Search

Abstract Deep learning (DL) is one of the fastest-growing topics in materials data science, with rapidly emerging applications spanning atomistic, image-based, spectral, and textual data modalities. DL allows analysis of unstructured data and automated identification of features. The recent development of large materials databases has fueled the application of DL methods in atomistic prediction in particular. In contrast, advances in image and spectral data have largely leveraged synthetic data enabled by high-quality forward models as well as by generative unsupervised DL methods. In this article, we present a high-level overview of deep learning methods followed by a detailed discussion of recent developments of deep learning in atomistic simulation, materials imaging, spectral analysis, and natural language processing. For each modality we discuss applications involving both theoretical and experimental data, typical modeling approaches with their strengths and limitations, and relevant publicly available software and datasets. We conclude the review with a discussion of recent cross-cutting work related to uncertainty quantification in this field and a brief perspective on limitations, challenges, and potential growth areas for DL methods in materials science.

Abstract Background Management of type 1 diabetes (T1D) requires the use of insulin, which can cause hypoglycaemia (low blood glucose levels). While most hypoglycaemic episodes can be self-treated, all episodes can be sudden, inconvenient, challenging to prevent or manage, unpleasant and/or cause unwanted attention or embarrassment. Severe hypoglycaemic episodes, requiring assistance from others for recovery, are rare but potentially dangerous. Repeated exposure to hypoglycaemia can reduce classic warning symptoms (‘awareness’), thereby increasing risk of severe episodes. Thus, fear of hypoglycaemia is common among adults with T1D and can have a negative impact on how they manage their diabetes, as well as on daily functioning, well-being and quality of life. While advances in glycaemic technologies and group-based psycho-educational programmes can reduce fear, frequency and impact of hypoglycaemia, they are not universally or freely available, nor do they fully resolve problematic hypoglycaemia or associated worries. This study aims to determine the effectiveness of a fully online, self-directed, scalable, psycho-educational intervention for reducing fear of hypoglycaemia: the Hypoglycaemia Prevention, Awareness of Symptoms, and Treatment (HypoPAST) programme. Methods A 24-week, two-arm, parallel-group, hybrid type 1 randomised controlled trial, conducted remotely (online and telephone). Australian adults (≥ 18 years) with self-reported T1D and fear of hypoglycaemia will be recruited, and allocated at random (1:1) to HypoPAST or control (usual care). The primary outcome is the between-group difference in fear of hypoglycaemia (assessed using HFS-II Worry score) at 24 weeks. A sample size of N = 196 is required to detect a 9-point difference, with 90% power and allowing for 30% attrition. Multiple secondary outcomes include self-reported psychological, behavioural, biomedical, health economic, and process evaluation data. Data will be collected at baseline, 12 and 24 weeks using online surveys, 2-week ecological momentary assessments, website analytics and semi-structured interviews. Discussion This study will provide evidence regarding the effectiveness, cost-effectiveness and acceptability of a novel, online psycho-educational programme: HypoPAST. Due to the fully online format, HypoPAST is expected to provide an inexpensive, convenient, accessible and scalable solution for reducing fear of hypoglycaemia among adults with T1D. Trial registration Australian and New Zealand Clinical Trials Registry (ANZCTR): ACTRN12623000894695 (21 August 2023).

Abstract Background Diabetes-specific quality of life (QoL) questionnaires are commonly used to assess the impact of diabetes and its management on an individual’s quality of life. While several valid and reliable measures of diabetes-specific QoL exist, there is no consensus on which to use and in what setting. Furthermore, there is limited evidence of their acceptability to people with diabetes. Our aim was to explore perceptions of adults with type 1 diabetes (T1D) toward five diabetes-specific QoL measures. Methods Adults (aged 18 + years) with T1D living in Australia or the United Kingdom (UK) were eligible to take part in ‘YourSAY: QoL’, an online cross-sectional survey. Recruitment involved study promotion on diabetes-related websites and social media, as well as direct invitation of people with T1D via a hospital client list (UK only). In random order, participants completed five diabetes-specific QoL measures: Audit of Diabetes-Dependent Quality of Life (ADDQoL-19); Diabetes Care Profile: Social and Personal Factors subscale (DCP); DAWN Impact of Diabetes Profile (DIDP); Diabetes-Specific Quality of Life Scale: Burden Subscale (DSQoLS); Diabetes Quality of Life Questionnaire (Diabetes QOL-Q). They were invited to provide feedback on each questionnaire in the form of a brief free-text response. Responses were analysed using inductive, thematic template analysis. Results Of the N = 1,946 adults with T1D who completed the survey, 20% (UK: n = 216, Australia: n = 168) provided qualitative responses about ≥ 1 measure. All measures received both positive and negative feedback, across four themes: (1) clarity and ease of completion, e.g., difficulty isolating impact of diabetes, dislike of hypothetical questions, and preference for ‘not applicable’ response options; (2) relevance and comprehensiveness, e.g., inclusion of a wide range of aspects of life to improve personal relevance; (3) length and repetition, e.g., length to be balanced against respondent burden; (4) framing and tone, e.g., preference for respectful language and avoidance of extremes. Conclusions These findings suggest opportunities to improve the relevance and acceptability of existing diabetes-specific QoL measures, and offer considerations for developing new measures, which need to be better informed by the preferences of people living with diabetes.

Abstract Background Clinical practice guidelines (CPGs) synthesize high-quality information to support evidence-based clinical practice. In primary care, numerous CPGs must be integrated to address the needs of patients with multiple risks and conditions. The BETTER program aims to improve prevention and screening for cancer and chronic disease in primary care by synthesizing CPGs into integrated, actionable recommendations. We describe the process used to harmonize high-quality cancer and chronic disease prevention and screening (CCDPS) CPGs to update the BETTER program. Methods A review of CPG databases, repositories, and grey literature was conducted to identify international and Canadian (national and provincial) CPGs for CCDPS in adults 40–69 years of age across 19 topic areas: cancers, cardiovascular disease, chronic obstructive pulmonary disease, diabetes, hepatitis C, obesity, osteoporosis, depression, and associated risk factors (i.e., diet, physical activity, alcohol, cannabis, drug, tobacco, and vaping/e-cigarette use). CPGs published in English between 2016 and 2021, applicable to adults, and containing CCDPS recommendations were included. Guideline quality was assessed using the Appraisal of Guidelines for Research and Evaluation (AGREE) II tool and a three-step process involving patients, health policy, content experts, primary care providers, and researchers was used to identify and synthesize recommendations. Results We identified 51 international and Canadian CPGs and 22 guidelines developed by provincial organizations that provided relevant CCDPS recommendations. Clinical recommendations were extracted and reviewed for inclusion using the following criteria: 1) pertinence to primary prevention and screening, 2) relevance to adults ages 40–69, and 3) applicability to diverse primary care settings. Recommendations were synthesized and integrated into the BETTER toolkit alongside resources to support shared decision-making and care paths for the BETTER program. Conclusions Comprehensive care requires the ability to address a person’s overall health. An approach to identify high-quality clinical guidance to comprehensively address CCDPS is described. The process used to synthesize and harmonize implementable clinical recommendations may be useful to others wanting to integrate evidence across broad content areas to provide comprehensive care. The BETTER toolkit provides resources that clearly and succinctly present a breadth of clinical evidence that providers can use to assist with implementing CCDPS guidance in primary care.

Abstract To address our climate emergency, “we must rapidly, radically reshape society”—Johnson & Wilkinson, All We Can Save. In science, reshaping requires formidable technical (cloud, coding, reproducibility) and cultural shifts (mindsets, hybrid collaboration, inclusion). We are a group of cross‐government and academic scientists that are exploring better ways of working and not being too entrenched in our bureaucracies to do better science, support colleagues, and change the culture at our organizations. We share much‐needed success stories and action for what we can all do to reshape science as part of the Open Science movement and 2023 Year of Open Science.

Summary: AML is characterized by mutations in genes associated with growth regulation such as internal tandem duplications (ITD) in the receptor kinase FLT3. Inhibitors targeting FLT3 (FLT3i) are being used to treat patients with FLT3-ITD+ but most relapse and become resistant. To elucidate the resistance mechanism, we compared the gene regulatory networks (GRNs) of leukemic cells from patients before and after relapse, which revealed that the GRNs of drug-responsive patients were altered by rewiring their AP-1-RUNX1 axis. Moreover, FLT3i induces the upregulation of signaling genes, and we show that multiple cytokines, including interleukin-3 (IL-3), can overcome FLT3 inhibition and send cells back into cycle. FLT3i leads to loss of AP-1 and RUNX1 chromatin binding, which is counteracted by IL-3. However, cytokine-mediated drug resistance can be overcome by a pan-RAS inhibitor. We show that cytokines instruct AML growth via the transcriptional regulators AP-1 and RUNX1 and that pan-RAS drugs bypass this barrier.

Summary: Acute myeloid leukemia (AML) is a heterogeneous disease caused by different mutations. Previously, we showed that each mutational subtype develops its specific gene regulatory network (GRN) with transcription factors interacting within multiple gene modules, many of which are transcription factor genes themselves. Here, we hypothesize that highly connected nodes within such networks comprise crucial regulators of AML maintenance. We test this hypothesis using FLT3-ITD-mutated AML as a model and conduct an shRNA drop-out screen informed by this analysis. We show that AML-specific GRNs predict crucial regulatory modules required for AML growth. Furthermore, our work shows that all modules are highly connected and regulate each other. The careful multi-omic analysis of the role of one (RUNX1) module by shRNA and chemical inhibition shows that this transcription factor and its target genes stabilize the GRN of FLT3-ITD+ AML and that its removal leads to GRN collapse and cell death.

BackgroundSelf-monitoring of blood glucose (SMBG) is considered of little clinical benefit for adults with non-insulin-treated type 2 diabetes, but no comprehensive review of a structured approach to SMBG has been published to date.PurposeTo conduct a systematic review and meta-analysis of the impact of sSMBG on HbA1c, treatment modifications, behavioral and psychosocial outcomes, and; examine the moderating effects of sSMBG protocol characteristics on HbA1c.Data sourcesFour databases searched (November 2020; updated: February 2022).Study selectionInclusion criteria: non-randomized and randomized controlled trials (RCTs) and prospective observational studies; reporting effect of sSMBG on stated outcomes; among adults (≥18 years) with non-insulin-treated type 2 diabetes. Studies excluded if involving children or people with insulin-treated or other forms of diabetes.Data extraction and analysisOutcome data extracted, and risk of bias/quality assessed independently by two researchers. Meta-analysis was conducted for RCTs, and moderators explored (HbA1c only).Data synthesisFrom 2,078 abstracts, k=23 studies were included (N=5,372). Risk of bias was evident and study quality was low. Outcomes assessed included: HbA1c (k=23), treatment modification (k=16), psychosocial/behavioral outcomes (k=12). Meta-analysis revealed a significant mean difference favoring sSMBG in HbA1c (-0·29%, 95% CI: -0·46 to -0·11, k=13) and diabetes self-efficacy (0.17%, 95% CI: 0.01 to 0.33, k=2). Meta-analysis revealed no significant moderating effects by protocol characteristics.LimitationsFindings limited by heterogeneity in study designs, intervention characteristics, and psychosocial assessments.ConclusionA small positive effect of sSMBG on HbA1c and diabetes self-efficacy was observed. Narrative synthesis of sSMBG intervention characteristics may guide future implementation.PROSPERO registrationhttps://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42020208857, identifier CRD42020208857.

Abstract Introduction Point-of-care ultrasound (US) is used in clinical practice across many specialties. Ultrasound (US) curricula for medical students are increasingly common. Optimal timing, structure, and effect of ultrasound education during medical school remains poorly understood. This study aims to retrospectively determine the association between participation in a preclinical, longitudinal US curriculum and medical student academic performance. Methods All first-year medical students at a medical school in the Midwest region of the United States were offered a voluntary longitudinal US curriculum. Participants were selected by random lottery. The curriculum consisted of five three-hour hands on-sessions with matching asynchronous content covering anatomy and pathologic findings. Content was paired with organ system blocks in the standard first year curriculum at our medical school. Exam scores between the participating and non-participating students were compared to evaluate the objective impact of US education on performance in an existing curriculum. We hypothesized that there would be an association between participation in the curriculum and improved medical student performance. Secondary outcomes included shelf exam scores for the surgery, internal medicine, neurology clerkships and USMLE Step 1. A multivariable linear regression model was used to evaluate the association of US curriculum participation with student performance. Scores were adjusted for age, gender, MCAT percentile, and science or engineering degree. Results 76 of 178 students applied to participate in the curriculum, of which 51 were accepted. US curriculum students were compared to non-participating students (n = 127) from the same class. The US curriculum students performed better in cardiovascular anatomy (mean score 92.1 vs. 88.7, p = 0.048 after adjustment for multiple comparisons). There were no significant differences in cumulative cardiovascular exam scores, or in anatomy and cumulative exam scores for the gastroenterology and neurology blocks. The effect of US curriculum participation on cardiovascular anatomy scores was estimated to be an improvement of 3.48 points (95% CI 0.78-6.18). No significant differences were observed for USMLE Step 1 or clerkship shelf exams. There were no significant differences in either preclinical, clerkship or Step 1 score for the 25 students who applied and were not accepted and the 102 who did not apply. Conclusions Participation in a preclinical longitudinal US curriculum was associated with improved exam performance in cardiovascular anatomy but not examination of other cardiovascular system concepts. Neither anatomy or comprehensive exam scores for neurology and gastrointestinal organ system blocks were improved.

Abstract Background The African continent currently experiences 25% of the global burden of disease with only 1.3% of the world’s healthcare workers. The COVID-19 pandemic has caused unprecedented disruption to medical education systems, increasing the strain on already-vulnerable regions. Our study examines the impact of COVID-19 on medical students across 33 countries in the African continent. Methods A 39-item anonymous electronic survey was developed and distributed to medical students across Africa through social networks to assess the impact of the COVID-19 pandemic on medical education. The survey assessed the domains of: class structure changes and timing, patient interactions, exam administration, learning environment satisfaction, mental health impacts, and volunteer opportunities/engagement. Results 694 students across 33 countries participated. 80% of respondents had their classes suspended for varied lengths of time during the pandemic, and from these students 59% of them resumed their classes. 83% of students felt they were in a supportive learning environment before the pandemic, which dropped to 32% since the start. The proportion of students taking exams online increased (6–26%, p

AimsManaging weight in the context of type 2 diabetes presents unique hormonal, medicinal, behavioural and psychological challenges. The relationship between weight management and personality has previously been reviewed for general and cardiovascular disease populations but is less well understood in diabetes. This systematic review investigated the relationship between personality constructs and weight management outcomes and behaviours among adults with type 2 diabetes.MethodsMedline, PubMed, Embase, PsycINFO and SPORTDiscus databases were searched to July 2021. Eligibility: empirical quantitative studies; English language; adults with type 2 diabetes; investigation of personality-weight management association. Search terms included variants of: diabetes, physical activity, diet, body mass index (BMI), adiposity, personality constructs and validated scales. A narrative synthesis, with quality assessment, was conducted.ResultsSeventeen studies were identified: nine cross-sectional, six cohort and two randomised controlled trials (N=6,672 participants, range: 30-1,553). Three studies had a low risk of bias. Personality measurement varied. The Big Five and Type D personality constructs were the most common measures. Higher emotional instability (neuroticism, negative affect, anxiety, unmitigated communion and external locus of control) was negatively associated with healthy diet and physical activity, and positively associated with BMI. Conscientiousness had positive associations with healthy diet and physical activity and negative associations with BMI and anthropometric indices.ConclusionsAmong adults with type 2 diabetes, evidence exists of a relationship between weight management and personality, specifically, negative emotionality and conscientiousness. Consideration of personality may be important for optimising weight management and further research is warranted.Systematic review registrationwww.crd.york.ac.uk/prospero/, identifier CRD42019111002.

Objective To develop a theory and evidence-based web intervention to reduce psychological barriers towards insulin therapy among adults with non-insulin-treated type 2 diabetes (T2D).Methods Salient psychological barriers towards insulin were identified from the literature and classified using the Theoretical Domains Framework (TDF). Relevant TDF domains were mapped to evidence-based behaviour change techniques (BCTs), which informed the content for each barrier. Acceptability was explored using cognitive debriefing interviews (n=6 adults with T2D).Results ‘Is Insulin Right for Me’ addresses eight barriers, phrased as common questions: Does insulin mean my diabetes is more serious? Do insulin injections cause complications? Is it my fault I need to inject insulin? Will I gain weight? Will injecting hurt? What about hypos? Will injecting insulin be a burden? What will others think of me? BCTs, including persuasive communication and modelling/demonstration, were delivered using appropriate methods (eg, demonstration of the injection process). Participant suggestions for improvement included clear and direct messages, normalising insulin and avoiding confronting images.Conclusions ‘Is Insulin Right for Me’ is the first theory and evidence-based, web intervention designed to reduce psychological barriers towards insulin therapy for adults with T2D. Evaluation is needed to determine its impact on negative appraisals and receptiveness towards insulin.

IntroductionBleeding is a concern after percutaneous coronary intervention (PCI) and subsequent dual antiplatelet therapy (DAPT). We herein report the incidence and risk factors for major bleeding in the Norwegian Coronary Stent Trial (NORSTENT).Materials and methodsNORSTENT was a randomized, double blind, pragmatic trial among patients with acute coronary syndrome or stable coronary disease undergoing PCI during 2008-11. The patients (N = 9,013) were randomized to receive either a drug-eluting stent or a bare-metal stent, and were treated with at least nine months of DAPT. The patients were followed for a median of five years, with Bleeding Academic Research Consortium (BARC) 3-5 major bleeding as one of the safety endpoints. We estimated cumulative incidence of major bleeding by a competing risks model and risk factors through cause-specific Cox models.ResultsThe 12-month cumulative incidence of major bleeding was 2.3%. Independent risk factors for major bleeding were chronic kidney disease, low bodyweight (< 60 kilograms), diabetes mellitus, and advanced age (> 80 years). A myocardial infarction (MI) or PCI during follow-up increased the risk of major bleeding (HR = 1.67, 95% CI 1-29-2.15).ConclusionsThe 12-month cumulative incidence of major bleeding in NORSTENT was higher than reported in previous, explanatory trials. This analysis strengthens the role of chronic kidney disease, advanced age, and low bodyweight as risk factors for major bleeding among patients receiving DAPT after PCI. The presence of diabetes mellitus or recurrent MI among patients is furthermore a signal of increased bleeding risk.Clinical trial registrationUnique identifier NCT00811772; http://www.clinicaltrial.gov.

Abstract Background General Practice Optimising Structured Monitoring to Improve Clinical Outcomes in Type 2 Diabetes (GP-OSMOTIC) is a multicentre, individually randomised controlled trial aiming to compare the use of intermittent retrospective continuous glucose monitoring (r-CGM) to usual care in patients with type 2 diabetes attending general practice. The study protocol was published in the British Medical Journal Open and described the principal features of the statistical methods that will be used to analyse the trial data. This paper provides greater detail on the statistical analysis plan, including background and justification for the statistical methods chosen, in accordance with SPIRIT guidelines. Objective To describe in detail the data management process and statistical methods that will be used to analyse the trial data. Methods An overview of the trial design and primary and secondary research questions are provided. Sample size assumptions and calculations are explained, and randomisation and data management processes are described in detail. The planned statistical analyses for primary and secondary outcomes and sub-group analyses are specified along with the intended table layouts for presentation of the results. Conclusion In accordance with best practice, all analyses outlined in the document are based on the aims of the study and have been pre-specified prior to the completion of data collection and outcome analyses. Trial registration Australian New Zealand Clinical Trials Registry, ACTRN12616001372471. Registered on 3 August 2016.

A Paleoproterozoic carbon isotope anomaly is likely linked to burial of oceanic cyanobacteria, but it is not clear how burial occurred. Here, the authors find that, under Paleoproterozoic pCO2 conditions, planktonic cyanobacteria increase exopolysaccharide production and mineralization, leading to aggregation and faster sinking.

Summary: Neoantigen-specific T cells are increasingly viewed as important immunotherapy effectors, but physically isolating these rare cell populations is challenging. Here, we describe a sensitive method for the enumeration and isolation of neoantigen-specific CD8+ T cells from small samples of patient tumor or blood. The method relies on magnetic nanoparticles that present neoantigen-loaded major histocompatibility complex (MHC) tetramers at high avidity by barcoded DNA linkers. The magnetic particles provide a convenient handle to isolate the desired cell populations, and the barcoded DNA enables multiplexed analysis. The method exhibits superior recovery of antigen-specific T cell populations relative to literature approaches. We applied the method to profile neoantigen-specific T cell populations in the tumor and blood of patients with metastatic melanoma over the course of anti-PD1 checkpoint inhibitor therapy. We show that the method has value for monitoring clinical responses to cancer immunotherapy and might help guide the development of personalized mutational neoantigen-specific T cell therapies and cancer vaccines. : Peng et al. report a sensitive method to detect tumor-associated neoantigen-specific T cells. Neoantigens and fluorescent DNA barcodes, presented on nanoparticle scaffolds, permit multiplex capture and analysis of specific T cell populations from blood or tumor. Neoantigen-specific T cell numbers track tumor volume in a melanoma patient responding to immunotherapy. Keywords: neoantigens, cancer immunotherapy, nanotechnology, microfluidics, T cell receptor

Objective In a sample of adults with type 2 diabetes mellitus (T2DM), the aim of this study was to examine whether self-reported physical activity level is associated with recall of specific physical activity-related interactions used by general practitioners (GP).Research design and methods Adults with T2DM completed an online survey reporting physical activity behaviors and recall of 14 GP–patient interactions about physical activity, mapped onto discrete behavior change techniques (BCT). Stepped logistical regression examined associations between recommended physical activity (≥600 MET-min/week) and GP–patient interactions, controlling for body mass index, diabetes-related comorbidities, depressive symptoms and self-efficacy.Results In total, 381 respondents (55% men, mean±SD age: 62±10 years and T2DM duration 8±8 years) provided complete data. Most (73%) reported receiving ‘general advice’, while interactions related to goal setting, monitoring, and relapse prevention were least commonly reported (all

Introduction Medical student mistreatment is a prevalent and significant challenge for medical schools across the country, associated with negative emotional and professional consequences for students. The Association of American Medical Colleges and Liaison Committee on Medical Education have increasingly emphasized the issue of mistreatment in recent years, and medical schools are tasked with creating a positive learning climate. Methods The authors describe the efforts of an emergency department (ED) to improve its clerkship learning environment, using a multifaceted approach for collecting mistreatment data and relaying them to educators and clerkship leadership. Data are gathered through end-of-rotation evaluations, teaching evaluations, and an online reporting system available to medical students. Mistreatment data are then relayed to the ED during semi-annual meetings between clerkship leadership and medical school assistant deans, and through annual mistreatment reports provided to department chairs. Results Over a two-year period, students submitted a total of 56 narrative comments related to mistreatment or unprofessional behavior during their emergency medicine (EM) clerkship. Of these comments, 12 were submitted in 2015–16 and 44 were submitted in 2016–17. The most frequently observed themes were students feeling ignored or marginalized by faculty (14 comments); students being prevented from speaking or working with patients and/or attending faculty (11 comments); and students being treated in an unprofessional manner by staff (other than faculty, 8 comments). Conclusion This article details an ED’s efforts to improve its EM clerkship learning environment by tracking mistreatment data and intentionally communicating the results to educators and clerkship leadership. Continued mistreatment data collection and faculty development will be necessary for these efforts to have a measurable effect on the learning environment.

Neutropenic fever (NF) is a life-threatening complication of myelosuppressive chemotherapy in patients with hematologic malignancies and triggers the administration of broad-spectrum antimicrobials. The ability to accurately predict NF would permit initiation of antimicrobials earlier in the course of infection with the goal of decreasing morbid complications and progression to septic shock and death. Changes in the blood level of inflammatory biomarkers may precede the occurrence of NF. To identify potential biomarkers for the prediction of NF, we performed serial meas- urements of nine biomarkers [C-reactive protein (CRP), protein C, interleukin (IL)-6, IL-8, IL-10, IL-1β, tumor necrosis factor-α, monocyte chemotactic protein-1, and intercellular adhesion molecule-1] using a multiplex ELISA array platform every 6-8 hours in patients undergoing myelosuppressive chemotherapy for hematologic malignancies. We found that the blood levels of IL-6 and CRP increased significantly 24 to 48 hours prior to the onset of fever. In addition, we showed that frequent biomarker monitoring is feasible using a bedside micro sample test device. The results of this pilot study suggest that serial monitoring of IL-6 and CRP levels using a bedside device may be useful in the prediction of NF. Prospective studies involving a larger cohort of patients to validate this observation are warranted. This trial is registered at ClinicalTrials.gov (NCT01144793).