methods in terms of detection accuracy and robustness, validating its effectiveness in complex agricultural environments. Compared to the original RT-DETR, the proposed model increased by 1.5%, reduced Params by 36.5%, decreased GFLOPs by 26.0%, and improved FPS by 32.9%.

The Gini index is a widely reported measure of income inequality. Because the underlying microdata are often confidential, releasing the Gini index may leak information. We present an approach for bounding this information leakage by releasing a differentially private version of the Gini index. In doing so, we analyze how adding, deleting, or altering a single observation in any specific dataset can affect the computation of the Gini index; this is known as the local sensitivity. We then derive a smooth upper bound on the local sensitivity. Using this bound, we define a mechanism that adds noise to the Gini index, thereby satisfying differential privacy. Using simulated and genuine income data, we show that the mechanism can reduce the errors from noise injection substantially relative to differentially private algorithms that rely on the global sensitivity, that is, the maximum of the local sensitivities over all possible datasets. We characterize settings where using smooth sensitivity can provide highly accurate estimates, as well as settings where the noise variance is simply too large to provide reliably useful results. We also present a post-processing step that provides interval estimates about the value of the Gini index computed with the confidential data.

Graphene, a two-dimensional carbon allotrope, exhibits outstanding mechanical, electrical, thermal, and chemical properties enabling widespread applications in flexible electronics, energy storage and harvesting, catalysis, and biosensing. Among the various graphene synthesis strategies, laser-induced graphene (LIG) offers distinct advantages in terms of scalability, simplicity, and direct patternability. However, the functional diversification of LIG remains a significant challenge, primarily due to reliance on multistep post-processing, high-temperature treatments, limited chemical tenability, poor area selectivity, and difficulties in integrating electrical connections and multifunctional graphene architectures within device platforms. Herein, we report, for the first time, an in-situ, under-solvent laser-directed approach that enables the direct, single-step formation and integration of chemically functionalized LIG structures without any post-synthesis. In this strategy, a solid polyimide (PI) film is laser-ablated under a predesigned solvent environment, where the solvent governs the incorporation of specific chemical functionalities into the emerging graphene network, allowing precise control over the chemical composition and morphology of the resulting LIG in a single laser-based step. By systematically tailoring both the solvent chemistry and laser parameters, the structural, physical, electrochemical, and catalytic properties of LIG can be finely engineered. Importantly, this work establishes a universal “toolbox” and efficient blueprint for producing a tunable framework for the area-selective, single-step integration of chemically functionalized LIG into microfluidic and electrochemical platforms. The spatially controlled functionalization enables the direct fabrication of multifunctional electrochemical circuits, in which distinct regions of LIG are selectively modified to perform complementary functions. As proof of concept, we have successfully fabricated a device using region-specific functionalized LIG for an electrochemical sensing platform for hydrogen peroxide (H 2 O 2 ). Cyclic voltammetry was performed at a scan rate of 50 mV/s, with a LOD of 0.2 mM. Furthermore, Platinum (Pt) nanoparticles decorated with LIG electrodes were fabricated through this approach, exhibiting enhanced electrocatalytic activity for water splitting. The electrodes were analyzed for hydrogen evolution (HER) activity in 1 M KOH electrolyte, where the LIG@Pt electrode showed an overpotential of ⁓68 mV at 10 mA cm −2 and a stability overpotential of ⁓68 mV for over 10 h. This solvent-assisted laser strategy provides a scalable and versatile pathway for the direct fabrication of advanced, multifunctional graphene-based devices. • Single-step, under-solvent laser writing enables in-situ chemically-functionalized LIG formation. • Solvent-controlled laser ablation enables direct synthesis of metal nanoparticle-embedded LIG active composites. • Laser-written LIG forms direct, electrically-connected circuits without post-processing steps. • Area-selective functionalization enables integrated multifunctional circuits for next-generation applications. • Pt–LIG composite exhibits low HER overpotential, and stable H₂O₂ sensing with a low LOD of 0.26 mM.

The special quasirandom structure (SQS) method is widely used for modeling disordered materials under periodic boundary conditions, with the ATAT mcsqs module being one of the most established implementations. However, SQS generation with mcsqs typically relies on manual preparation of input files, ad hoc execution scripts, and post-processing steps, which introduces user-dependent errors and limits reproducibility. Here, we present SimplySQS ( https://simplysqs.com ), an automated and reproducible workflow for SQS generation that is delivered through an online, interactive interface. SimplySQS guides users through structure import, compositional and supercell definition, and cluster parameter selection, while automatically generating all required ATAT input files and a single all-in-one execution script that encapsulates the complete search process. By standardizing input preparation, execution, and output analysis, the framework minimizes errors associated with manual file handling and enables consistent reproducibility of SQS searches. The workflow is demonstrated on the Pb 1- x Sr x TiO 3 (PSTO, including PbTiO 3 (PTO) and SrTiO 3 (STO)) perovskite system. SQSs spanning the entire concentration range were generated using a single automated bash script produced by SimplySQS , after which all resulting structures were subjected to geometry optimization using a universal machine-learning interatomic potential (MACE MATPES-r²SCAN-0). This approach reliably reproduced the experimentally observed cubic-to-tetragonal transition near x ≈ 0.5, with lattice parameters deviating by less than 1% in the cubic region ( x > 0.5) and less than 4% in the tetragonal region ( x ≤ 0.5). Overall, SimplySQS transforms SQS generation with ATAT into an intuitive, reproducible, and systematic framework for modeling disordered materials.

Direction-aware deformable registration for respiratory motion correction in PET/CT imaging.

Small, non-interconnected island systems are at the forefront of the energy transition but their isolated condition makes them exposed to the natural variability of renewable resources. This work develops a static robust optimization framework to design multi-technology portfolios that explicitly account for the inter-annual variability of the wave climate. The methodology couples the EnergyPLAN simulation tool with a MATLAB-based NSGA-II algorithm. The design vector includes the installed capacities of onshore and offshore wind, photovoltaics, two different wave energy converters (Pelamis and CorPower) and a battery energy storage system (BESS). Robustness is assessed over a discrete uncertainty set composed of twenty years of hourly meteomarine data from Copernicus reanalysis for La Gomera (Canary Islands). Three system-level indicators are optimized in a worst-case sense: annual C O 2 emissions, a demand-generation mismatch metric ϕ , and the BESS exploitation index. A post-processing step selects all robust portfolios that satisfy a stringent emissions target, corresponding to roughly a 70% reduction with respect to the validated reference configuration. Within this low-carbon subset, C O 2 is treated as a saturated objective and the remaining trade-offs are explored in the two-dimensional ϕ - R B E S S plane, where a secondary Pareto front is identified. The resulting portfolios reveal a clear interaction between storage use and temporal balancing, with different BESS levels but a quite constant wave contribution. The framework is generic and transferable to other island systems and renewable technology combinations, providing a practical tool for integrating long-term resource variability and explicit decarbonization targets into energy system planning. • Static robust optimization of multi-RES portfolios. • Robustness evaluated over twenty years of hourly Copernicus wave climate data. • La Gomera island used as real-world testbed for energy system planning strategies. • Robust low-carbon portfolios consistently retain wave energy in the RES mix. • Results demonstrate the strategic value of robust wave energy inclusive portfolios.

The fabrication of complex architectures remains a central challenge in 3D bioprinting, as the low mechanical properties of hydrogels limit the range of achievable geometries. Four-dimensional (4D) bioprinting can address these limitations by enabling programmed shape-morphing behavior; however, in most approaches, this functionality is introduced after hydrogel formation, limiting the complexity of the resulting deformation. Here, a proof-of-concept strategy is presented, in which shape-morphing is directly encoded during fabrication. By modulating light exposure time layer-by-layer in vat photopolymerization, spatial variations in crosslinking density are introduced in situ within Gelatin Methacryloyl (GelMA) hydrogel constructs. Exposure times in the range of 20–70 s were investigated, enabling controlled bending of the printed structures upon immersion in aqueous media, with radii of curvature between 11 and 20 mm depending on the geometry. This approach allows deformation pathways to be programmed during printing, without requiring additional materials or post-processing steps. The morphing behavior was further supported by finite element simulations, which reproduced the experimentally observed deformation and enabled prediction of the shape change. In addition, high cell viability (>95%) was maintained after material contact and UV exposure. Overall, this study demonstrates that swelling-driven actuation can be encoded during fabrication. Although demonstrated on simplified geometries, this approach provides a versatile framework for process-driven shape-morphing and represents a step toward more spatially resolved and potentially volumetric 4D bioprinting strategies.

Background/Objectives: The limited aqueous solubility of therapeutically active drugs remains a significant challenge in their pharmaceutical application. This study presents a novel solid dispersion matrix (NSDM) that utilizes the inverted thermoresponsive behavior of Pluronic F127 to enhance drug dissolution while addressing the industrial and ecological limitations of conventional methods. Methods: For comparative assessment, a solid dispersion formulation of dapagliflozin was formulated using the NSDM approach and three conventional approaches: heat fusion (HFSD), microwave (MWSD), and lyophilization (LPSD). Differential scanning calorimetry (DSC), Fourier transform infrared spectroscopy (FTIR), and X-ray diffraction (XRD) were used to characterize the prepared formulations. In vitro dissolution test was performed to compare the pharmaceutical performance of NSDM against conventional approaches. Results: The NSDM exhibited a unique thermal transition to the liquid state at 32.4 °C. Moreover, the physiological assessment revealed complete liquefaction within 81.7 s. DSC and XRD confirmed amorphization of dapagliflozin in all formulations. In addition, FTIR revealed that dapagliflozin was integrated within the formulation without any chemical interaction with the excipient. Dissolution studies showed remarkable superiority of NSDM, with 97.30 ± 2.26% dissolution efficiency and a mean dissolution time of 2.40 ± 0.80 min. A multi-criteria assessment of ecological impact, worker friendliness, industrial effectiveness, and pharmaceutical performance demonstrated NSDM’s comprehensive advantages. Conclusions: The present approach provides a sustainable paradigm compared to conventional solid dispersion approaches. It eliminates energy-intensive operations and post-processing steps through direct capsule filling. This affords superior pharmaceutical performance while supporting sustainability and industrial applicability.

Objective.Low-field magnetic resonance imaging (MRI) offers distinct advantages in terms of affordability, portability, and accessibility. However, its widespread adoption is limited by an inherently low signal-to-noise ratio (SNR) and reduced spatial resolution. This study proposes an AI-assisted framework to enhance low-field MRI image quality and overcome these limitations.Approach.We propose a two-stage framework to generate high-quality low-field MRI images. In the first stage, realistic low-field images are synthesized from high-field acquisitions using a physics-informed forward model that incorporates spiralk-space trajectories and accounts for nonlinear magnetic field gradients,B0inhomogeneity,k-space undersampling, and image reconstruction characteristics of low-field systems. In the second stage, a 3D U-Net enhanced with a multi-head attention in a vision transformer (ViT) module is trained on paired synthetic low- and high-field images to serve as a post processing following conventional image reconstruction.Main results.On the synthetic test set, our framework demonstrates strong performance, achieving a peak SNR (PSNR) of 19.08 ± 2.85 dB for the baseline U-Net model and 21.00 ± 2.50 dB with the ViT block, demonstrating high reconstruction fidelity. The structural similarity index measure reaches 0.6456 ± 0.0779 (without ViT) and 0.6639 ± 0.0798 (with ViT), along with low normalized root mean squared error values of 0.3866 ± 0.0952 and 0.3084 ± 0.0695, respectively. These results highlight significant improvements in both image quality and reconstruction robustness. The trained network, applied as a post-processing step after conventional reconstruction, consistently enhances the contrast-to-noise ratio of the output images, supporting the qualitative observations of improved image contrast and clarity.Significance.The proposed framework addresses key limitations hindering the broader adoption of low-field MRI, including noise, artifacts, and resolution loss inherent to low-field acquisitions. By integrating deep learning with physics-based simulations, the approach achieves notable qualitative and quantitative enhancements in denoising, artifact removal, and overall image quality. These results highlight the framework's potential to improve the practical utility of low-field MRI substantially.

This paper introduces a novel framework for improving social recommender systems by incorporating a personalized notion of item novelty grounded in user's social interactions. Unlike conventional approaches that treat novelty as a static, item-specific feature, our method estimates the novelty of each item relative to a given user by analyzing behavioral patterns within the user's social communities. Additionally, we model each user's individual tendency toward novel content, allowing for personalized calibration of the novelty-relevance trade-off in recommendations. The proposed method operates independently of the underlying recommendation algorithm and can be seamlessly integrated as a post-processing step over candidate lists generated by various base models. Experimental evaluations on two real-world datasets-Epinions, and LastFM-demonstrate that our framework consistently enhances diversity, coverage, and novelty while preserving recommendation relevance. These findings underscore the value of socially contextualized and user-personalized novelty modeling in elevating the effectiveness and user satisfaction of recommender systems.

Combinations of additive manufacturing methods such as inkjet printing and various 3D printing techniques have led to advances in customizable electronic devices by reducing the need for the complex assembly of individually fabricated components. Such combinations are particularly attractive for the fabrication of electrochemical sensing platforms, where the sizes and spatial configurations of electrodes enable (or limit) the sensitivity of the sensors. However, to realize the potential of the combined fabrication methods, the parameters of each printing technique must be mutually compatible. For instance, electrochemical sensing assemblies consisting of inkjet-printed metals on 3D-printed plastics require post-processing steps that convert nanoparticular metals in precursor ink to conformal, mechanically stable, conductive films with negligible alteration to the underlying plastic substrate. While traditional sintering techniques convert inks to conductive metal films, the use of high temperatures are not compatible with structures printed via fused deposition modeling, for instance. In this proof-of-concept study, commercial gold inks were inkjet-printed onto 3D-printed poly-(lactic acid) (PLA) substrates. Optimization of the printing parameters and IR sintering resulted in stable, electrochemically active working electrodes. Surface characterization via scanning electron microscopy confirmed the formation of a homogeneous coating, and X-ray photoelectron spectroscopy data revealed the presence of gold on the surface. Electrical characterization via a four-point probe demonstrated a low sheet resistance, indicating the suitability of these electrodes for electrochemical measurements. Electrochemical data revealed that these gold electrodes are electrochemically active, and the diffusion-limited behavior of a redox probe was observed as expected. Additionally, these electrodes were tested for detecting lead in standard solutions, showing a linear response to lead concentrations, which indicates their potential in sensing applications. However, the detection range is significantly higher than the EPA-approved limit for lead concentrations in water, suggesting that electrode sensitivity requires further improvement. These findings have implications for fabricating inkjet-printed gold electrodes as sensors while highlighting the need for additional modifications to meet regulatory detection limits for heavy metal analysis.

Objective: The material composition, manufacturing system, and post-processing steps used to fabricate acrylic-based occlusal devices may affect their clinical performance. This study aims to assess how manufacturing techniques and post-processing treatments influence the material properties of acrylic-based occlusal devices and the propensity of Streptococcus mutans to form biofilms on the surfaces. Materials and methods: Based on applied manufacturing technique and post-processing treatment, disc‑shaped specimens were manufactured using four 3D printing workflows (Splint 2.0‑Asiga Flash/Otoflash (OF), and LT Clear‑Form Cure/OF), one milling workflow (Therapon), and one autopolymerization workflow (PalaXtreme). Water sorption and solubility, surface free energy (SFE), average ­surface roughness, and Vickers hardness were tested across these workflows. The ATCC 700610 Streptococcus mutans strain served as a model for biofilm formation on the material surfaces. Two biofilm methods were employed: a 24-hour bioreactor approach and a 72-hour culture plate approach. Biofilm was quantified as colony-forming units per cm2. Results and conclusion: The Therapon and PalaXtreme workflows exhibited the lowest solubility, ­suggesting that these materials have the lowest release of material components in water. The Splint 2.0 workflows exhibited the lowest water sorption, indicating enhanced material integrity in humid conditions. Therapon showed the highest Vickers hardness, followed by PalaXtreme. The lower hardness of the print materials may make them susceptible to wear, which may not be optimal for treating patients with bruxism. No significant differences in SFE were observed between workflows. Low roughness values across all workflows indicate good polishability, which can enhance resistance to bacterial adhesion. In the 72-hour biofilm experiment, the Therapon workflow exhibited the most biofilm formation on material surfaces while PalaXtreme showed the least (p &lt; 0.05). No significant differences between workflow groups were shown in the 24-hour biofilm experiment.In summary, material properties are influenced by material chemistry, manufacturing method, and ­associated post-processing treatment.

Considering optical Braille patterns has been investigated in several studies. There is a huge number of studies which analyzed Braille patterns in different natural languages. However, the Arabic patterns have not been examined as same as the other languages. This is due to the lack of the datasets of the Arabic patterns and the shortage of researches in this area. This study utilizes YOLOv11 model as a detection tool because of its relative effectiveness and the level of accuracy as well as the staged training approach with the AdamW optimizer and Automatic Mixed Precision. For the translation of the Arabic patterns into text, post-processing steps are performed including: detecting cells vertically clustered, horizontally sorted within a line, adaptively defined word boundaries, and corrected reading order of right-to-left. In the analysis of experiments, the best findings achieved is 0.99 of all the precision, recall, and F1 scores. Moreover, the framework-level runtime indicates that the total processing time (inference + post-processing for text extraction) ranges between 29 and 82 ms per image. The proposed framework has been examined with a primary dataset of 5924 pages of images of Braille patterns of 45 classes of Arabic letters and diacritics. The yielded results show that the proposed framework is a robust approach toward effective, scalable, responsive Arabic Braille recognition (OBR) for assistive technologies to be mobile and wearable. By building the first dedicated corpus in Arabic Braille and providing an end-to-end recognition suite, this study laid the groundwork for future research and applications in the field. This research bridges the accessibility gap, so of allow sighted individuals to access content encoded in Braille.