PrivFR: Privacy-Enhanced Federated Recommendation With Shared Hash Embedding.

Federated recommender systems (FRSs), with their improved privacy-preserving advantages to jointly train recommendation models from numerous devices while keeping user data distributed, have been widely explored in modern recommender systems (RSs). However, conventional FRSs require transmitting the entire model between the server and clients, which brings a huge carbon footprint for cost-conscious cross-device learning tasks. While several efforts have been dedicated to improving the efficiency of FRSs, it's suboptimal to treat the whole model as the objective of compact design. Besides, current research fails to handle the out-of-vocabulary (OOV) issue in real-world FRSs, where the items only occasionally appear in the testing phase but were not observed during the training process, which is another practical challenge and has not been well studied yet. To this end, we propose a privacy-enhanced federated recommendation framework with shared hash embedding, PrivFR, in cross-device settings, which is an efficient representation mechanism specialized for the embedding parameters without compromising the model capability. Specifically, it represents items in a resource-efficient way by delicately utilizing shared hash embedding and multiple hash functions. As such, it just maintains a small shared pool of hash embedding in local clients, rather than fitting all embedding vectors for each item, which can exactly achieve the dual advantages of conserving resources and handling the OOV issue. What's more, we prove that this mechanism can protect the data privacy of local clients from a theoretical perspective. Extensive experiments show that our method not only effectively reduces storage and communication overheads, but also outperforms state-of-the-art FRSs.

Investigating the efficacy and importance of mobile-based assessments for Parkinson's disease: uncovering the potential of novel digital tests.

This study introduces PDMotion, a mobile application comprising 11 digital tests, including those adapted from the MDS-Unified Parkinson's Disease Rating Scale (MDS-UPDRS) Part III and novel assessments, for remote Parkinson's Disease (PD) motor symptoms evaluation. Employing machine learning techniques on data from 50 PD patients and 29 healthy controls, PDMotion achieves accuracies of 0.878 for PD status prediction and 0.715 for severity assessment. A post-hoc explanation model is employed to assess the importance of features and tasks in diagnosis and severity evaluation. Notably, novel tasks that are not adapted from MDS-UPDRS Part III like the circle drawing, coordination test, and alternative tapping test are found to be highly important, suggesting digital assessments for PD can go beyond digitizing existing tests. The alternative tapping test emerges as the most significant task. Using its features alone achieves prediction accuracies comparable to the full task set, underscoring its potential as an independent screening tool. This study addresses a notable research gap by digitalizing a wide array of tests, including novel ones, and conducting a comparative analysis of their feature and task importance. These insights provide guidance for task selection and future development in PD mobile assessments, a field previously lacking such comparative studies.

Reversing skin cancer adversarial examples by multiscale diffusive and denoising aggregation mechanism.

Reliable skin cancer diagnosis models play an essential role in early screening and medical intervention. Prevailing computer-aided skin cancer classification systems employ deep learning approaches. However, recent studies reveal their extreme vulnerability to adversarial attacks - often imperceptible perturbations to significantly reduce the performances of skin cancer diagnosis models. To mitigate these threats, this work presents a simple, effective, and resource-efficient defense framework by reverse engineering adversarial perturbations in skin cancer images. Specifically, a multiscale image pyramid is first established to better preserve discriminative structures in the medical imaging domain. To neutralize adversarial effects, skin images at different scales are then progressively diffused by injecting isotropic Gaussian noises to move the adversarial examples to the clean image manifold. Crucially, to further reverse adversarial noises and suppress redundant injected noises, a novel multiscale denoising mechanism is carefully designed that aggregates image information from neighboring scales. We evaluated the defensive effectiveness of our method on ISIC 2019, a largest skin cancer multiclass classification dataset. Experimental results demonstrate that the proposed method can successfully reverse adversarial perturbations from different attacks and significantly outperform some state-of-the-art methods in defending skin cancer diagnosis models.

Emotional Attention: From Eye Tracking to Computational Modeling.

Attending selectively to emotion-eliciting stimuli is intrinsic to human vision. In this research, we investigate how emotion-elicitation features of images relate to human selective attention. We create the EMOtional attention dataset (EMOd). It is a set of diverse emotion-eliciting images, each with (1) eye-tracking data from 16 subjects, (2) image context labels at both object- and scene-level. Based on analyses of human perceptions of EMOd, we report an emotion prioritization effect: emotion-eliciting content draws stronger and earlier human attention than neutral content, but this advantage diminishes dramatically after initial fixation. We find that human attention is more focused on awe eliciting and aesthetic vehicle and animal scenes in EMOd. Aiming to model the above human attention behavior computationally, we design a deep neural network (CASNet II), which includes a channel weighting subnetwork that prioritizes emotion-eliciting objects, and an Atrous Spatial Pyramid Pooling (ASPP) structure that learns the relative importance of image regions at multiple scales. Visualizations and quantitative analyses demonstrate the model's ability to simulate human attention behavior, especially on emotion-eliciting content.

Effectiveness and usability of the system for assessment and intervention of frailty for community-dwelling pre-frail older adults: A pilot study.

Background: Effective multicomponent interventions in the community targeted at preventing frailty in at-risk older adults can promote healthy ageing. However, there is a lack of studies exploring the effectiveness of technology-enabled autonomous multi-domain community-based interventions for frailty. We developed a novel end-to-end System for Assessment and Intervention of Frailty (SAIF) with exercise, nutrition, and polypharmacy components. This pilot study aimed to explore SAIF's effectiveness in improving frailty status, physical performance and strength, and its usability in pre-frail older adults. Materials and methods: This is a single arm 8-week pilot study in 20 community-dwelling older adults who were pre-frail, defined using the Clinical Frailty Scale (CFS) as CFS 3 + (CFS 3 and FRAIL positive) or CFS 4. For outcomes, we assessed frailty status using the modified Fried Frailty Phenotype (FFP) and CFS; physical performance using Short Physical Performance Battery (SPPB); and Hand Grip Strength (HGS) at baseline and 8-week. User experience was explored using the System Usability Scale (SUS), interest-enjoyment subscale of the Intrinsic Motivation Inventory and open-ended questions. We analyzed effectiveness using repeated-measures tests on pre-post scores, and usability using a convergent mixed-method approach via thematic analysis of open-ended responses and descriptive statistics of usability/interest-enjoyment scales. Results: Sixteen participants (71.8 ± 5.5 years) completed the 8-week study. There was a significant improvement in FFP score (-0.5, p < 0.05, effect size, r = 0.43), but not CFS (-1.0, p = 0.10, r = 0.29). Five (31.3%) improved in frailty status for both FFP and CFS. SPPB (+1.0, p < 0.05, r = 0.42) and HGS (+3.5, p < 0.05, r = 0.45) showed significant improvements. Three themes were identified: "Difficulty in module navigation" (barriers for SAIF interaction); "User engagement by gamification" (facilitators that encourage participation); and "Perceived benefits to physical health" (subjective improvements in physical well-being), which corroborated with SUS (68/100) and interest-enjoyment (3.9/5.0) scores. Taken together, user experience results cohere with the Senior Technology Acceptance and Adoption Model. Conclusion: Our pilot study provides preliminary evidence of the effectiveness of SAIF in improving frailty status, physical performance and strength of pre-frail older adults, and offers user experience insights to plan the follow-up large-scale randomized controlled trial.

Mobile device: a useful tool to teach inhaler devices to healthcare professionals.

BACKGROUND: Proper inhaler device usage is paramount for control of underlying obstructive airway disease. Hence, education to healthcare professionals who will eventually educate patients need to be done effectively. We developed an application for mobile devices for education on six medical inhaler devices, the metered-dose inhaler (MDI), Turbuhaler, Accuhaler, Breezhaler, Ellipta and Respimat, and studied if there were any difference between the application and the manufacturer's instructions on inhaler technique. The aim of this study is to see if inhaler education via a mobile phone app is comparable to manual instruction for health care professions. METHODS: Participants, who were nursing students, were randomized to learn the inhaler devices via the manufacturer's instruction guide or a mobile device app designed specifically for education on inhaler devices. RESULTS: There were 45 participants in each group. 78% of them were females with a median age of 21 (IQR 3). 67% used an Apple mobile device and the remainder used an Android device. The mobile device showed better total improvement points for the Turbuhaler device (262 vs 287 points; P = 0.02). Participants learning from the manufacturer's guide had a significantly higher total improvement points in the Breezhaler (370 vs 327 points; P < 0.01) and Ellipta (214 vs 174 points; P < 0.01) device. Both interventions showed improvement in total scores for demonstrating the correct usage of all inhaler devices. MDI has the least number of correct steps for both interventions. The participants' reported their mean (SD) self-rated knowledge was significantly higher for those using the app for all devices as compared to those that did not (4.33 (0.68) vs 4.73 (0.42); P = < 0.01). Self-reported confidence level was found to be higher in the mobile app group, but this was not statistically significant. The app was well received and scored of 4.42 of 5 with regards to its quality. CONCLUSION: Using a mobile inhaler app is just as effective to teach inhaler device techniques to healthcare professionals and is likely a more convenient, versatile and important adjunct to learning. TRIAL REGISTRATION: National Healthcare Group Ethics Board (2018/00960).

Corrigendum: Effectiveness and usability of the system for assessment and intervention of frailty for community-dwelling pre-frail older adults: A pilot study.

[This corrects the article DOI: 10.3389/fmed.2022.955785.].

Image Visual Realism: From Human Perception to Machine Computation.

Visual realism is defined as the extent to which an image appears to people as a photo rather than computer generated. Assessing visual realism is important in applications like computer graphics rendering and photo retouching. However, current realism evaluation approaches use either labor-intensive human judgments or automated algorithms largely dependent on comparing renderings to reference images. We develop a reference-free computational framework for visual realism prediction to overcome these constraints. First, we construct a benchmark dataset of 2,520 images with comprehensive human annotated attributes. From statistical modeling on this data, we identify image attributes most relevant for visual realism. We propose both empirically-based (guided by our statistical modeling of human data) and deep convolutional neural network models to predict visual realism of images. Our framework has the following advantages: (1) it creates an interpretable and concise empirical model that characterizes human perception of visual realism; (2) it links computational features to latent factors of human image perception.

A dataset of human decision-making in teamwork management.

Today, most endeavours require teamwork by people with diverse skills and characteristics. In managing teamwork, decisions are often made under uncertainty and resource constraints. The strategies and the effectiveness of the strategies different people adopt to manage teamwork under different situations have not yet been fully explored, partially due to a lack of detailed large-scale data. In this paper, we describe a multi-faceted large-scale dataset to bridge this gap. It is derived from a game simulating complex project management processes. It presents the participants with different conditions in terms of team members' capabilities and task characteristics for them to exhibit their decision-making strategies. The dataset contains detailed data reflecting the decision situations, decision strategies, decision outcomes, and the emotional responses of 1,144 participants from diverse backgrounds. To our knowledge, this is the first dataset simultaneously covering these four facets of decision-making. With repeated measurements, the dataset may help establish baseline variability of decision-making in teamwork management, leading to more realistic decision theoretic models and more effective decision support approaches.

Stripe-PZT Sensor-Based Baseline-Free Crack Diagnosis in a Structure with a Welded Stiffener.

This paper proposes a stripe-PZT sensor-based baseline-free crack diagnosis technique in the heat affected zone (HAZ) of a structure with a welded stiffener. The proposed technique enables one to identify and localize a crack in the HAZ using only current data measured using a stripe-PZT sensor. The use of the stripe-PZT sensor makes it possible to significantly improve the applicability to real structures and minimize man-made errors associated with the installation process by embedding multiple piezoelectric sensors onto a printed circuit board. Moreover, a new frequency-wavenumber analysis-based baseline-free crack diagnosis algorithm minimizes false alarms caused by environmental variations by avoiding simple comparison with the baseline data accumulated from the pristine condition of a target structure. The proposed technique is numerically as well as experimentally validated using a plate-like structure with a welded stiffener, reveling that it successfully identifies and localizes a crack in HAZ.

Insights into the hierarchical structure and digestion rate of alkali-modulated starches with different amylose contents.

Combined analytical techniques were used to explore the effects of alkali treatment on the multi-scale structure and digestion behavior of starches with different amylose/amylopectin ratios. Alkali treatment disrupted the amorphous matrix, and partial lamellae and crystallites, which weakened starch molecular packing and eventually enhanced the susceptibility of starch to alkali. Stronger alkali treatment (0.5% w/w) made this effect more prominent and even transformed the dual-phase digestion of starch into a triple-phase pattern. Compared with high-amylose starch, regular maize starch, which possesses some unique structure characteristics typically as pores and crystallite weak points, showed evident changes of hierarchical structure and in digestion rate. Thus, alkali treatment has been demonstrated as a simple method to modulate starch hierarchical structure and thus to realize the rational development of starch-based food products with desired digestibility.

Extubation outcome after a successful spontaneous breathing trial: A multicenter validation of a 3-factor prediction model.

The aim of the present study was to validate, and if necessary update, a predictive model previously developed using a classification and regression tree (CART) algorithm for predicting successful extubation (ES) using a new cohort. This prospective cohort study enrolled adults admitted to 10 intensive care units, who had successfully passed a spontaneous breathing trial (SBT) and were considered ready for extubation. After extubation, the patients were followed up for 48 h. The primary outcome measure was ES, defined as the ability to maintain spontaneous unassisted breathing for >48 h after extubation. The 3-factor CART model was applied to patients in this cohort. The predicted probability of ES for each patient in this validation cohort was calculated based on the original CART model using the Laplace correction method. The performance was assessed by discrimination and calibration. A decision curve analysis was used assess the clinical net benefit (NB). Extubation failure (EF) occurred in 90/530 patients (17%). Among the 90 patients, 72 (13.6%) were reintubated, while 18 patients remained on rescue noninvasive ventilation within 48 h after extubation. The original CART model showed high discrimination but only moderate calibration with predicted probabilities that were systematically lower than expected. The original CART model was updated, and the updated model preserved excellent discrimination (area under the receiver operating characteristic curve, 0.91; 95% confidence interval, 0.87 to 0.93), but exhibited near-perfect calibration (calibration slope, 1; intercept, 0). Between threshold probabilities of 50 and 80%, the NB of using this updated model is significantly improved compared with the current strategy. The updated CART model may be used to estimate the predicted probability of ES after a successful SBT for individual patients. Applying this model appears to produce a substantial clinical consequence with regard to potential reduction in unexpected EFs.

Administration of a loading dose of atorvastatin before percutaneous coronary intervention prevents inflammation and reduces myocardial injury in STEMI patients: a randomized clinical study.

BACKGROUND: Administration of a loading dose of atorvastatin 80 mg/d has been shown to be beneficial in patients with stable coronary artery disease and acute coronary syndromes. However, little is known about the impact and mechanism behind the beneficial effects of loading-dose atorvastatin treatment before percutaneous coronary intervention (PCI), especially for those patients experiencing cardiovascular inflammation in ST-segment elevation myocardial infarction (STEMI). OBJECTIVE: The goal of this randomized clinical study was to investigate whether, before emergency PCI, administration of loading-dose atorvastatin therapy in STEMI patients inhibits inflammation and improves cardiac function during 24 weeks of follow-up. METHODS: A total of 102 STEMI patients were enrolled into 3 groups: group A (n = 32) received 80 mg of atorvastatin before emergency PCI, post-PCI follow-up atorvastatin 40 mg for 4 weeks, and atorvastatin 20 mg for 20 weeks; group B (n = 32) received no pre-PCI loading dose of atorvastatin but did receive atorvastatin 40 mg for 4 weeks and then atorvastatin 20 mg for 20 weeks; and group C (n = 38) received only post-PCI atorvastatin 20 mg for 24 weeks. RESULTS: No differences were found in baseline demographic and angiographic characteristics among the 3 groups. Patients in group A had the lowest plasma levels of high-sensitivity C-reactive protein (hs-CRP), B-type natriuretic peptide (BNP), and matrix metalloproteinase type 9 (MMP-9) (P < 0.05). Patients in group A also showed improvement in heart performance, with significant increases in their left ventricular ejection fraction. To a lesser extent, group B displayed reductions in the plasma levels of hs-CRP, BNP, and MMP-9 at later time points (P < 0.05). Compared with those in group C, patients in group B also exhibited significant improvement in left ventricular ejection fraction (P < 0.05). CONCLUSIONS: Loading-dose atorvastatin therapy before emergency PCI reduced the inflammatory response and myocardial dysfunction in these STEMI patients by lowering hs-CRP, BNP, and MMP-9. Pre-PCI loading-dose atorvastatin treatment may help prevent inflammatory response and improve cardiac function in patients with acute coronary syndromes undergoing emergency PCI. ClinicalTrials.gov identifier: NCT01334671.

Creating an immersive game world with evolutionary fuzzy cognitive maps.

The Evolutionary Fuzzy Cognitive Map improves on serious games by modeling both fuzzy and probabilistic causal relationships among the game's variables. It permits asynchronous updates of the variables so that they can evolve dynamically and stochastically. These improvements give players a more engaging, immersive experience.