14 research outputs found
Author Correction: Automation of surgical skill assessment using a three-stage machine learning algorithm.
No full textTransactions across serverless functions leveraging stateful dataflows
No full textServerless computing is currently the fastest-growing cloud services segment. The most prominent serverless offering is Function-as-a-Service (FaaS), where users write functions and the cloud automates deployment, maintenance, and scalability. Although FaaS is a good fit for executing stateless functions, it does not adequately support stateful constructs like microservices and scalable, low-latency cloud applications. Recently, there have been multiple attempts to add first-class support for state in FaaS systems, such as Microsoft Orleans, Azure Durable Functions, or Beldi. These approaches execute business code inside stateless functions, handing over their state to an external database. In contrast, approaches such as Apache Flink's StateFun follow a different design: a dataflow system such as Apache Flink handles all state management, messaging, and checkpointing by executing a stateful dataflow graph providing exactly-once state processing guarantees. This design relieves programmers from having to “pollute” their business logic with distributed systems error checking, management, and mitigation. Although programmers can easily develop applications without worrying about messaging and state management, executing transactions across stateful functions remains an open problem. In this paper, we introduce a programming model and implementation for transaction orchestration of stateful serverless functions. Our programming model supports serializable distributed transactions with two-phase commit, as well as eventually consistent workflows with Sagas. We design and implement our programming model on Apache Flink StateFun to leverage Flink's exactly-once processing and state management guarantees. Our experiments show that the approach of building transactional systems on top of dataflow graphs can achieve very high throughput, but with latency overhead due to checkpointing mechanism guaranteeing the exactly-once processing. We compare our approach to Beldi that implements two-phase commit on AWS lambda functions backed by DynamoDB for state management, as well as an implementation of a system that makes use of CockroachDB as its backend.Web Information System
Author Correction: Uncoupling of invasive bacterial mucosal immunogenicity from pathogenicity.
No full textISSN:2041-172
ETAT DE LA RECHERCHE ANGLOPHONE ET FRANCOPHONE EN SYSTEMES D'INFORMATION COMPTABLES SUR LA PERIODE 1990 - 2007
No full textInternational audienceThis paper aims to perform a critical analysis of the academic production, from 1990 to 2007, within AIS field and to propose future research perspectives. This study is based upon the record and content analysis of published papers within specialized Anglo-Saxons journals, and French journals and main academic conferences. The results are classified into three categories. The first one identifies main vectors for research dissemination, particularly the major role of IJAIS and JIS. The second one uncover AIS research characteristics, such as acceptation of varied methods, themes evolution, and no domination by any author, university, or paradigm. The last one highlights similarities and divergences between French and Anglo-Saxon researches, especially a so faster cadence of the Anglo-Saxon production. We conclude that AIS research raise to the rank of distinct research field.Cet article procède à une analyse critique de la production académique, de 1990 à 2007, dans le domaine des systèmes d'information comptable (SIC) et propose des voies futures à explorer. L'étude est fondée sur le recensement et l'analyse de contenu des articles publiés dans les revues françaises et anglo-saxonnes ainsi que dans les principales conférences académiques en comptabilité et systèmes d'information en France. Les résultats de recherche peuvent être classés en trois catégories. La première identifie des vecteurs majeurs de la dissémination de la recherche, en particulier le rôle majeur des revues anglo-saxonnes comme l'IJAIS et le JIS. La deuxième révèle les caractéristiques de la recherche en SIC, à savoir l'acceptation de méthodes variées, l'évolution des thématiques, l'absence d'une domination par un auteur ou un courant de pensée ou une université. La dernière met en lumière les différences et les similitudes entre la recherche anglophone et francophone dans ce domaine de recherche, en particulier une cadence nettement plus soutenue de la production scientifique anglo-saxonne. Nous concluons que les travaux académiques en SIC semblent acquérir le statut d'un domaine de recherche distinct
Mouse models for bacterial enteropathogen infections: insights into the role of colonization resistance
No full textABSTRACTGlobally, enteropathogenic bacteria are a major cause of morbidity and mortality.1-3 Campylobacter, Salmonella, Shiga-toxin-producing Escherichia coli, and Listeria are among the top five most commonly reported zoonotic pathogens in the European Union.4 However, not all individuals naturally exposed to enteropathogens go on to develop disease. This protection is attributable to colonization resistance (CR) conferred by the gut microbiota, as well as an array of physical, chemical, and immunological barriers that limit infection. Despite their importance for human health, a detailed understanding of gastrointestinal barriers to infection is lacking, and further research is required to investigate the mechanisms that underpin inter-individual differences in resistance to gastrointestinal infection. Here, we discuss the current mouse models available to study infections by non-typhoidal Salmonella strains, Citrobacter rodentium (as a model for enteropathogenic and enterohemorrhagic E. coli), Listeria monocytogenes, and Campylobacter jejuni. Clostridioides difficile is included as another important cause of enteric disease in which resistance is dependent upon CR. We outline which parameters of human infection are recapitulated in these mouse models, including the impact of CR, disease pathology, disease progression, and mucosal immune response. This will showcase common virulence strategies, highlight mechanistic differences, and help researchers from microbiology, infectiology, microbiome research, and mucosal immunology to select the optimal mouse model
Decoupling bile acid 7α-dehydroxylation from colonization resistance to <i>Clostridioides difficile</i>
No full textSecondary bile acids, generated through microbial transformation of primary bile acids secreted in bile, play a role in shaping intestinal microbial communities, modulating host immunity, and regulating energy metabolism. In vitro studies have shown that the balance between primary and secondary bile acids strongly affects spore germination, growth, and cellular physiology of Clostridioides difficile , a major nosocomial gut pathogen. In vivo correlations between microbiome composition, bile acid metabolome, and colonization resistance have led to the hypothesis that 7α-dehydroxylating bacteria such as Clostridium scindens protect against C. difficile infection by producing secondary bile acids like deoxycholic acid. However, due to the genetic intractability of known 7α-dehydroxylating species, direct experimental validation of this causal relationship has been challenging. In this study, we leveraged the first available 7α-dehydroxylation-deficient baiH mutant to test the direct role of 7-dehydroxylated bile acid production in C. difficile colonization resistance in vivo. We colonized gnotobiotic mice with isogenic wild-type or baiH strains of the recently described 7α-dehydroxylating species Faecalicatena contorta , including wild-type C. scindens -colonized mice as a positive control. Wild-type F. contorta accumulated 7-dehydroxylated bile acids at levels equivalent to C. scindens , in a strictly baiH -dependent manner. However, despite equivalent bile acid profiles, wild-type F. contorta failed to replicate the C. difficile -restrictive phenotype observed with C. scindens . These findings demonstrate that commensal clostridial 7α-dehydroxylation alone is not sufficient for enhancing colonization resistance to C. difficile . Our results highlight the existence of additional, potentially bile acid-independent mechanisms by which certain commensals mediate protection, with important implications for microbiota-based therapies. Importance 7α-dehydroxylated secondary bile acids, including deoxycholic acid and lithocholic acid, produced by commensal clostridia are widely assumed to inhibit the important nosocomial pathogen Clostridioides difficile , yet their precise role in colonization resistance remains unresolved. Using a defined mouse microbiota and an isogenic Faecalicatena contorta strain pair differing in a single 7α-dehydroxylation gene ( baiH ), we show that restoration of secondary bile acid production is not sufficient to delay C. difficile colonization in vivo. This contrasts with the protective effect of Clostridium scindens , which generates a similar bile acid profile. Our findings uncouple bile acid metabolism from protection and suggest that additional, strain-specific functions – such as nutrient competition or antimicrobial production – play a critical role. Understanding these mechanisms is essential for the rational design of next-generation microbiota-based therapies to prevent or treat recurrent C. difficile infection.EM
Perioperative Fully Closed-Loop Insulin Delivery in Patients Undergoing Elective Surgery: An Open-Label, Randomized Controlled Trial.
No full textOBJECTIVE
Perioperative management of glucose levels remains challenging. We aimed to assess whether fully closed-loop subcutaneous insulin delivery would improve glycemic control compared with standard insulin therapy in insulin-requiring patients undergoing elective surgery.
RESEARCH DESIGN AND METHODS
We performed a single-center, open-label, randomized controlled trial. Patients with diabetes (other than type 1) undergoing elective surgery were recruited from various surgical units and randomly assigned using a minimization schedule (stratified by HbA1c and daily insulin dose) to fully closed-loop insulin delivery with fast-acting insulin aspart (closed-loop group) or standard insulin therapy according to local clinical practice (control group). Study treatment was administered from hospital admission to discharge (for a maximum of 20 days). The primary end point was the proportion of time with sensor glucose in the target range (5.6-10.0 mmol/L).
RESULTS
Forty-five patients were enrolled and assigned to the closed-loop (n = 23) or the control (n = 22) group. One patient (closed-loop group) withdrew from the study before surgery and was not analyzed. Participants underwent abdominal (57%), vascular (23%), orthopedic (9%), neuro (9%), or thoracic (2%) surgery. The mean proportion of time that sensor glucose was in the target range was 76.7 ± 10.1% in the closed-loop and 54.7 ± 20.8% in the control group (mean difference 22.0 percentage points [95% CI 11.9; 32.0%]; P < 0.001). No episodes of severe hypoglycemia (<3.0 mmol/L) or hyperglycemia with ketonemia or any study-related adverse events occurred in either group.
CONCLUSIONS
In the context of mixed elective surgery, the use of fully closed-loop subcutaneous insulin delivery improves glucose control without a higher risk of hypoglycemia
Patient and Context Factors in the Adoption of Active Surveillance for Low-Risk Prostate Cancer
No full textImportance Although active surveillance for patients with low-risk prostate cancer (LRPC) has been recommended for years, its adoption at the population level is often limited.Objective To make active surveillance available for patients with LRPC using a research framework and to compare patient characteristics and clinical outcomes between those who receive active surveillance vs radical treatments at diagnosis.Design, Setting, and Participants This population-based, prospective cohort study was designed by a large multidisciplinary group of specialists and patients' representatives. The study was conducted within all 18 urology centers and 7 radiation oncology centers in the Piemonte and Valle d'Aosta Regional Oncology Network in Northwest Italy (approximate population, 4.5 million). Participants included patients with a new diagnosis of LRPC from June 2015 to December 2021. Data were analyzed from January to May 2023.Exposure At diagnosis, all patients were informed of the available treatment options by the urologist and received an information leaflet describing the benefits and risks of active surveillance compared with active treatments, either radical prostatectomy (RP) or radiation treatment (RT). Patients choosing active surveillance were actively monitored with regular prostate-specific antigen testing, clinical examinations, and a rebiopsy at 12 months.Main Outcomes and Measures Outcomes of interest were proportion of patients choosing active surveillance or radical treatments, overall survival, and, for patients in active surveillance, treatment-free survival. Comparisons were analyzed with multivariable logistic or Cox models, considering centers as clusters.Results A total of 852 male patients (median [IQR] age, 70 [64-74] years) were included, and 706 patients (82.9%) chose active surveillance, with an increasing trend over time; 109 patients (12.8%) chose RP, and 37 patients (4.3%) chose RT. Median (IQR) follow-up was 57 (41-76) months. Worse prostate cancer prognostic factors were negatively associated with choosing active surveillance (eg, stage T2a vs T1c: odds ratio [OR], 0.51; 95% CI, 0.28-0.93), while patients who were older (eg, age >= 75 vs = 2 vs 0: OR, 1.98; 95% CI, 1.02-3.85), underwent an independent revision of the first prostate biopsy (OR, 2.35; 95% CI, 1.26-4.38) or underwent a multidisciplinary assessment (OR, 2.65; 95% CI, 1.38-5.11) were more likely to choose active surveillance vs active treatment. After adjustment, center at which a patient was treated continued to be an important factor in the choice of treatment (intraclass correlation coefficient, 18.6%). No differences were detected in overall survival between active treatment and active surveillance. Treatment-free survival in the active surveillance cohort was 59.0% (95% CI, 54.8%-62.9%) at 24 months, 54.5% (95% CI, 50.2%-58.6%) at 36 months, and 47.0% (95% CI, 42.2%-51.7%) at 48 months.Conclusions and Relevance In this population-based cohort study of patients with LRPC, a research framework at system level as well as favorable prognostic factors, a multidisciplinary approach, and an independent review of the first prostate biopsy at patient-level were positively associated with high uptake of active surveillance, a practice largely underused before this study
