Search

Integrating eye gaze data with chest X-ray images in deep learning (DL) has led to contradictory conclusions in the literature. Some authors assert that eye gaze data can enhance prediction accuracy, while others consider eye tracking irrelevant for predictive tasks. We argue that this disagreement lies in how researchers process eye-tracking data as most remain agnostic to the human component and apply the data directly to DL models without proper preprocessing. We present EyeXNet, a multimodal DL architecture that combines images and radiologists’ fixation masks to predict abnormality locations in chest X-rays. We focus on fixation maps during reporting moments as radiologists are more likely to focus on regions with abnormalities and provide more targeted regions to the predictive models. Our analysis compares radiologist fixations in both silent and reporting moments, revealing that more targeted and focused fixations occur during reporting. Our results show that integrating the fixation masks in a multimodal DL architecture outperformed the baseline model in five out of eight experiments regarding average Recall and six out of eight regarding average Precision. Incorporating fixation masks representing radiologists’ classification patterns in a multimodal DL architecture benefits lesion detection in chest X-ray (CXR) images, particularly when there is a strong correlation between fixation masks and generated proposal regions. This highlights the potential of leveraging fixation masks to enhance multimodal DL architectures for CXR image analysis. This work represents a first step towards human-centered DL, moving away from traditional data-driven and human-agnostic approaches.

This work aims to explore the efficiency of ZnO nanoparticles synthesized via the non-thermal gliding arc discharge-assisted plasma (NT-GAD) technique for inhibiting the corrosion of X60 API 5L steel in a 1M HCl environment. The XRD pattern revealed that the ZnO nanoparticles exhibit hexagonal wurtzite structure with average particle size of ∼24 nm. UV–visible spectroscopy analysis revealed an absorption peak centering at 365 nm, corresponding to an energy band gap of 3.29 eV. SEM and TEM analysis revealed that the nanoparticles exhibit an agglomerated and irregular morphology. The corrosion inhibition of ZnO NPs was investigated via the electrochemical impedance spectroscopy (EIS) and potentiodynamic polarization tests (PDP), while varying both concentration and temperature. The results revealed that the increase in inhibitor concentration resulted in a higher activity at ambient temperature, with an optimal efficiency of 93 % at a concentration of 100 mg/L. However, the increase in temperature remarkably reduced the inhibition efficiency, suggesting a physisorption behavior of ZnO NPs onto the steel surface. AFM and FE-SEM analysis confirmed the formation of a protective layer on the X60 API 5L steel surface. This study emphasizes the significant potential of ZnO NPs synthesized via the NT-GAD assisted plasma technique as corrosion inhibitor for X60 API 5L carbon steel in 1M HCl corrosive media.

Sol–gel method was used to prepare Sr1-xCdxZn2Fe4O11 (x = 0.00, 0.02, 0.06, 0.1) R-type hexaferrite. The synthesized materials were sintered at 850oC and desired phase was obtained. X-ray diffraction analysis confirms that R-type hexaferrite exists only as a single phase. Using the Scherer formula, crystallite size for all of the prepared samples was found to be in the range of 10.39–12.62 nm. Crystallite size (D), the lattice parameters (a, c), and the cell volume (Vcell), d-spacing, bulk density, X-ray density, porosity, dislocation density and micro strain were determined in structural analysis. Fourier transform infrared spectroscopy method was used to identify the metal-oxygen vibrations at different locations. FT-IR verifies the presence of the Fe–O stretching vibration band at 743 and 867 cm−1. The typical grain size in surface morphology investigation ranges from 0.56 to 0.82 μm. Dielectric response of ferrite ceramic samples replaced with Cd2+ was investigated in the frequency range of 1 MHz–3 GHz. The AC conductivity rises with an increase in frequency because they are proportional to one another. This increasing tendency is effectively described by the theory of Maxwell-Wagner and Koop. Q-values remain constant as frequency rises and behaves independently of frequency as long as frequency reaches 1.7 GHz. These types of materials are utilized in high-frequency applications including frequency filters and resonant circuits. All of the magnetic properties determined by analyzing the M − H loops, including saturation magnetization (Ms), retentivity (Mr), and coercivity (Hc), exhibit an increasing trend as the substitution of Cd2+ rises. Ms (49.76–56.38 emu/g), Mr (15.82–18.30 emu/g) and Hc range from 203.20 Oe to 215.80 Oe. Grain size decreases cause arise in coercivity, which is caused by an enhancement in magneto-crystalline anisotropy. Overall results suggest that Cd2+ replaced R-type hexagonal ferrites are a great resource for longitudinal recording media; they have the potential to be used in a wide variety of electronic applications, including resonant circuits and high-frequency filters, security, detecting and switching.

Abstract This study investigates the effects of including patients’ clinical information on the performance of deep learning (DL) classifiers for disease location in chest X-ray images. Although current classifiers achieve high performance using chest X-ray images alone, consultations with practicing radiologists indicate that clinical data is highly informative and essential for interpreting medical images and making proper diagnoses. In this work, we propose a novel architecture consisting of two fusion methods that enable the model to simultaneously process patients’ clinical data (structured data) and chest X-rays (image data). Since these data modalities are in different dimensional spaces, we propose a spatial arrangement strategy, spatialization, to facilitate the multimodal learning process in a Mask R-CNN model. We performed an extensive experimental evaluation using MIMIC-Eye, a dataset comprising different modalities: MIMIC-CXR (chest X-ray images), MIMIC IV-ED (patients’ clinical data), and REFLACX (annotations of disease locations in chest X-rays). Results show that incorporating patients’ clinical data in a DL model together with the proposed fusion methods improves the disease localization in chest X-rays by 12% in terms of Average Precision compared to a standard Mask R-CNN using chest X-rays alone. Further ablation studies also emphasize the importance of multimodal DL architectures and the incorporation of patients’ clinical data in disease localization. In the interest of fostering scientific reproducibility, the architecture proposed within this investigation has been made publicly accessible( https://github.com/ChihchengHsieh/multimodal-abnormalities-detection ).

Musculoskeletal conditions affect an estimated 1.7 billion people worldwide, causing intense pain and disability. These conditions lead to 30 million emergency room visits yearly, and the numbers are only increasing. However, diagnosing musculoskeletal issues can be challenging, especially in emergencies where quick decisions are necessary. Deep learning (DL) has shown promise in various medical applications. However, previous methods had poor performance and a lack of transparency in detecting shoulder abnormalities on X-ray images due to a lack of training data and better representation of features. This often resulted in overfitting, poor generalisation, and potential bias in decision-making. To address these issues, a new trustworthy DL framework has been proposed to detect shoulder abnormalities (such as fractures, deformities, and arthritis) using X-ray images. The framework consists of two parts: same-domain transfer learning (TL) to mitigate imageNet mismatch and feature fusion to reduce error rates and improve trust in the final result. Same-domain TL involves training pre-trained models on a large number of labelled X-ray images from various body parts and fine-tuning them on the target dataset of shoulder X-ray images. Feature fusion combines the extracted features with seven DL models to train several ML classifiers. The proposed framework achieved an excellent accuracy rate of 99.2%, F1Score of 99.2%, and Cohen's kappa of 98.5%. Furthermore, the accuracy of the results was validated using three visualisation tools, including gradient-based class activation heat map (Grad CAM), activation visualisation, and locally interpretable model-independent explanations (LIME). The proposed framework outperformed previous DL methods and three orthopaedic surgeons invited to classify the test set, who obtained an average accuracy of 79.1%. The proposed framework has proven effective and robust, improving generalisation and increasing trust in the final results.

Hybrid nanogenerators based on the principle of surface charging of functional films are significant in self-powering sensing and energy conversion devices due to their multiple functions and high conversion efficiency, although applications remain limited due to a lack of suitable materials and structures. Here, we investigate a triboelectric-piezoelectric hybrid nanogenerator (TPHNG) in the form of a mousepad for computer user behavior monitoring and energy harvesting. Triboelectric and piezoelectric nanogenerators with different functional films and structures work independently to detect sliding and pressing movements, and the profitable coupling between the two nanogenerators leads to enhanced device outputs/sensitivity. Different mouse operations such as clicking, scrolling, taking-up/putting-down, sliding, moving rate, and pathing can be detected by the device via distinguishable patterns of voltage ranging from 0.6 to 36 V. Based on operation recognition, human behavior monitoring is realized, with monitoring of tasks such as browsing a document and playing a computer game being successfully demonstrated. Energy harvesting from mouse sliding, patting, and bending of the device is realized with output voltages up to 37 V and power up to 48 μW while exhibiting good durability up to 20,000 cycles. This work presents a TPHNG utilizing surface charging for self-powered human behavior sensing and biomechanical energy harvesting.

Introduction Circulating intact parathyroid hormone (iPTH) levels include full-length (1-84) PTH and long C-PTH fragments, but primarily (7-84) PTH, which have been reported to have antagonistic effects on the bones and kidneys. However, their effects on the cardiovascular system remain unclear. In this study, the relationships between the plasma PTH fragments levels and heart rate variability (HRV) in stage 5 chronic kidney disease (CKD5) patients are explored. Furthermore, the effects of parathyroidectomy (PTX) on the above indices are investigated. Methods In this cross-sectional study, 164 healthy controls and 354 CKD5 patients, including 208 secondary hyperparathyroidism (SHPT) subgroup with PTX, were enrolled. Circulating (7-84) PTH levels were calculated by subtracting plasma (1-84) PTH levels from iPTH levels. The HRV parameters were measured using a 24-hour Holter. Results The baseline levels of plasma iPTH, (1-84) PTH, and (7-84) PTH in the CKD5 patients were 930.40 (160.65, 1792.50) pg/mL, 448.60 (99.62, 850.45) pg/mL, and 468.20 (54.22, 922.55) pg/mL, respectively. In the CKD5 patients, plasma (1-84) PTH levels were independently correlated with the standard deviation of the normal-to-normal R-R intervals (SDNN) and the standard deviation of the five-minute average of the normal R-R intervals (SDANN). With a median follow up time of 6.50 months after PTX in the SHPT patients (n = 30), improved SDNN and SDANN markers were related with decreased (1-84) PTH levels. Furthermore, an improved SDNN was related with decreased (7-84) PTH levels. Conclusions The CKD5 patients’ baseline (1-84) PTH levels were correlated with the SDNN and SDANN. After PTX, an improved SDNN was related with decreased (1-84) PTH and (7-84) PTH levels, while improved SDANN was related with decreased (1-84) PTH levels. No antagonistic effects of (1-84) PTH and (7-84) PTH on HRV were found in the CKD5 patients.

Purpose Microwave ablation (MWA) is feasible for severe renal secondary hyperparathyroidism (SHPT) and primary hyperparathyroidism (PHPT) patients ineligible for parathyroidectomy (PTX). Here we compared the clinical manifestations and characteristics of parathyroid glands in these two groups, and summarized the techniques, safety and efficacy of MWA. Methods Baseline clinical characteristics, ablation-related techniques, adverse events/complications, and efficacy were recorded. Results In SHPT group, malnutrition, cardiovascular/pulmonary complications, and abnormal bone metabolism were severe. SHPT patients had more hyperplastic parathyroid glands. The volume of each gland was smaller, and the time of ablation for a single parathyroid was shorter in the SHPT group, although there were no significant differences compared with patients in the PHPT group. Three patients in both groups had recurrent laryngeal nerve injuries and all recovered, except for one SHPT patient. By the end of follow-up, serum iPTH levels had decreased from 2400.26 ± 844.26 pg/mL to 429.39 ± 407.93 pg/mL (p

Despite their rapid growth, the utilisation of application programming interfaces (APIs) poses challenges for companies under pressure to yield productive systems integration. APIs of larger systems tend to be large, complex and have reduced modularity and quality, which makes them cumbersome to comprehend and use. These challenges can be addressed by static API analysis that focuses on studying API code itself and deriving business entities and dependencies from operational signatures. However, existing techniques for static analysis of APIs face the challenges in deriving a sufficient coverage of business entity relationship types from implementation-oriented API operational signatures carrying limited semantic insights. The paper aims to address such problems by supporting static analysis techniques for APIs that improve their modularity. Our approach adopts an object-oriented paradigm where the concept of “object” is exemplified by the notion of business entity. It systematically applies interface analysis methods and techniques for eliciting knowledge of business entities and their attributes, for deriving the temporal order of calling operations across multiple business entities, and for learning and extracting various ways of invoking a service via APIs. The approach is implemented as an open-source tool and applied to a group of widely-deployed services in practice for validation. The research contributes to identifying key aspects of both the structure and behaviour of APIs, which will lead to building a simplified but comprehensive interface (presentation) layer to assist service users in understanding complex and overloaded interfaces as well as to facilitate efficient and effective service integration.

Available, effectively converting low-frequency vibration into available electricity, triboelectric nanogenerator (TENG) is always research hot nowadays. However, the enhancing effect of the existing methods for the output have all sorts of drawbacks, i.e., low efficiency and unstable, and its practical applications still need to be further explored. Here, leveraging core-shell nanoparticles Ag@SiO2 doping into tribo-materials generates the surface plasmon effect to boost the output performance of the TENG. On one hand, the shell alleviated the seepage effect from conventional nanoparticles; on the other hand, the surface plasmon effect enabled the core-shell nanoparticles to further boost the output performance of TENG. We circumvent the limitations and present a TENG whose output power density can be up to 4.375 mW/cm2. Points is that this article novelty investigate the high-performance TENG applicating for traditional Chinese medicine and develop a pratical self-powered acupuncture system. This technology enables rapid, routine regulation of human health at any age, which has potential applications in nearly any setting across healthcare platforms alike.

Tensile tests of 8009Al alloy reinforced with SiC and Al₂O₃ particles fabricated by powder metallurgy (PM) were conducted at temperatures of 250–350°C and strain rates of 0.001–0.1 s⁻¹. The ultimate tensile strength and yield strength of the samples decreased while the temperature and strain rate increased. The elongation slightly decreased at first and then increased with growing temperature because of the medium-temperature brittleness of the alloy matrix. When the strain rate was 0.1 s⁻¹, the elongation of the 8009Al/Al₂O₃ composites always decreased with an increase in temperature because of the poorly coordinated deformation and weak bonding between the matrix and Al₂O₃ particles at such a high strain rate. The work-hardening rates of the composites sharply increased to maxima and then decreased rapidly as the strain increased. Meanwhile, the 8009Al/SiCₚ composites displayed superior UTS, YS, elongation, and work-hardening rates than those of the 8009Al/Al₂O₃ composites under the same conditions. Compared to 8009Al alloys reinforced with spherical Al₂O₃ particle, 8009Al alloys reinforced with irregular SiC particles exhibited a better strengthening effect. The fracture mechanism of the 8009Al/SiCₚ composites was mainly ductile, while that of the 8009Al/Al₂O₃ composites was primarily debonding at the matrix–particle interfaces in a brittle mode.

Purpose: Accurate preoperative parathyroid localization is important for successful parathyroidectomy (PTX). The aim of our study was to investigate whether SPECT/CT has enhanced effect in preoperative localization of parathyroids. Methods: In our retrospective cohort study, we evaluated the effects of technetium-99m methoxyisobutylisonitrile-single-photon emission computed tomography/computed tomography (99mTc-MIBI SPECT/CT) on preoperative parathyroid localization for 645 secondary hyperparathyroidism (SHPT) patients. Among them, 569 successful PTX patients were divided into group A (received 99mTc-MIBI scintigraphy, n = 175) and group B (received 99mTc-MIBI scintigraphy and SPECT/CT imaging, n = 394). Sensitivity, specificity, and consistency of two imaging methods in preoperative localization of parathyroids were compared. Results: Overall sensitivity and consistency were higher in group B, while there was no difference in specificity between the two groups. In group A, the sensitivity of 99mTc-MIBI was 50.00%, 77.11%, 61.76%, and 76.54% in the right upper gland (RU), right lower gland (RL), left upper gland (LU), and left lower gland (LL) subgroups, while the consistency was 52.00%, 76.57%, 61.71%, and 75.43%, respectively. In group B, the sensitivity of 99mTc-MIBI with SPECT/CT was 69.39%, 90.03%, 78.07%, and 84.27%, and the consistency was 69.54%, 88.32%, 78.43%, and 84.26%, respectively. The sensitivity and consistency in lower glands were higher than in upper glands in both groups. Sensitivity for eutopic parathyroid was higher in group B, while there was no difference for ectopic parathyroid. Conclusions: 99mTc-MIBI SPECT/CT can increase the sensitivity and consistency of preoperative localization of eutopic parathyroid glands, and it can accurately locate ectopic parathyroid without sensitivity improvement.

This study investigates the transportation of mass and heat in the 3D flow of hybrid nanofluids over a rotating disk in the presence of uniform magnetic field acting along the transverse direction to the flow. Water is being used as a base fluid mixed with nanoparticles of aluminum oxide and copper. The mathematical model of the nano-fluidic problem in terms of partial differential equations (PDEs) is formulated by considering the velocity slip at the boundary and heat transport mechanism due to thermal convection and radiation effects. Suitable similarity transformations are being used to transform the governing system of PDEs into a system of nonlinear ordinary differential equations. The transformed model is then analyzed by employing the competency of the Lobatto IIIA technique. Numerical and graphical illustrations are presented for the analysis of different flow parameters by varying velocities and temperature profiles. The behavior of heat transfer rate and coefficient skin friction is also studied numerically for the presented flow model.

Film and TV productions, a key area in production screen business, comprise of processes with high demand for creativity and flexibility. However, despite the era of fast developing technology, film production processes are carried out in an old fashioned way. This is reflected, for example, by the fact that document processing accompanied by daily shooting activities is still paper-based and coordinating geographically distributed cast and crew is purely manual. There is an opportunity to bring process innovation into this industry, which can streamline and optimise film production processes and thus reduce production costs. Business Process Management (BPM) is the mainstream contemporary technology-enabled business improvement method. It is proven to provide significant benefits to an organisation in terms of cost savings and responsiveness to changes. In this paper, we apply BPM technology to process innovation for film production. We also share the experience in how to deal with the innovation barriers in the film industry. Over the course of the investigation, a prototype called YAWL4Film was developed on top of a state-of-the-art BPM system. YAWL4Film supports collection and entering of production related data and automatic generation of reports required during film production. The system was deployed in two student productions at the Australian Film Television and Radio School (AFTRS), as well as in a feature film production by Porchlight, an independent film production company.