Search

This study addresses the need for inclusive AI education by focusing on marginalized female students who historically lack access to learning opportunities in computing. It applies the theoretical framework of intersectionality to understand how gender, race and ethnicity intersect to shape these students' learning experiences and outcomes. Specifically, this study investigated 27 high-school students' cognitive engagement in machine learning practices. We conducted the Wilcoxon-Mann-Whitney test to explore differences in cognitive engagement between marginalized female students and their peers, employed comparative content analysis to delve into significant differences and analysed interview data thematically to gain deeper insights into students' machine learning model development processes. The findings indicated that, when engaging in machine learning practices requiring drawing diverse cultural perspectives, marginalized female students demonstrated significantly higher performance compared to their peers. In particular, marginalized female students exhibited strengths in holistic language analysis, paying attention to writers' intentions and recognizing cultural nuances in language. This study suggests that integrating language analysis and machine learning across subjects has the potential to empower marginalized female students and amplify their perspectives. Furthermore, it calls for a strengths-based approach to reshape the narrative of underrepresentation and promote equitable participation in machine learning and AI.

Abstract Background A surgical site infection (SSI) is a postoperative infection that occurs at or near the surgical incision. SSIs significantly increase morbidity, mortality, length of hospital stay, and healthcare costs. The World Health Organization (WHO) has established hospital hygiene precaution guidelines for the prevention of SSIs, which were enhanced during the COVID-19 pandemic. The current study aims to explore the effect of the COVID-19 pandemic on SSI incidence among initially uninfected postoperative patients. We hypothesize that these enhanced precautions would reduce the incidence of SSIs. Materials and methods A retrospective study comparing surgical outcomes before and during the pandemic. Patients who had abdominal surgery between June and December 2019 (Non-COVID-19) or between February and June 2020 (COVID-19) were included. The two groups were matched in a 1:1 ratio based on age, Sex, acuity (elective or emergent), surgical approach, and comorbidities. Electronic medical records were reviewed to identify SSIs and hospital readmissions within 30 days after surgery. Pearson's chi-square test and Fisher's exact test were used. Results Data was collected and analyzed from 976 patients who had surgery before the COVID-19 pandemic (non-COVID group) and 377 patients who had surgery during the pandemic (COVID group). After matching, there were 377 patients in each group. In our study, we found 23 surgical site infections (SSIs) in both laparoscopic and open surgeries. The incidence of SSIs was significantly higher in the non-COVID period compared to the COVID period [17 cases (4.5%) vs. 6 cases (1.6%), respectively, p = 0.032], especially in non-COVID open surgeries. The incidence of SSIs in laparoscopic surgeries was also higher during the non-COVID period, but not statistically significant. Conclusions Enhanced hygiene precautions during the COVID -19 pandemic may have reduced SSIs rates following abdominal surgery.

Both areal and phylogenetic affiliation have been discussed as driving factors of the distribution of word order in the languages of the world. However, disentangling the interaction of these two factors is challenging. Here we take Indo-European as a test case. Word order in this family is largely homogeneous both within areas and within branches, which makes it difficult to assess which factor was more important in shaping the present-day distribution. To break out of this impasse we turn to corpus data and explicit statistical modeling. Building on a parallel corpus of movie subtitles, we investigate word order on the sentence level under stable pragmatic conditions. We measure the similarity of word order variation between pairs of languages with an information-theoretic distance metric. Using cluster analysis and variation partitioning methods these distance metrics show that phylogenetic distance predicts more variation than geographical distance, but the most important predictor is the shared fraction where phylogeny and area overlap. We conclude that word order has evolved along both dimensions and cannot be reduced to a single one.

Abstract A zinc metallopeptidase neurolysin (Nln) processes diverse bioactive peptides to regulate signaling in the mammalian nervous system. To understand how Nln interacts with various peptides with dissimilar sequences, we determined crystal structures of Nln in complex with diverse peptides including dynorphins, angiotensin, neurotensin, and bradykinin. The structures show that Nln binds these peptides in a large dumbbell-shaped interior cavity constricted at the active site, making minimal structural changes to accommodate different peptide sequences. The structures also show that Nln readily binds similar peptides with distinct registers, which can determine whether the peptide serves as a substrate or a competitive inhibitor. We analyzed the activities and binding of Nln toward various forms of dynorphin A peptides, which highlights the promiscuous nature of peptide binding and shows how dynorphin A (1–13) potently inhibits the Nln activity while dynorphin A (1–8) is efficiently cleaved. Our work provides insights into the broad substrate specificity of Nln and may aid in the future design of small molecule modulators for Nln.

Abstract Chloride deposits are markers for early Mars’ aqueous past, with important implications for our understanding of the martian climate and habitability. The Colour and Stereo Surface Imaging System (CaSSIS) onboard ESA’s Trace Gas Orbiter provides high-resolution color-infrared images, enabling a planet-wide search for (small) potentially chloride-bearing deposits. Here, we use a neural network to map potentially chloride-bearing deposits in CaSSIS images over a significant fraction of the planet. We identify 965 chloride deposit candidates with diameters ranging from 3000 m, including previously unknown deposits, 136 (~14%) of which are located in the highlands north of the equator, up to ~36°N. Northern chloride candidates tend to be smaller than in the south and are predominantly located in small-scale topographic depressions in low-albedo Noachian and Hesperian highland terranes. Our new dataset augments existing chloride deposit maps, informs current and future imaging campaigns, and enables future modelling work towards a better understanding of the distribution of near-surface water in Mars’ distant past.

Abstract Hyperpolarization and cyclic nucleotide (HCN) activated ion channels are critical for the automaticity of action potentials in pacemaking and rhythmic electrical circuits in the human body. Unlike most voltage-gated ion channels, the HCN and related plant ion channels activate upon membrane hyperpolarization. Although functional studies have identified residues in the interface between the voltage-sensing and pore domain as crucial for inverted electromechanical coupling, the structural mechanisms for this unusual voltage-dependence remain unclear. Here, we present cryo-electron microscopy structures of human HCN1 corresponding to Closed, Open, and a putative Intermediate state. Our structures reveal that the downward motion of the gating charges past the charge transfer center is accompanied by concomitant unwinding of the inner end of the S4 and S5 helices, disrupting the tight gating interface observed in the Closed state structure. This helix-coil transition at the intracellular gating interface accompanies a concerted iris-like dilation of the pore helices and underlies the reversed voltage dependence of HCN channels.

The language comprehension system preferentially assumes that agents come first during incremental processing. While this might reflect a biologically fixed bias, shared with other domains and other species, the evidence is limited to languages that place agents first, and so the bias could also be learned from usage frequency. Here, we probe the bias with electroencephalography (EEG) in Äiwoo, a language that by default places patients first, but where sentence-initial nouns are still locally ambiguous between patient or agent roles. Comprehenders transiently interpreted nonhuman nouns as patients, eliciting a negativity when disambiguation was toward the less common agent-initial order. By contrast and against frequencies, human nouns were transiently interpreted as agents, eliciting an N400-like negativity when the disambiguation was toward patient-initial order. Consistent with the notion of a fixed property, the agent bias is robust against usage frequency for human referents. However, this bias can be reversed by frequency experience for nonhuman referents.

Abstract Focusing on a specific conversation amidst multiple interfering talkers is challenging, especially for those with hearing loss. Brain‐controlled assistive hearing devices aim to alleviate this problem by enhancing the attended speech based on the listener's neural signals using auditory attention decoding (AAD). Departing from conventional AAD studies that relied on oversimplified scenarios with stationary talkers, a realistic AAD task that involves multiple talkers taking turns as they continuously move in space in background noise is presented. Invasive electroencephalography (iEEG) data are collected from three neurosurgical patients as they focused on one of the two moving conversations. An enhanced brain‐controlled assistive hearing system that combines AAD and a binaural speaker‐independent speech separation model is presented. The separation model unmixes talkers while preserving their spatial location and provides talker trajectories to the neural decoder to improve AAD accuracy. Subjective and objective evaluations show that the proposed system enhances speech intelligibility and facilitates conversation tracking while maintaining spatial cues and voice quality in challenging acoustic environments. This research demonstrates the potential of this approach in real‐world scenarios and marks a significant step toward developing assistive hearing technologies that adapt to the intricate dynamics of everyday auditory experiences.

Over the last two decades, an emerging body of research has demonstrated that non-human animals exhibit the ability to combine context-specific calls into larger sequences. These structures have frequently been compared with language’s syntax, whereby linguistic units are combined to form larger structures, and leveraged to argue that syntax might not be unique to language. Currently, however, the overwhelming majority of examples of call combinations are limited to simple sequences comprising just two calls which differ dramatically from the open-ended hierarchical structuring of the syntax found in language. We revisit this issue by taking a whole-repertoire approach to investigate combinatoriality in common marmosets (Callithrix jacchus). We use Markov chain models to quantify the vocal sequences produced by marmosets providing evidence for structures beyond the bigram, including three-call and even combinations of up to eight or nine calls. Our analyses of these longer vocal sequences are suggestive of potential further internal organization, including some amount of recombination, nestedness and non-adjacent dependencies. We argue that data-driven, whole-repertoire analyses are fundamental to uncovering the combinatorial complexity of non-human animals and will further facilitate meaningful comparisons with language’s combinatoriality.

Summary: Background: Blood-based biomarkers offer a promising, less invasive, and more cost-effective alternative for Alzheimer's disease screening compared to cerebrospinal fluid or imaging biomarkers. However, they have been extensively studied only in memory clinic-based cohorts. We aimed to validate them in a more heterogeneous, older patient population from primary care. Methods: We measured plasma Aβ42/Aβ40, P-tau181, NfL, and GFAP in 1007 individuals without dementia, aged 79–94 years, from the longitudinal, primary care-based German AgeCoDe study. We assessed the association with cognitive decline, disease progression, and the capacity to predict future dementia of the Alzheimer's type (DAT). We also evaluated biomarker dynamics in 305 individuals with a follow-up sample (∼8 years later). Findings: Higher levels of P-tau181 (HR = 1.32 [95% CI: 1.17–1.51]), NfL (HR = 1.19 [95% CI: 1.03–1.36]), and GFAP (HR = 1.36 [95% CI: 1.22–1.52]), and a lower Aβ42/Aβ40 ratio (HR = 0.80 [95% CI: 0.68–0.95]) were associated with an increased risk of progressing to clinically-diagnosed DAT. Additionally, higher levels of P-tau181 (β = −0.49 [95% CI: −0.71 to 0.26]), NfL (β = −0.29 [95% CI: −0.52 to 0.06]), and GFAP (β = −0.60 [95% CI: −0.83 to 0.38]) were linked to faster cognitive decline. A two-step DAT prediction strategy combining initial MMSE with biomarkers improved the identification of individuals in the prodromal stage for potential treatment eligibility. Biomarker levels changed over time, with increases in P-tau181 (β = 0.19 [95% CI: 0.14–0.25]), NfL (β = 2.88 [95% CI: 2.18–3.59]), and GFAP (β = 8.23 [95% CI: 6.71–9.75]). NfL (β = 2.47 [95% CI: 1.04–3.89]) and GFAP (β = 4.45 [95% CI: 1.38–7.51]) exhibited a faster increase in individuals progressing to DAT. Interpretation: Evaluating plasma biomarkers, alongside brief cognitive assessments, might enhance the precision of risk assessment for DAT progression in primary care. Funding: Alzheimer Forschung Initiative, Bundesministerium für Bildung und Forschung.

BackgroundThe public launch of OpenAI’s ChatGPT platform generated immediate interest in the use of large language models (LLMs). Health care institutions are now grappling with establishing policies and guidelines for the use of these technologies, yet little is known about how health care providers view LLMs in medical settings. Moreover, there are no studies assessing how pediatric providers are adopting these readily accessible tools. ObjectiveThe aim of this study was to determine how pediatric providers are currently using LLMs in their work as well as their interest in using a Health Insurance Portability and Accountability Act (HIPAA)–compliant version of ChatGPT in the future. MethodsA survey instrument consisting of structured and unstructured questions was iteratively developed by a team of informaticians from various pediatric specialties. The survey was sent via Research Electronic Data Capture (REDCap) to all Boston Children’s Hospital pediatric providers. Participation was voluntary and uncompensated, and all survey responses were anonymous. ResultsSurveys were completed by 390 pediatric providers. Approximately 50% (197/390) of respondents had used an LLM; of these, almost 75% (142/197) were already using an LLM for nonclinical work and 27% (52/195) for clinical work. Providers detailed the various ways they are currently using an LLM in their clinical and nonclinical work. Only 29% (n=105) of 362 respondents indicated that ChatGPT should be used for patient care in its present state; however, 73.8% (273/368) reported they would use a HIPAA-compliant version of ChatGPT if one were available. Providers’ proposed future uses of LLMs in health care are described. ConclusionsDespite significant concerns and barriers to LLM use in health care, pediatric providers are already using LLMs at work. This study will give policy makers needed information about how providers are using LLMs clinically.

Dietary modifications such as caloric restriction (CR) and intermittent fasting (IF) have gained popularity due to their proven health benefits in aged populations. In time restricted feeding (TRF), a form of intermittent fasting, the amount of time for food intake is regulated without restricting the caloric intake. TRF is beneficial for the central nervous system to support brain health in the context of aging. Therefore, we here ask whether TRF also exerts beneficial effects in the aged retina. We compared aged mice (24 months) on a TRF paradigm (access to food for six hours per day) for either 6 or 12 months against young control mice (8 months) and aged control mice on an ad libitum diet. We examined changes in the retina at the functional (electroretinography), structural (histology and fluorescein angiograms) and molecular (gene expression) level. TRF treatment showed amelioration of age-related reductions in both scotopic and photopic b-wave amplitudes suggesting benefits for retinal interneuron signaling. TRF did not affect age-related signs of retinal inflammation or microglial activation at either the molecular or histological level. Our data indicate that TRF helps preserve some aspects of retinal function that are decreased with aging, adding to our understanding of the health benefits that altered feeding patterns may confer.

Cocoa pulp, a by-product of the cocoa supply chain, has gained the food sector's attention for its pleasant flavour. Enzyme-assisted technologies to facilitate its separation from the cocoa seeds remain largely unexplored. We studied the effects of temperature, enzyme activity and enzyme combinations on the physicochemical properties of cocoa pulp by Response Surface Methodology and D-optimal design. Endo-polygalacturonase, endo-cellulase and hemicellulase activities were investigated. A reduced quadratic model described the relative reduction in viscosity, of the particle diameter dv,0.9 and of the browning index. Polygalacturonase had the strongest influence on viscosity, while cellulase had the least effect. Synergistic effects of polygalacturonase with cellulase and/or hemicellulase were identified. The model predicted maximal reductions in viscosity by 70.3 % with 275 U of polygalacturonase and 275 U of hemicellulase at 40 °C. Fresh pulp's dv,0.9 was 613 μm. Combining all enzyme activities reduced the dv,0.9–418 μm (40 °C, total activity: 580 U). Only the temperature proved to significantly influence the browning index, with higher temperatures causing a higher colouring effect. Lastly, the model was successfully validated for the relative reduction in viscosity and the dv,0.9. Our study provides key insights into the enzyme-assisted processing of cocoa pulp, facilitating its separation from the seeds. By utilising the pulp, innovative ingredients can be developed for the food sector, and at the same time, the added value in the cocoa supply chain can be improved.

IntroductionThe antiviral activity of recombinant bovine interferon lambda 3 (bovIFN-λ3) against bovine viral diarrhea virus (BVDV) has been demonstrated in vitro in Madin-Darby bovine kidney cells (MDBK) and in vivo in cattle. However, anti-BVDV activity of bovIFN-λ3 has not been studied in bovine respiratory tract epithelial cells, supposedly a primary target of BVDV infection when entering the host by the oronasal route.MethodsHere we investigated the anti-BVDV activity of bovIFN-λ3 in bovine turbinate-derived primary epithelial cells (BTu) using BVDV infection and immunoperoxidase staining, TCID50, RT-qPCR, DNA and transcriptome sequencing, and transfection with plasmids containing the two subunits, IL-28Rα and IL-10Rβ that constitute the bovIFN-λ3 receptor.ResultsOur immunoperoxidase staining, RT-qPCR, and TCID50 results show that while BVDV was successfully cleared in MDBK cells treated with bovIFN-λ3 and bovIFN-α, only the latter, bovIFN-α, cleared BVDV in BTu cells. Preincubation of MDBK cells with bovIFN-λ3 before BVDV infection was needed to induce optimal antiviral state. Both cell types displayed intact type I and III IFN signaling pathways and expressed similar levels of IL-10Rβ subunit of the type III IFN receptor. Sequencing of PCR amplicon of the IL-28Rα subunit revealed intact transmembrane domain and lack of single nucleotide polymorphisms (SNPs) in BTu cells. However, RT-qPCR and transcriptomic analyses showed a lower expression of IL-28Rα transcripts in BTu cells as compared to MDBK cells. Interestingly, transfection of BTu cells with a plasmid encoding IL-28Rα subunit, but not IL-10Rβ subunit, established the bovIFN-λ3 sensitivity showing similar anti-BVDV activity to the response in MDBK cells.ConclusionOur results demonstrate that the sensitivity of cells to bovIFN-λ3 depends not only on the quality but also of the quantity of the IL-28Rα subunit of the heterodimeric receptor. A reduction in IL-28Rα transcript expression was detected in BTu as compared to MDBK cells, despite the absence of spliced variants or SNPs. The establishment of bovIFN-λ3 induced anti-BVDV activity in BTu cells transfected with an IL-28Rα plasmid suggests that the level of expression of this receptor subunit is crucial for the specific antiviral activity of type III IFN in these cells.

Summary: Background: In care of people living with human immunodeficiency virus (HIV), early diagnosis of infection is one of the greatest challenges remaining. A promising approach to increase early diagnosis could be optimized HIV testing in persons with indicator conditions (ICs). ICs are conditions which are AIDS-defining in people living with HIV, conditions that may have significant adverse consequences for the individual's clinical management if the presence of HIV infection is not detected, and conditions with an (undiagnosed) HIV prevalence of ≥0.1%. Methods: In this cohort study, anonymous routine healthcare data of German statutory health insurances from 07/01/2016 to 06/30/2021 based on insured persons with an ICD-10-based diagnosis of selected ICs were analyzed. In a primary analysis, two stratifications (gender and age), and four sensitivity analyses HIV prevalence/incidence were calculated for persons with at least one of 26 IC described in international literature. This study is registered in the German Clinical Trials Register (identifier: DRKS00028743). Findings: Routine healthcare data from 513,509 insured persons were selected for analysis. In the primary analysis, only in malignant neoplasm of bronchus and lung a HIV prevalence was observed with a 95%-CI < 0.1%. ICs with particularly high HIV prevalence were pneumocystosis (40.33%), oral hairy leukoplakia (36.71%), and Kaposi's sarcoma (29.86%). When stratified by gender, it was observed that in female patients, the 95%-CI of HIV prevalence fell below 0.1% for seven ICs. No such effect was observed in male patients. Stratified by age, among patients aged 30 to

Summary: When conversing, humans instantaneously predict meaning from fragmentary and ambiguous mspeech, long before utterance completion. They do this by integrating priors (initial assumptions about the world) with contextual evidence to rapidly decide on the most likely meaning. One powerful prior is attentional preference for agents, which biases sentence processing but universally so only if agents are animate. Here, we investigate the evolutionary origins of this preference, by allowing chimpanzees, gorillas, orangutans, human children, and adults to freely choose between agents and patients in still images, following video clips depicting their dyadic interaction. All participants preferred animate (and occasionally inanimate) agents, although the effect was attenuated if patients were also animate. The findings suggest that a preference for animate agents evolved before language and is not reducible to simple perceptual biases. To conclude, both humans and great apes prefer animate agents in decision tasks, echoing a universal prior in human language processing.

The NASA Lunar Reconnaissance Orbiter (LRO) has returned petabytes of lunar high spatial resolution surface imagery over the past decade, impractical for humans to fully review manually. Here, we develop an automated method using a deep generative visual model that rapidly retrieves scientifically interesting examples of LRO surface imagery representing the first planetary image anomaly detector. We give quantitative experimental evidence that our method preferentially retrieves anomalous samples such as notable geological features and known human landing and spacecraft crash sites. Our method addresses a major capability gap in planetary science and presents a novel way to unlock insights hidden in ever-increasing remote-sensing data archives, with numerous applications to other science domains.

Abstract Background Long-term care facilities (LTCFs) are vulnerable to disease outbreaks. Here, we jointly analyze SARS-CoV-2 genomic and paired epidemiologic data from LTCFs and surrounding communities in Washington state (WA) to assess transmission patterns during 2020–2022, in a setting of changing policy. We describe sequencing efforts and genomic epidemiologic findings across LTCFs and perform in-depth analysis in a single county. Methods We assessed genomic data representativeness, built phylogenetic trees, and conducted discrete trait analysis to estimate introduction sizes over time, and explored selected outbreaks to further characterize transmission events. Results We found that transmission dynamics among cases associated with LTCFs in WA changed over the course of the COVID-19 pandemic, with variable introduction rates into LTCFs, but decreasing amplification within LTCFs. SARS-CoV-2 lineages circulating in LTCFs were similar to those circulating in communities at the same time. Transmission between staff and residents was bi-directional. Conclusions Understanding transmission dynamics within and between LTCFs using genomic epidemiology on a broad scale can assist in targeting policies and prevention efforts. Tracking facility-level outbreaks can help differentiate intra-facility outbreaks from high community transmission with repeated introduction events. Based on our study findings, methods for routine tree building and overlay of epidemiologic data for hypothesis generation by public health practitioners are recommended. Discrete trait analysis added valuable insight and can be considered when representative sequencing is performed. Cluster detection tools, especially those that rely on distance thresholds, may be of more limited use given current data capture and timeliness. Importantly, we noted a decrease in data capture from LTCFs over time. Depending on goals for use of genomic data, sentinel surveillance should be increased or targeted surveillance implemented to ensure available data for analysis.