Étude cas-témoins | EPFL Graph Search

Une étude cas-témoins est une étude statistique observationnelle rétrospective utilisée en épidémiologie. Les études cas-témoins sont utilisées pour mettre en évidence des facteurs qui peuvent contribuer à l'apparition d'une maladie en comparant des sujets qui ont cette maladie (les cas) avec des sujets qui n'ont pas la maladie mais qui sont similaires par ailleurs (les témoins). Les études cas-témoins posent donc la question : Y a-t-il plus de gens exposés à un facteur (par exemple plus de fumeurs) chez les cas que chez les témoins ? Les études cas-témoins sont des études qui sont relativement peu onéreuses et faciles à mettre en place. L'un des succès les plus significatifs de l'étude cas-témoins a été de démontrer le lien entre le tabagisme et le cancer du poumon. Sir Richard Doll a été capable de montrer une association statistiquement significative entre les deux. Les opposants ont notamment fait valoir, à juste titre, pendant de nombreuses années que ce type d'étude seul ne pe

Quantum-based approach for laboratory risk assessment and decision-making

Ruoxing Liao

With the rising focus on academic safety, there has been an effort to improve the academic safety climate and develop lab-specific risk assessment tools. Despite the progress made in recent years, there is still a deficit of reliable data statistics on safety events and risk assessment in academia, making it difficult to perform full-scale analysis. Moreover, the rapid turnover of personnel and the highly dynamic research tasks in a laboratory environment impede the implementation of classical and well-established risk management strategies, which are usually time-consuming and resource-intensive. Laboratory safety management is always challenging due to lack of data and effective methods. The end of safety management is usually the selection of risk mitigation measures. In the laboratory risk management process, decisions are often made through discussion and compromise. In this regard, it would be beneficial to provide a tool that can support in resource allocation and prioritization of actions. The goal of this project is to develop a decision support tool to assist in the selection of optimal and appropriate preventive or corrective measures in the lab safety management process. The use of decision tools such as Multi-Criteria Decision Making (MCDM) can effectively abstract complex laboratory management problems into a decision framework. This study focuses on the ranking process of competitive corrective measures. Combining the results of interviews and questionnaires, five evaluation criteria were proposed for the normative process of multi-criteria decision making. They are: Residual Risk, Risk reduction, Personnel reliability, Technical reliability, and Cost. classical methods TOPSIS and WSM have been evaluated during the study, and emerging quantum decision making methods have been explored. All three methods are used to simulate the laboratory safety decision making process, and they all present good consistency. Different weighting methods have also been applied to the model and two case studies have been conducted. The first case study is designed to construct the model and verify its feasibility, and the second model is used to investigate the process of group decision making. By applying different MCDM methods, the obtained results are discussed and compared to verify the stability and accuracy of the model. The approach proposed in this research is a promising tool. It offers the possibility to be compatible with different MCDM methods and risk assessment methods, and the proposed decision model could be used as a decision support system to help lab managers make better decisions, or it can provide a neutral point of view as an aid. Future work may include fine-tuning the model. In the following stage, decision criteria should be precisely defined, and corresponding quantitative formulas should be proposed to serve the flexible and dynamic laboratory environment.

2022

Text as signal. A tutorial with case studies focusing on social media (Twitter)

Erick Jorge Canales Rodriguez

Sentiment analysis is the automated coding of emotions expressed in text. Sentiment analysis and other types of analyses focusing on the automatic coding of textual documents are increasingly popular in psychology and computer science. However, the potential of treating automatically coded text collected with regular sampling intervals as a signal is currently overlooked. We use the phrase "text as signal" to refer to the application of signal processing techniques to coded textual documents sampled with regularity. In order to illustrate the potential of treating text as signal, we introduce the reader to a variety of such techniques in a tutorial with two case studies in the realm of social media analysis. First, we apply finite response impulse filtering to emotion-coded tweets posted during the US Election Week of 2020 and discuss the visualization of the resulting variation in the filtered signal. We use changepoint detection to highlight the important changes in the emotional signals. Then we examine data interpolation, analysis of periodicity via the fast Fourier transform (FFT), and FFT filtering to personal value-coded tweets from November 2019 to October 2020 and link the variation in the filtered signal to some of the epoch-defining events occurring during this period. Finally, we use block bootstrapping to estimate the variability/uncertainty in the resulting filtered signals. After working through the tutorial, the readers will understand the basics of signal processing to analyze regularly sampled coded text.

SPRINGER2022

Microphysics and dynamics of snowfall associated with a warm conveyor belt over Korea

Alexis Berne, Nikola Besic, Josué Etienne Gehring, Nicolas Jean François Jullien, Etienne Gabriel Henri Vignon

On 28 February 2018, 57mm of precipitation associated with a warm conveyor belt (WCB) fell within 21 h over South Korea. To investigate how the large-scale circulation influenced the microphysics of this intense precipitation event, we used radar measurements, snowflake photographs and radiosounding data from the International Collaborative Experiments for Pyeongchang 2018 Olympic and Paralympic-Winter Games (ICE-POP 2018). The WCB was identified with trajectories computed with analysis wind fields from the Integrated Forecast System global atmospheric model. The WCB was collocated with a zone of enhanced wind speed of up to 45ms(-1) at 6500ma.s.l., as measured by a radiosonde and a Doppler radar. Supercooled liquid water (SLW) with concentrations exceeding 0.2 g kg(-1) was produced during the rapid ascent within the WCB. During the most intense precipitation period, vertical profiles of polarimetric radar variables show a peak and subsequent decrease in differential reflectivity as aggregation starts. Below the peak in differential reflectivity, the specific differential phase shift continues to increase, indicating early riming of oblate crystals and secondary ice generation. We hypothesise that the SLW produced in the WCB led to intense riming. Moreover, embedded updraughts in the WCB and turbulence at its lower boundary enhanced aggregation by increasing the probability of collisions between particles. This suggests that both aggregation and riming occurred prominently in this WCB. This case study shows how the large-scale atmospheric flow of a WCB provides ideal conditions for rapid precipitation growth involving SLW production, riming and aggregation. Future microphysical studies should also investigate the synoptic conditions to understand how observed processes in clouds are related to large-scale circulation.

2020