Duke University

Presently, the prevalent authentication approaches in smartphones are susceptible to interference from light,0 noise, temperature, and the risk of replay attacks. In light of these vulnerabilities, and taking into account user behavior alongside smartphone interaction patterns, we have developed an innovative behavioral-based authentication system. This system harnesses the distinctiveness of individual keystroke dynamics for secure user authentication, offering resilience against noise and light fluctuations. In this unique approach, our smartphone’s speakers and microphones emit and capture high-frequency acoustic signals (AS). To the best of our knowledge, this is the first instance of employing the Doppler effect generated by the high-frequency AS in response to keystroke activity as a distinctive user feature. Our definition of ’keystroke behavior’ encompasses the motions involved in tapping screen buttons while holding the smartphone, effectively capturing unique user attributes without necessitating any special procedures or passwords. Our initial experiments have convincingly shown that the AS Doppler effect, triggered by keystroke actions, is uniquely identifiable per user during button presses. Subsequently, we utilized a Convolutional Autoencoder (CAE) to distill keystroke behaviors from the reflected signals, employing a One-Class Support Vector Machine (OCSVM) for user authentication and identification processes. We then implemented a prototype of this scheme on smartphones and rigorously tested its performance across four real-world scenarios. The outcomes are promising, demonstrating that our scheme not only withstands disturbances from noise and light but also achieves an impressive average accuracy rate of 95.08%. Regarding security, it effectively thwarts replay and record attacks, further underscoring its robustness and reliability.

The cardiac pericardium is a multilayer tissue that envelops the heart. Physiologic roles of the pericardium include: (1) protecting the heart, (2) creating a compartment for containing the pericardial fluid, and (3) providing this compartment for optimizing cardiac movements and functionality within the thoracic cavity. Compared to other cardiothoracic tissues, the pericardium exhibits material properties that vary based on location within the greater pericardial sac. These properties are a result of varied collagen fiber alignments, collagen densities throughout the pericardial sac, and overall cardiac health. Initiatives to better understand the tensile properties of human and large mammalian pericardium have driven both therapies and medical device manufacturing. Recent therapeutic approaches have been directed to exploit the space that exists between the pericardium and the epicardial surface of the heart. New devices and techniques are continually being developed to access this space with minimally invasive approaches. The pharmacokinetics of many drugs may be greatly enhanced if the drug is delivered into the pericardium. Advancements in transcatheter bioprosthetic heart valves have also motivated studies to better translate the biomechanical properties of large mammalian pericardium into more efficient heart valve designs. Developing preclinical research methods to better test and analyze native and treated pericardium will in turn improve the performances of bioprosthetic pericardial implanted devices and the quality of life of implant recipients. Therefore, as the complexities of implantable pericardial devices and pericardial access therapies increase, it remains crucial to better understand the properties of pericardium from both humans and commonly used animal models and tissues.

Nonstructural carbohydrate (NSC) concentrations might reflect the strategies described in the leaf economic spectrum (LES) due to their dependence on photosynthesis and respiration. We examined if NSC concentrations correlate with leaf structure, chemistry, and physiology traits for 114 species from 19 sites and 5 biomes around the globe. Total leaf NSC concentrations varied greatly from 16 to 199 mg g⁻¹ dry mass and were mostly independent of leaf gas exchange and the LES traits. By contrast, leaf NSC residence time was shorter in species with higher rates of photosynthesis, following the fast‐slow strategies in the LES. An average leaf held an amount of NSCs that could sustain one night of leaf respiration and could be replenished in just a few hours of photosynthesis under saturating light, indicating that most daily carbon gain is exported. Our results suggest that NSC export is clearly linked to the economics of return on resource investment.

This paper describes laser exposure to tune the infrared (IR) emissivity of a film of eutectic gallium indium (EGaIn) particles. EGaIn – a liquid metal at room temperature – forms a native oxide that keeps particles of the metal from spontaneously percolating. Photothermal energy from a CO2 laser percolates the particles into a conductive network. Here, it also causes a decrease in the IR emissivity of the film of particles from 0.4 to 0.24 over the range of 7.5–13 µm wavelength (measured by an IR camera) with the increase of laser fluence from 1.4 to 1.9 J cm⁻². The particles percolate most prominently at the bottom of the film, and thus, the apparent surface roughness does not change with laser exposure. This finding suggests the decrease in emissivity is not due to changes in the film's topography. Instead, the change in IR emissivity is attributed to a loss of the surface plasmonic resonance effect of EGaIn particles in the IR range after the sintering, which is confirmed by optical simulations. As a demonstration, it is shown that the ability to change the emissivity makes it possible to encrypt messages and camouflage laser‐processed patterns.

We study a family of structure-preserving deterministic numerical schemes for Lindblad equations. This family of schemes has a simple form and can systemically achieve arbitrary high-order accuracy in theory. Moreover, these schemes can also overcome the non-physical issues that arise from many traditional numerical schemes. Due to their preservation of physical nature, these schemes can be straightforwardly used as backbones for further developing randomized and quantum algorithms in simulating Lindblad equations. In this work, we systematically study this family of structure-preserving deterministic schemes and perform a detailed error analysis, which is validated through numerical examples.

In 2017, the Economic Community of West African States launched its Policy for Gender Mainstreaming in Energy Access with each member state tasked to create a National Action Plan (NAP). This study explores the case of Sierra Leone to understand how stakeholders have influenced the NAP process, and what that might mean for implementation. Using the Actors, Objectives, Context framework with interview data from key gender and energy actors, we find elements that make Sierra Leone’s NAP unique, such as addressing systematic barriers, explicitly promoting solar energy and heavily relying on civil society for implementation. The study finds that a policy focus on gender reflects both a push from international donor organisations and the national efforts of civil society. Sierra Leone’s NAP reflects the tension between the high priority of gender politics at the national and international level and more immediate interest from the Ministry of Energy and government actors on economic development. It affirms the uneasy co-existence of the liberal and liberating view of women in international development with some evidence that the liberal view prevails. Still, despite the government’s focus on economic development, poor engagement with the private sector and cultural barriers are likely to hamper the inclusion of women into the energy industry.

Importance Medicare finances health care for most US patients with end-stage kidney disease (ESKD), regardless of age. Medicare enrollment may have slowed for patients with incident ESKD who gained access to new private insurance options with the 2014 passage of the Affordable Care Act (ACA) and introduction of the ACA Marketplace. Objective To describe trends in public and private insurance coverage and dialysis spending among patients with incident ESKD from 2012 to 2017. Design, Setting, and Participants This serial cross-sectional study included patients 18 to 64 years old in Colorado who were not enrolled in Medicare at dialysis initiation. Data analysis was conducted from May to August 2023. Exposure Introduction of the ACA Marketplace in 2014. Main Outcomes and Measures Medicare, Medicaid, or private insurance enrollment in the first year after dialysis initiation, and dialysis spending by insurance type. Results Of 2005 patients included in the sample, 1416 (70.6%) were 45 to 64 years old, and 1259 (62.8%) were male. A lower proportion of patients with incident ESKD starting dialysis were newly enrolled in Medicare in the years following the ACA (361 of 713 [50.6%]) compared to 2 years prior (420 of 595 [70.6%]). Unadjusted rates of switching from Medicaid to Medicare 1 year after dialysis initiation decreased 14.3 percentage points over time (68.9% in 2012-2013 vs 58.3% and 54.6% in 2014-2015 and 2016-2017, respectively). Unadjusted rates of switching from private insurance to Medicare 1 year after dialysis initiation decreased by 22.3 percentage points (68.1% in 2012-2013 vs 52.2% and 45.8% in 2014-2015 and 2016-2017, respectively). Over the entire 2012 to 2017 period, quarterly dialysis spending in the first year of dialysis among patients with private insurance was higher than among those with Medicare coverage ($26 351-$29 781 vs $10 039-$12 741). Conclusions and Relevance This cross-sectional study demonstrates that lower Medicare enrollment rates over time among those initiating dialysis may be inducing higher social spending. This finding raises concerns about the effectiveness of Medicare policies and federal leverage to improve access, outcomes, and value of dialysis care.

Let $\mathscr {B}$ be a set of n unit balls in ${\mathbb {R}}^3$. We present a linear-size data structure for storing $\mathscr {B}$ that can determine in $O^*(\sqrt{n})$ time whether a query line intersects any ball of $\mathscr {B}$ and report all k such balls in additional O(k) time. The data structure can be constructed in $O(n\log n)$ time. (The $O^*(\cdot )$ notation hides subpolynomial factors, e.g., of the form $O(n^{{\varepsilon }})$, for arbitrarily small ${\varepsilon }> 0$, and their coefficients which depend on ${\varepsilon }$.) We also consider the dual problem: Let $\mathscr {L}$ be a set of n lines in ${\mathbb {R}}^3$. We preprocess $\mathscr {L}$, in $O^*(n^2)$ time, into a data structure of size $O^*(n^2)$ that can determine in $O(\log {n})$ time whether a query unit ball intersects any line of $\mathscr {L}$, or report all k such lines in additional O(k) time.

In this study, we consider the question of repairing and recovering a low-dimensional manifold embedded in high-dimensional space from noisy scattered data. Given a noisy point cloud sampled from a low-dimensional manifold, suppose that part of the scattered data is missing, which results in holes. In these settings, the main goal is to accurately and efficiently reconstruct the information within these gaps. While in three-dimensions the problem has been extensively studied, the challenge of reinstating missing information for high-dimensional manifolds remains open. In this paper, we propose a new approach named Manifold Repairing via Locally Optimal Projection (R-MLOP). The method is defined as a minimization problem with three terms. First, we leverage the spatial proximity to the holes to balance between denoising the data and preserving geometric continuity among points situated along the hole’s boundaries. In addition, a penalty term is added to guarantee a quasi-uniform sampling of the unknown manifold. We prove that the suggested solution recovers missing information inside the hole, with an approximation order that is controlled by the density of the given scattered data as well as the size of the amended hole. The effectiveness of our approach is demonstrated by considering different manifold topologies, for single and multiple-hole repairing, in low and high dimensions.

OBJECTIVE The Endoscopic Third Ventriculostomy Success Score (ETVSS) is a useful decision-making heuristic when considering the probability of surgical success, defined traditionally as no repeat cerebrospinal fluid diversion surgery needed within 6 months. Nonetheless, the performance of the logistic regression (LR) model in the original 2009 study was modest, with an area under the receiver operating characteristic curve (AUROC) of 0.68. The authors sought to use a larger dataset to develop more accurate machine learning (ML) models to predict endoscopic third ventriculostomy (ETV) success and also to perform the largest validation of the ETVSS to date. METHODS The authors queried the MarketScan national database for the years 2005–2022 to identify patients < 18 years of age who underwent first-time ETV and subsequently had at least 6 months of continuous enrollment in the database. The authors collected data on predictors matching the original ETVSS: age, etiology of hydrocephalus, and history of any previous shunt placement. Next, they used 6 ML algorithms—LR, support vector classifier, random forest, k-nearest neighbors, Extreme Gradient Boosted Regression (XGBoost), and naive Bayes—to develop predictive models. Finally, the authors used nested cross-validation to assess the models’ comparative performances on unseen data. RESULTS The authors identified 2047 patients who met inclusion criteria, and 1261 (61.6%) underwent successful ETV. The performances of most ML models were similar to that of the original ETVSS, which had an AUROC of 0.693 on the validation set and 0.661 (95% CI 0.600–0.722) on the test set. The authors’ new LR model performed comparably with AUROCs of 0.693 on both the validation and test sets, with 95% CI 0.633–0.754 on the test set. Among the more complex ML algorithms, XGBoost performed best, with AUROCs of 0.683 and 0.672 (95% CI 0.609–0.734) on the validation and test sets, respectively. CONCLUSIONS This is the largest external validation of the ETVSS, and it confirms modest performance. More sophisticated ML algorithms do not meaningfully improve predictive performance compared to ETVSS; this underscores the need for higher utility, novelty, and dimensionality of input data rather than changes in modeling strategies.

Background Evidence is limited on insured patients’ use of safety net providers as vertically integrated health systems spread throughout the United States. Objectives To examine whether market-level health system penetration is associated with: (1) switches in Medicare beneficiaries’ usual source of primary care from federally qualified health centers (FQHCs) to health systems; and (2) FQHCs’ overall Medicare patient and visit volume. Research Design Beneficiary-level discrete-time survival analysis and market-level linear regression analysis using Medicare fee-for-service claims data from 2013 to 2018. Subjects A total of 659,652 Medicare fee-for-service beneficiaries aged 65 and older lived in one of 27,386 empirically derived primary care markets whose usual source of care in 2013 was an FQHC or a non-FQHC–independent physician organization that predominantly served low-income patients. Measures Beneficiary-year measure of the probability of switching to health system-affiliated physician organizations and market-year measures of the number of FQHC visits by Medicare beneficiaries, number of beneficiaries attributed to FQHCs, and FQHC Medicare market shares. Results During 2013–2018, 16.5% of beneficiaries who sought care from FQHCs switched to health systems. When health system penetration increases from the 25th to 75th percentile, the probability of Medicare FQHC patient switching increases by 4.6 percentage points, with 22 fewer Medicare FQHC visits and 4 fewer beneficiaries attributed to FQHCs per market year. Complex patients and patients who sought care from non-FQHC, independent physician organizations exhibited higher rates of switching to health systems. Conclusions Health system expansion was associated with the loss of Medicare patients by FQHCs, suggesting potential negative spillovers of vertical integration on independent safety net providers.

Background Suicide is the third-leading cause of death among US adolescents aged 10-19 years, and about 10% attempt suicide each year. School-based universal prevention may reduce youth suicidal behavior. Sources of Strength uses a peer leader network diffusion model to promote healthy norms across a school population. A key challenge within schoolwide programs is reaching a large and diverse array of students, especially those less engaged with their peers. Motivated by this challenge, we developed and field-tested Text4Strength—a program of automated text messages targeting help-seeking attitudes and norms, social coping resources, and emotion regulation skills. Objective This study conducted a pilot randomized controlled trial of Text4Strength in 1 high school as an extension of an ongoing schoolwide program (Sources of Strength), to test its impact on targets that have the potential to reduce suicidal behavior. Methods Students at an upstate New York high school (N=223) received 1-2 text messages per week for 9 weeks, targeting strategies for coping with difficult feelings and experiences through clarifying emotions and focusing on positive affect concepts, awareness, and strengthening of youth-adult relationships; and positive help-seeking norms, skills, and resources. Surveys were administered at baseline, immediately post intervention and 3 months after texting ended. We measured proximal intervention targets (methods of coping during stressful events, ability to make sense of their own emotions, feelings of powerlessness during emotion management and recovery, relations with trusted adults at school, and help-seeking behaviors), symptoms and suicide ideation, and student replies to messages. Results No significant effects were observed for any outcome at either follow-up time point. Results showed that if there is a true (but undetected) intervention effect, it is small. Students with fewer friend nominations did not interact any more or less with the text messages. Exploratory moderation analyses observed no interaction between the intervention condition and the number of friends or baseline suicide ideation at any time point. Conclusions In contrast to a promising previous field test, these results suggest that Text4Strength is unlikely to have impacted the outcomes of interest and that undetected moderate or large effects can be ruled out with high confidence. Although motivated by the need to reach more isolated students, students with fewer friends did not engage more or show a greater effect than other participants. This study was conducted in a single high school that was already implementing Sources of Strength, so the bar for showing a distinct effect from texting alone was high. Many further channels for reaching youth through private messaging remain unexplored. Alternative delivery systems should be investigated, such as embedding messaging in gaming chat systems and other media. More sophisticated systems drawing on chatbots may also achieve better outcomes. Trial Registration ClinicalTrials.gov NCT03145363; https://clinicaltrials.gov/study/NCT03145363

In studying the association between clinical measurements and time‐to‐event outcomes within a cure model, utilizing repeated observations rather than solely baseline values may lead to more accurate estimation. However, there are two main challenges in this context. First, longitudinal measurements are usually observed at discrete time points and second, for diseases that respond well to treatment, a high censoring proportion may occur by the end of the trial. In this article, we propose a joint modelling approach to simultaneously study the longitudinal observations and time‐to‐event outcome with an assumed cure fraction. We employ the functional principal components analysis (FPCA) to model the longitudinal data, offering flexibility by not assuming a specific form for the longitudinal curve. We used a Cox's proportional hazards mixture cure model to study the survival outcome. To investigate the longitudinal binary observations, we adopt a quasi‐likelihood method which builds pseudo normal distribution for the binary data and use the E‐M algorithm to estimate the parameters. The tuning parameters are selected using the Akaike information criterion. Our proposed method is evaluated through extensive simulation studies and applied to a clinical trial data to study the relationship between the longitudinal prostate specific antigen (PSA) measurements and overall survival in men with metastatic prostate cancer.

HIV incidence among transgender women remains high and disproportionately impacts young, Black, and Latina transgender women. Data on preferred PrEP modalities among this population are limited. Participants in The LITE Cohort completed a survey module on PrEP modality preferences during 24-month study visits. We summarized ranked preferences based on an exhaustive set of 10 head-to-head comparisons of 5 PrEP modalities (pill, injection, implantable device, topical gel, and intravenous antibodies) and conducted in-depth interviews to contextualize findings. Between 2020 and 2022, 789 participants completed the PrEP modality survey module. The most preferred PrEP modality was the implant (ranked first among 45% of respondents), followed by pill (21%), injection (19%), gel (10%), and intravenous antibodies (4%). The implant ranked highest among Latina transgender women (36%), young adult transgender women (ages 18–24 years; 41%), those living in the South (47%), and those with PrEP indication(s) (45%), while injection was the top-ranked modality among Black transgender women (30%). Qualitative analysis of in-depth interviews (n = 45) revealed that PrEP modality preferences were individualized, context-dependent, considered gender-related factors (e.g. gender-affirming hormone injections), and informed by prior healthcare experiences, personal values, and anticipated modality-specific facilitators and barriers. Our findings suggest high interest in long-acting PrEP options, including implants and injections, and daily pills among transgender women.