Search

Your search keyword '"Jouvet P"' showing total 999 results
Start Over Author "Jouvet P"

Refine your results

999 results on '"Jouvet P"'

1. Empowering Clinicians with Medical Decision Transformers: A Framework for Sepsis Treatment

Author

Rahman, Aamer Abdul, Agarwal, Pranav, Noumeir, Rita, Jouvet, Philippe, Michalski, Vincent, and Kahou, Samira Ebrahimi

Subjects
Computer Science - Machine Learning, Computer Science - Artificial Intelligence
Abstract

Offline reinforcement learning has shown promise for solving tasks in safety-critical settings, such as clinical decision support. Its application, however, has been limited by the lack of interpretability and interactivity for clinicians. To address these challenges, we propose the medical decision transformer (MeDT), a novel and versatile framework based on the goal-conditioned reinforcement learning paradigm for sepsis treatment recommendation. MeDT uses the decision transformer architecture to learn a policy for drug dosage recommendation. During offline training, MeDT utilizes collected treatment trajectories to predict administered treatments for each time step, incorporating known treatment outcomes, target acuity scores, past treatment decisions, and current and past medical states. This analysis enables MeDT to capture complex dependencies among a patient's medical history, treatment decisions, outcomes, and short-term effects on stability. Our proposed conditioning uses acuity scores to address sparse reward issues and to facilitate clinician-model interactions, enhancing decision-making. Following training, MeDT can generate tailored treatment recommendations by conditioning on the desired positive outcome (survival) and user-specified short-term stability improvements. We carry out rigorous experiments on data from the MIMIC-III dataset and use off-policy evaluation to demonstrate that MeDT recommends interventions that outperform or are competitive with existing offline reinforcement learning methods while enabling a more interpretable, personalized and clinician-directed approach.

Published
2024

2. SwinFuSR: an image fusion-inspired model for RGB-guided thermal image super-resolution

Author

Arnold, Cyprien, Jouvet, Philippe, and Seoud, Lama

Subjects
Electrical Engineering and Systems Science - Image and Video Processing, Computer Science - Computer Vision and Pattern Recognition
Abstract

Thermal imaging plays a crucial role in various applications, but the inherent low resolution of commonly available infrared (IR) cameras limits its effectiveness. Conventional super-resolution (SR) methods often struggle with thermal images due to their lack of high-frequency details. Guided SR leverages information from a high-resolution image, typically in the visible spectrum, to enhance the reconstruction of a high-res IR image from the low-res input. Inspired by SwinFusion, we propose SwinFuSR, a guided SR architecture based on Swin transformers. In real world scenarios, however, the guiding modality (e.g. RBG image) may be missing, so we propose a training method that improves the robustness of the model in this case. Our method has few parameters and outperforms state of the art models in terms of Peak Signal to Noise Ratio (PSNR) and Structural SIMilarity (SSIM). In Track 2 of the PBVS 2024 Thermal Image Super-Resolution Challenge, it achieves 3rd place in the PSNR metric. Our code and pretained weights are available at https://github.com/VisionICLab/SwinFuSR., Comment: Accepted at 20th IEEE Workshop on Perception Beyond the Visible Spectrum, CVPR 2024

Published
2024

3. Acute pediatric hyperammonemia: current diagnosis and management strategies

Author

Savy N, Brossier D, Brunel-Guitton C, Ducharme-Crevier L, Du Pont-Thibodeau G, and Jouvet P

Subjects
Pediatrics, Critical Care, hyperammonemia, sodium phenylacetate, sodium benzoate, hemodialysis., Diseases of the digestive system. Gastroenterology, RC799-869
Abstract

Nadia Savy,1 David Brossier,2 Catherine Brunel-Guitton,1 Laurence Ducharme-Crevier,1 Geneviève Du Pont-Thibodeau,1 Philippe Jouvet1 1Department of Pediatrics, CHU Sainte-Justine, University of Montreal, Montreal, QC, Canada; 2Department of Pediatrics, Pediatric Intensive Care Unit, CHU Caen, Caen, France Abstract: Acute hyperammonemia may induce a neurologic impairment leading to an acute life-threatening condition. Coma duration, ammonia peak level, and hyperammonemia duration are the main risk factors of hyperammonemia-related neurologic deficits and death. In children, hyperammonemia is mainly caused by severe liver failure and inborn errors of metabolism. In an acute setting, obtaining reliable plasma ammonia levels can be challenging because of the preanalytical difficulties that need to be addressed carefully. The management of hyperammonemia includes 1) identification of precipitating factors and cerebral edema presence, 2) a decrease in ammonia production by reducing protein intake and reversing catabolism, and 3) ammonia removal with pharmacologic treatment and, in the most severe cases, with extracorporeal therapies. In case of severe coma, transcranial Doppler ultrasound could be the method of choice to noninvasively monitor cerebral blood flow and titrate therapies. Keywords: pediatrics, critical care, hyperammonemia, sodium phenylacetate, sodium benzoate, hemodialysis

Published
2018

4. Label Propagation Techniques for Artifact Detection in Imbalanced Classes using Photoplethysmogram Signals

Author

Macabiau, Clara, Le, Thanh-Dung, Albert, Kevin, Shahriari, Mana, Jouvet, Philippe, and Noumeir, Rita

Subjects
Computer Science - Machine Learning, Electrical Engineering and Systems Science - Signal Processing, 68T02
Abstract

This study aimed to investigate the application of label propagation techniques to propagate labels among photoplethysmogram (PPG) signals, particularly in imbalanced class scenarios and limited data availability scenarios, where clean PPG samples are significantly outnumbered by artifact-contaminated samples. We investigated a dataset comprising PPG recordings from 1571 patients, wherein approximately 82% of the samples were identified as clean, while the remaining 18% were contaminated by artifacts. Our research compares the performance of supervised classifiers, such as conventional classifiers and neural networks (Multi-Layer Perceptron (MLP), Transformers, Fully Convolutional Network (FCN)), with the semi-supervised Label Propagation (LP) algorithm for artifact classification in PPG signals. The results indicate that the LP algorithm achieves a precision of 91%, a recall of 90%, and an F1 score of 90% for the "artifacts" class, showcasing its effectiveness in annotating a medical dataset, even in cases where clean samples are rare. Although the K-Nearest Neighbors (KNN) supervised model demonstrated good results with a precision of 89%, a recall of 95%, and an F1 score of 92%, the semi-supervised algorithm excels in artifact detection. In the case of imbalanced and limited pediatric intensive care environment data, the semi-supervised LP algorithm is promising for artifact detection in PPG signals. The results of this study are important for improving the accuracy of PPG-based health monitoring, particularly in situations in which motion artifacts pose challenges to data interpretation, Comment: Under preparation to submit to IEEE for possible publications

Published
2023
Full Text
View/download PDF

5. Self-supervised learning with diffusion-based multichannel speech enhancement for speaker verification under noisy conditions

Author

Dowerah, Sandipana, Kulkarni, Ajinkya, Serizel, Romain, and Jouvet, Denis

Subjects
Computer Science - Sound, Electrical Engineering and Systems Science - Audio and Speech Processing
Abstract

The paper introduces Diff-Filter, a multichannel speech enhancement approach based on the diffusion probabilistic model, for improving speaker verification performance under noisy and reverberant conditions. It also presents a new two-step training procedure that takes the benefit of self-supervised learning. In the first stage, the Diff-Filter is trained by conducting timedomain speech filtering using a scoring-based diffusion model. In the second stage, the Diff-Filter is jointly optimized with a pre-trained ECAPA-TDNN speaker verification model under a self-supervised learning framework. We present a novel loss based on equal error rate. This loss is used to conduct selfsupervised learning on a dataset that is not labelled in terms of speakers. The proposed approach is evaluated on MultiSV, a multichannel speaker verification dataset, and shows significant improvements in performance under noisy multichannel conditions.

Published
2023

6. Evaluation of Speaker Anonymization on Emotional Speech

Author

Nourtel, Hubert, Champion, Pierre, Jouvet, Denis, Larcher, Anthony, and Tahon, Marie

Subjects
Electrical Engineering and Systems Science - Audio and Speech Processing, Computer Science - Artificial Intelligence, Computer Science - Computation and Language
Abstract

Speech data carries a range of personal information, such as the speaker's identity and emotional state. These attributes can be used for malicious purposes. With the development of virtual assistants, a new generation of privacy threats has emerged. Current studies have addressed the topic of preserving speech privacy. One of them, the VoicePrivacy initiative aims to promote the development of privacy preservation tools for speech technology. The task selected for the VoicePrivacy 2020 Challenge (VPC) is about speaker anonymization. The goal is to hide the source speaker's identity while preserving the linguistic information. The baseline of the VPC makes use of a voice conversion. This paper studies the impact of the speaker anonymization baseline system of the VPC on emotional information present in speech utterances. Evaluation is performed following the VPC rules regarding the attackers' knowledge about the anonymization system. Our results show that the VPC baseline system does not suppress speakers' emotions against informed attackers. When comparing anonymized speech to original speech, the emotion recognition performance is degraded by 15\% relative to IEMOCAP data, similar to the degradation observed for automatic speech recognition used to evaluate the preservation of the linguistic information.

Published
2023
Full Text
View/download PDF

8. Improving Transformer Performance for French Clinical Notes Classification Using Mixture of Experts on a Limited Dataset

Author

Le, Thanh-Dung, Jouvet, Philippe, and Noumeir, Rita

Subjects
Computer Science - Computation and Language, Electrical Engineering and Systems Science - Signal Processing
Abstract

Transformer-based models have shown outstanding results in natural language processing but face challenges in applications like classifying small-scale clinical texts, especially with constrained computational resources. This study presents a customized Mixture of Expert (MoE) Transformer models for classifying small-scale French clinical texts at CHU Sainte-Justine Hospital. The MoE-Transformer addresses the dual challenges of effective training with limited data and low-resource computation suitable for in-house hospital use. Despite the success of biomedical pre-trained models such as CamemBERT-bio, DrBERT, and AliBERT, their high computational demands make them impractical for many clinical settings. Our MoE-Transformer model not only outperforms DistillBERT, CamemBERT, FlauBERT, and Transformer models on the same dataset but also achieves impressive results: an accuracy of 87\%, precision of 87\%, recall of 85\%, and F1-score of 86\%. While the MoE-Transformer does not surpass the performance of biomedical pre-trained BERT models, it can be trained at least 190 times faster, offering a viable alternative for settings with limited data and computational resources. Although the MoE-Transformer addresses challenges of generalization gaps and sharp minima, demonstrating some limitations for efficient and accurate clinical text classification, this model still represents a significant advancement in the field. It is particularly valuable for classifying small French clinical narratives within the privacy and constraints of hospital-based computational resources., Comment: Under the revision

Published
2023

9. Operational Definitions Related to Pediatric Ventilator Liberation

Author

Abu-Sultaneh, Samer, Iyer, Narayan Prabhu, Fernández, Analía, Gaies, Michael, González-Dambrauskas, Sebastián, Hotz, Justin Christian, Kneyber, Martin CJ, López-Fernández, Yolanda M, Rotta, Alexandre T, Werho, David K, Baranwal, Arun Kumar, Blackwood, Bronagh, Craven, Hannah J, Curley, Martha AQ, Essouri, Sandrine, Fioretto, Jose Roberto, Hartmann, Silvia MM, Jouvet, Philippe, Korang, Steven Kwasi, Rafferty, Gerrard F, Ramnarayan, Padmanabhan, Rose, Louise, Tume, Lyvonne N, Whipple, Elizabeth C, Wong, Judith Ju Ming, Emeriaud, Guillaume, Mastropietro, Christopher W, Napolitano, Natalie, Newth, Christopher JL, Khemani, Robinder G, and Network, Pediatric Acute Lung Injury and Sepsis Investigators

Subjects
Lung, Bioengineering, Pediatric, Assistive Technology, Patient Safety, Humans, Child, Respiration, Artificial, Ventilator Weaning, Ventilators, Mechanical, Research Design, Airway Extubation, airway extubation, extubation failure, high-flow nasal cannula, mechanical ventilation, noninvasive ventilation, pediatric ICU, ventilator weaning, Pediatric Acute Lung Injury and Sepsis Investigators (PALISI) Network, Clinical Sciences, Respiratory System
Abstract

BackgroundCommon, operational definitions are crucial to assess interventions and outcomes related to pediatric mechanical ventilation. These definitions can reduce unnecessary variability among research and quality improvement efforts, to ensure findings are generalizable, and can be pooled to establish best practices.Research questionCan we establish operational definitions for key elements related to pediatric ventilator liberation using a combination of detailed literature review and consensus-based approaches?Study design and methodsA panel of 26 international experts in pediatric ventilator liberation, two methodologists, and two librarians conducted systematic reviews on eight topic areas related to pediatric ventilator liberation. Through a series of virtual meetings, we established draft definitions that were voted upon using an anonymous web-based process. Definitions were revised by incorporating extracted data gathered during the systematic review and discussed in another consensus meeting. A second round of voting was conducted to confirm the final definitions.ResultsIn eight topic areas identified by the experts, 16 preliminary definitions were established. Based on initial discussion and the first round of voting, modifications were suggested for 11 of the 16 definitions. There was significant variability in how these items were defined in the literature reviewed. The final round of voting achieved ≥ 80% agreement for all 16 definitions in the following areas: what constitutes respiratory support (invasive mechanical ventilation and noninvasive respiratory support), liberation and failed attempts to liberate from invasive mechanical ventilation, liberation from respiratory support, duration of noninvasive respiratory support, total duration of invasive mechanical ventilation, spontaneous breathing trials, extubation readiness testing, 28 ventilator-free days, and planned vs rescue use of post-extubation noninvasive respiratory support.InterpretationWe propose that these consensus-based definitions for elements of pediatric ventilator liberation, informed by evidence, be used for future quality improvement initiatives and research studies to improve generalizability and facilitate comparison.

Published
2023

10. How to Leverage DNN-based speech enhancement for multi-channel speaker verification?

Author

Dowerah, Sandipana, Serizel, Romain, Jouvet, Denis, Mohammadamini, Mohammad, and Matrouf, Driss

Subjects
Computer Science - Sound, Computer Science - Human-Computer Interaction, Electrical Engineering and Systems Science - Audio and Speech Processing
Abstract

Speaker verification (SV) suffers from unsatisfactory performance in far-field scenarios due to environmental noise andthe adverse impact of room reverberation. This work presents a benchmark of multichannel speech enhancement for far-fieldspeaker verification. One approach is a deep neural network-based, and the other is a combination of deep neural network andsignal processing. We integrated a DNN architecture with signal processing techniques to carry out various experiments. Ourapproach is compared to the existing state-of-the-art approaches. We examine the importance of enrollment in pre-processing,which has been largely overlooked in previous studies. Experimental evaluation shows that pre-processing can improve the SVperformance as long as the enrollment files are processed similarly to the test data and that test and enrollment occur within similarSNR ranges. Considerable improvement is obtained on the generated and all the noise conditions of the VOiCES dataset.

Published
2022

11. Phenotypic resistant single-cell characteristics under recurring ampicillin antibiotic exposure in Escherichia coli

Author

Silvia Kollerová, Lionel Jouvet, Julia Smelková, Sara Zunk-Parras, Alexandro Rodríguez-Rojas, and Ulrich K. Steiner

Subjects
single-cell, microfluidics, antibiotic resistance, β-lactam antibiotics, phenotypic resistance, persistence, Microbiology, QR1-502
Abstract

ABSTRACT Non-heritable, phenotypic drug resistance toward antibiotics challenges antibiotic therapies. Characteristics of such phenotypic resistance have implications for the evolution of heritable resistance. Diverse forms of phenotypic resistance have been described, but phenotypic resistance characteristics remain less explored than genetic resistance. Here, we add novel combinations of single-cell characteristics of phenotypic resistant E. coli cells and compare those to characteristics of susceptible cells of the parental population by exposure to different levels of recurrent ampicillin antibiotic. Contrasting expectations, we did not find commonly described characteristics of phenotypic resistant cells that arrest growth or near growth. We find that under ampicillin exposure, phenotypic resistant cells reduced their growth rate by about 50% compared to growth rates prior to antibiotic exposure. The growth reduction is a delayed alteration to antibiotic exposure, suggesting an induced response and not a stochastic switch or caused by a predetermined state as frequently described. Phenotypic resistant cells exhibiting constant slowed growth survived best under ampicillin exposure and, contrary to expectations, not only fast-growing cells suffered high mortality triggered by ampicillin but also growth-arrested cells. Our findings support diverse modes of phenotypic resistance, and we revealed resistant cell characteristics that have been associated with enhanced genetically fixed resistance evolution, which supports claims of an underappreciated role of phenotypic resistant cells toward genetic resistance evolution. A better understanding of phenotypic resistance will benefit combatting genetic resistance by developing and engulfing effective anti-phenotypic resistance strategies.IMPORTANCEAntibiotic resistance is a major challenge for modern medicine. Aside from genetic resistance to antibiotics, phenotypic resistance that is not heritable might play a crucial role for the evolution of antibiotic resistance. Using a highly controlled microfluidic system, we characterize single cells under recurrent exposure to antibiotics. Fluctuating antibiotic exposure is likely experienced under common antibiotic therapies. These phenotypic resistant cell characteristics differ from previously described phenotypic resistance, highlighting the diversity of modes of resistance. The phenotypic characteristics of resistant cells we identify also imply that such cells might provide a stepping stone toward genetic resistance, thereby causing treatment failure.

Published
2024
Full Text
View/download PDF

12. Hyperosmolarity in children with hyperammonemia: a risk of brain herniation at the start of renal replacement therapy

Author

Yousra Maghmoul, Arnaud Wiedemann, Lucile Barcat, Fabienne Parente, Pierre Allard, Fernando Alvarez, and Philippe Jouvet

Subjects
hyperosmolarity, hyperammonemia, liver failure, children, brain herniation, Pediatrics, RJ1-570
Abstract

PurposeRenal replacement therapy (RRT) is used in hyperammonemia to reduce the concentration of ammonia in the blood. In the case of plasma hyperosmolarity, RRT can also rapidly decrease plasma osmolarity, which may increase cerebral edema in these patients and favor the occurrence of brain herniation.MethodsWe conducted a retrospective clinical study in a tertiary care university-affiliated hospital. All patients admitted in a Pediatric Intensive Care Unit (PICU), less than 18 years old with ammonemia >150 µmol/L and who underwent RRT between January 2015 and June 2023 were included. We collected data on plasma osmolarity levels, osmolar gap and blood ammonia levels before and during RRT.ResultsEleven patients were included (10 with acute liver failure and 1 with a urea cycle disorders). Their mean age was 36.2 months. Before RRT, the median highest measured osmolarity was 320 (305–324) mOsm/L, whereas the median calculated osmolarity was 303 (293–314) mOsm/L, corresponding to an osmolar gap of 14 mOsm/L. Ammonia blood level over 400 µmol/L are significantly associated with higher plasma osmolarity (P-Value

Published
2024
Full Text
View/download PDF

13. Executive Summary of the Second International Guidelines for the Diagnosis and Management of Pediatric Acute Respiratory Distress Syndrome (PALICC-2).

Author

Emeriaud, Guillaume, López-Fernández, Yolanda, Iyer, Narayan, Bembea, Melania, Agulnik, Asya, Barbaro, Ryan, Baudin, Florent, Bhalla, Anoopindar, Brunow de Carvalho, Werther, Carroll, Christopher, Cheifetz, Ira, Chisti, Mohammod, Cruces, Pablo, Curley, Martha, Dahmer, Mary, Dalton, Heidi, Erickson, Simon, Essouri, Sandrine, Fernández, Analía, Flori, Heidi, Grunwell, Jocelyn, Jouvet, Philippe, Killien, Elizabeth, Kneyber, Martin, Kudchadkar, Sapna, Korang, Steven, Lee, Jan, Macrae, Duncan, Maddux, Aline, Modesto I Alapont, Vicent, Morrow, Brenda, Nadkarni, Vinay, Napolitano, Natalie, Newth, Christopher, Pons-Odena, Martí, Quasney, Michael, Rajapreyar, Prakadeshwari, Rambaud, Jerome, Randolph, Adrienne, Rimensberger, Peter, Rowan, Courtney, Sanchez-Pinto, L, Sapru, Anil, Sauthier, Michael, Shein, Steve, Smith, Lincoln, Steffen, Katerine, Takeuchi, Muneyuki, Thomas, Neal, Tse, Sze, Valentine, Stacey, Ward, Shan, Watson, R, Yehya, Nadir, Zimmerman, Jerry, and Khemani, Robinder

Subjects
Child, Humans, Respiratory Distress Syndrome, Respiration, Artificial, Acute Lung Injury, Consensus
Abstract

OBJECTIVES: We sought to update our 2015 work in the Second Pediatric Acute Lung Injury Consensus Conference (PALICC-2) guidelines for the diagnosis and management of pediatric acute respiratory distress syndrome (PARDS), considering new evidence and topic areas that were not previously addressed. DESIGN: International consensus conference series involving 52 multidisciplinary international content experts in PARDS and four methodology experts from 15 countries, using consensus conference methodology, and implementation science. SETTING: Not applicable. PATIENTS: Patients with or at risk for PARDS. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Eleven subgroups conducted systematic or scoping reviews addressing 11 topic areas: 1) definition, incidence, and epidemiology; 2) pathobiology, severity, and risk stratification; 3) ventilatory support; 4) pulmonary-specific ancillary treatment; 5) nonpulmonary treatment; 6) monitoring; 7) noninvasive respiratory support; 8) extracorporeal support; 9) morbidity and long-term outcomes; 10) clinical informatics and data science; and 11) resource-limited settings. The search included MEDLINE, EMBASE, and CINAHL Complete (EBSCOhost) and was updated in March 2022. Grading of Recommendations, Assessment, Development, and Evaluation methodology was used to summarize evidence and develop the recommendations, which were discussed and voted on by all PALICC-2 experts. There were 146 recommendations and statements, including: 34 recommendations for clinical practice; 112 consensus-based statements with 18 on PARDS definition, 55 on good practice, seven on policy, and 32 on research. All recommendations and statements had agreement greater than 80%. CONCLUSIONS: PALICC-2 recommendations and consensus-based statements should facilitate the implementation and adherence to the best clinical practice in patients with PARDS. These results will also inform the development of future programs of research that are crucially needed to provide stronger evidence to guide the pediatric critical care teams managing these patients.

Published
2023

14. Executive Summary: International Clinical Practice Guidelines for Pediatric Ventilator Liberation, A Pediatric Acute Lung Injury and Sepsis Investigators (PALISI) Network Document.

Author

Abu-Sultaneh, Samer, Iyer, Narayan, Fernández, Analía, Gaies, Michael, González-Dambrauskas, Sebastián, Hotz, Justin, Kneyber, Martin, López-Fernández, Yolanda, Rotta, Alexandre, Werho, David, Baranwal, Arun, Blackwood, Bronagh, Craven, Hannah, Curley, Martha, Essouri, Sandrine, Fioretto, Jose, Hartmann, Silvia, Jouvet, Philippe, Korang, Steven, Rafferty, Gerrard, Ramnarayan, Padmanabhan, Rose, Louise, Tume, Lyvonne, Whipple, Elizabeth, Wong, Judith, Emeriaud, Guillaume, Mastropietro, Christopher, Napolitano, Natalie, Newth, Christopher, and Khemani, Robinder

Subjects
airway extubation, clinical protocols, mechanical ventilators, pediatric intensive care units, ventilator weaning, Humans, Child, Respiration, Artificial, Ventilator Weaning, Ventilators, Mechanical, Airway Extubation, Sepsis
Abstract

Rationale: Pediatric-specific ventilator liberation guidelines are lacking despite the many studies exploring elements of extubation readiness testing. The lack of clinical practice guidelines has led to significant and unnecessary variation in methods used to assess pediatric patients readiness for extubation. Methods: Twenty-six international experts comprised a multiprofessional panel to establish pediatrics-specific ventilator liberation clinical practice guidelines, focusing on acutely hospitalized children receiving invasive mechanical ventilation for more than 24 hours. Eleven key questions were identified and first prioritized using the Modified Convergence of Opinion on Recommendations and Evidence. A systematic review was conducted for questions that did not meet an a priori threshold of ⩾80% agreement, with Grading of Recommendations, Assessment, Development, and Evaluation methodologies applied to develop the guidelines. The panel evaluated the evidence and drafted and voted on the recommendations. Measurements and Main Results: Three questions related to systematic screening using an extubation readiness testing bundle and a spontaneous breathing trial as part of the bundle met Modified Convergence of Opinion on Recommendations criteria of ⩾80% agreement. For the remaining eight questions, five systematic reviews yielded 12 recommendations related to the methods and duration of spontaneous breathing trials, measures of respiratory muscle strength, assessment of risk of postextubation upper airway obstruction and its prevention, use of postextubation noninvasive respiratory support, and sedation. Most recommendations were conditional and based on low to very low certainty of evidence. Conclusions: This clinical practice guideline provides a conceptual framework with evidence-based recommendations for best practices related to pediatric ventilator liberation.

Published
2023

15. Adaptation of Autoencoder for Sparsity Reduction From Clinical Notes Representation Learning

Author

Le, Thanh-Dung, Noumeir, Rita, Rambaud, Jerome, Sans, Guillaume, and Jouvet, Philippe

Subjects
Computer Science - Machine Learning, Computer Science - Computation and Language, Electrical Engineering and Systems Science - Signal Processing
Abstract

When dealing with clinical text classification on a small dataset recent studies have confirmed that a well-tuned multilayer perceptron outperforms other generative classifiers, including deep learning ones. To increase the performance of the neural network classifier, feature selection for the learning representation can effectively be used. However, most feature selection methods only estimate the degree of linear dependency between variables and select the best features based on univariate statistical tests. Furthermore, the sparsity of the feature space involved in the learning representation is ignored. Goal: Our aim is therefore to access an alternative approach to tackle the sparsity by compressing the clinical representation feature space, where limited French clinical notes can also be dealt with effectively. Methods: This study proposed an autoencoder learning algorithm to take advantage of sparsity reduction in clinical note representation. The motivation was to determine how to compress sparse, high-dimensional data by reducing the dimension of the clinical note representation feature space. The classification performance of the classifiers was then evaluated in the trained and compressed feature space. Results: The proposed approach provided overall performance gains of up to 3% for each evaluation. Finally, the classifier achieved a 92% accuracy, 91% recall, 91% precision, and 91% f1-score in detecting the patient's condition. Furthermore, the compression working mechanism and the autoencoder prediction process were demonstrated by applying the theoretic information bottleneck framework., Comment: Submitted to IEEE journal for review

Published
2022

16. Are disentangled representations all you need to build speaker anonymization systems?

Author

Champion, Pierre, Jouvet, Denis, and Larcher, Anthony

Subjects
Computer Science - Sound, Computer Science - Artificial Intelligence, Computer Science - Cryptography and Security, Computer Science - Machine Learning, Electrical Engineering and Systems Science - Audio and Speech Processing
Abstract

Speech signals contain a lot of sensitive information, such as the speaker's identity, which raises privacy concerns when speech data get collected. Speaker anonymization aims to transform a speech signal to remove the source speaker's identity while leaving the spoken content unchanged. Current methods perform the transformation by relying on content/speaker disentanglement and voice conversion. Usually, an acoustic model from an automatic speech recognition system extracts the content representation while an x-vector system extracts the speaker representation. Prior work has shown that the extracted features are not perfectly disentangled. This paper tackles how to improve features disentanglement, and thus the converted anonymized speech. We propose enhancing the disentanglement by removing speaker information from the acoustic model using vector quantization. Evaluation done using the VoicePrivacy 2022 toolkit showed that vector quantization helps conceal the original speaker identity while maintaining utility for speech recognition.

Published
2022

17. High-resolution LGM climate of Europe and the Alpine region using the regional climate model WRF

Author

E. Russo, J. Buzan, S. Lienert, G. Jouvet, P. Velasquez Alvarez, B. Davis, P. Ludwig, F. Joos, and C. C. Raible

Subjects
Environmental pollution, TD172-193.5, Environmental protection, TD169-171.8, Environmental sciences, GE1-350
Abstract

In this study we present a series of sensitivity experiments conducted for the Last Glacial Maximum (LGM, ∼21 ka) over Europe using the regional climate Weather Research and Forecasting model (WRF). Using a four-step two-way nesting approach, we are able to reach a convection-permitting horizontal resolution over the inner part of the study area, covering central Europe and the Alpine region. The main objective of the paper is to evaluate a model version including a series of new developments better suitable for the simulation of paleo-glacial time slices with respect to the ones employed in former studies. The evaluation of the model is conducted against newly available pollen-based reconstructions of the LGM European climate and takes into account the effect of two main sources of model uncertainty: a different height of continental glaciers at higher latitudes of the Northern Hemisphere and different land cover. Model results are in good agreement with evidence from the proxies, in particular for temperatures. Importantly, the consideration of different ensemble members for characterizing model uncertainty allows for increasing the agreement of the model against the proxy reconstructions that would be obtained when considering a single model realization. The spread of the produced ensemble is relatively small for temperature, besides areas surrounding glaciers in summer. On the other hand, differences between the different ensemble members are very pronounced for precipitation, in particular in winter over areas highly affected by moisture advection from the Atlantic. This highlights the importance of the considered sources of uncertainty for the study of European LGM climate and allows for determining where the results of a regional climate model (RCM) are more likely to be uncertain for the considered case study. Finally, the results are also used to assess the effect of convection-permitting resolutions, at both local and regional scales, under glacial conditions.

Published
2024
Full Text
View/download PDF

18. Label Propagation Techniques for Artifact Detection in Imbalanced Classes Using Photoplethysmogram Signals

Author

Clara Macabiau, Thanh-Dung Le, Kevin Albert, Mana Shahriari, Philippe Jouvet, and Rita Noumeir

Subjects
Motion artifacts, imbalanced classes, label propagation algorithm, machine learning classifiers, photoplethysmogram (PPG) signals, Electrical engineering. Electronics. Nuclear engineering, TK1-9971
Abstract

This study aimed to investigate the application of label propagation techniques to propagate labels among photoplethysmogram (PPG) signals, particularly in imbalanced class scenarios and limited data availability scenarios, where clean PPG samples are significantly outnumbered by artifact-contaminated samples. We investigated a dataset comprising PPG recordings from 1571 patients, wherein approximately 82% of the samples were identified as clean, while the remaining 18% were contaminated by artifacts. Our research compares the performance of supervised classifiers, such as conventional classifiers and neural networks (Multi-Layer Perceptron (MLP), Transformers, Fully Convolutional Network (FCN)), with the semi-supervised Label Propagation (LP) algorithm for artifact classification in PPG signals. The results indicate that the LP algorithm achieves a precision of 91%, a recall of 90%, and an F1 score of 90% for the “artifacts” class, showcasing its effectiveness in annotating a medical dataset, even in cases where clean samples are rare. Although the K-Nearest Neighbors (KNN) supervised model demonstrated good results with a precision of 89%, a recall of 95%, and an F1 score of 92%, the semi-supervised algorithm excels in artifact detection. In the case of imbalanced and limited pediatric intensive care environment data, the semi-supervised LP algorithm is promising for artifact detection in PPG signals. The results of this study are important for improving the accuracy of PPG-based health monitoring, particularly in situations in which motion artifacts pose challenges to data interpretation.

Published
2024
Full Text
View/download PDF

19. Characteristics and outcomes of COVID-19 patients admitted to hospital with and without respiratory symptoms

Author

Barbara Wanjiru Citarella, Christiana Kartsonaki, Elsa D. Ibáñez-Prada, Bronner P. Gonçalves, Joaquin Baruch, Martina Escher, Mark G. Pritchard, Jia Wei, Fred Philippy, Andrew Dagens, Matthew Hall, James Lee, Demetrios James Kutsogiannis, Evert-Jan Wils, Marília Andreia Fernandes, Bharath Kumar Tirupakuzhi Vijayaraghavan, Prasan Kumar Panda, Ignacio Martin-Loeches, Shinichiro Ohshimo, Arie Zainul Fatoni, Peter Horby, Jake Dunning, Jordi Rello, Laura Merson, Amanda Rojek, Michel Vaillant, Piero Olliaro, Luis Felipe Reyes, S.A. Moharam, Sabriya Abdalasalam, Alaa Abdalfattah Abdalhadi, Naana Reyam Abdalla, Walaa Abdalla, Almthani Hamza Abdalrheem, Ashraf Abdalsalam, Saedah Abdeewi, Esraa Hassan Abdelgaum, Mohamed Abdelhalim, Mohammed Abdelkabir, Israa Abdelrahman, Sheryl Ann Abdukahil, Lamees Adil Abdulbaqi, Salaheddin Abdulhamid, Widyan Abdulhamid, Nurul Najmee Abdulkadir, Eman Abdulwahed, Rawad Abdunabi, Ryuzo Abe, Laurent Abel, Ahmed Mohammed Abodina, Amal Abrous, Lara Absil, Kamal Abu Jabal, Nashat Abu Salah, Abdurraouf Abusalama, Tareg Abdallah Abuzaid, Subhash Acharya, Andrew Acker, Elisabeth Adam, Safia Adem, Manuella Ademnou, Francisca Adewhajah, Diana Adrião, Anthony Afum-Adjei Awuah, Melvin Agbogbatey, Saleh Al Ageel, Aya Mustafa Ahmed, Musaab Mohammed Ahmed, Shakeel Ahmed, Zainab Ahmed Alaraji, Abdulrahman Ahmed Elhefnawy Enan, Reham Abdelhamid Ahmed Khalil, Ali Mostafa Ahmed Mohamed Abdelaziz, Kate Ainscough, Eka Airlangga, Tharwat Aisa, Ali Aisha, Bugila Aisha, Ali Ait Hssain, Younes Ait Tamlihat, Takako Akimoto, Ernita Akmal, Chika Akwani, Eman Al Qasim, Ahmed Alajeeli, Ahmed Alali, Razi Alalqam, Aliya Mohammed Alameen, Mohammed Al-Aquily, Zinah A. Alaraji, Khalid Albakry, Safa Albatni, Angela Alberti, Osama Aldabbourosama, Tala Al-dabbous, Amer Aldhalia, Abdulkarim Aldoukali, Senthilkumar Alegesan, Marta Alessi, Beatrice Alex, Kévin Alexandre, Abdulrahman Al-Fares, Asil Alflite, Huda Alfoudri, Qamrah Alhadad, Hoda Salem Alhaddad, Maali Khalid Mohamed Abdalla Alhasan, Ahmad Nabil Alhouri, Hasan Alhouri, Adam Ali, Imran Ali, Maha TagElser Mohammed Ali, Syed Ali Abbas, Yomna Ali Abdelghafar, Naseem Ali Sheikh, Kazali Enagnon Alidjnou, Mahmoud Aljadi, Sarah Aljamal, Mohammed Alkahlout, Akram Alkaseek, Qabas Alkhafajee, Clotilde Allavena, Nathalie Allou, Lana Almasri, Abdulrahman Almjersah, Raja Ahmed Alqandouz, Walaa Alrfaea, Moayad Alrifaee, Rawan Alsaadi, Yousef Al-Saba'a, Entisar Alshareea, Eslam Alshenawy, Aneela Altaf, João Melo Alves, João Alves, Rita Alves, Joana Alves Cabrita, Maria Amaral, Amro Essam Amer, Nur Amira, Amos Amoako Adusei, John Amuasi, Roberto Andini, Claire Andrejak, Andrea Angheben, François Angoulvant, Sophia Ankrah, Séverine Ansart, Sivanesen Anthonidass, Massimo Antonelli, Carlos Alexandre Antunes de Brito, Ardiyan Apriyana, Yaseen Arabi, Irene Aragao, Francisco Arancibia, Carolline Araujo, Antonio Arcadipane, Patrick Archambault, Lukas Arenz, Jean-Benoît Arlet, Christel Arnold-Day, Lovkesh Arora, Rakesh Arora, Elise Artaud-Macari, Diptesh Aryal, Angel Asensio, Elizabeth A. Ashley, Muhammad Ashraf, Muhammad Sheharyar Ashraf, Abir Ben Ashur, Franklin Asiedu-Bekoe, Namra Asif, Mohammad Asim, Grace Assi, Jean Baptiste Assie, Amirul Asyraf, Fouda Atangana, Ahmed Atia, Minahel Atif, Asia Atif Abdelrhman Abdallahrs, Anika Atique, Moad Atlowly, AM Udara Lakshan Attanyake, Johann Auchabie, Hugues Aumaitre, Adrien Auvet, Abdelmalek Awad Ali Mohammed, Eyvind W. Axelsen, Ared Ayad, Ahmed Ayman Hassan Helmi, Laurène Azemar, Mohammed Azizeldin, Cecile Azoulay, Hakeem Babatunde, Benjamin Bach, Delphine Bachelet, Claudine Badr, Roar Bævre-Jensen, Nadia Baig, John Kenneth Baillie, J Kevin Baird, Erica Bak, Agamemnon Bakakos, Nazreen Abu Bakar, Hibah Bileid Bakeer, Ashraf Bakri, Andriy Bal, Mohanaprasanth Balakrishnan, Irene Bandoh, Firouzé Bani-Sadr, Renata Barbalho, Nicholas Yuri Barbosa, Wendy S. Barclay, Saef Umar Barnett, Michaela Barnikel, Helena Barrasa, Cleide Barrigoto, Marie Bartoli, Joaquín Baruch, Romain Basmaci, Muhammad Fadhli Hassin Basri, AbdAlkarim Batool, Denise Battaglini, Jules Bauer, Diego Fernando Bautista Rincon, Denisse Bazan Dow, Abigail Beane, Alexandra Bedossa, Ker Hong Bee, Husna Begum, Sylvie Behilill, Albertus Beishuizen, Aleksandr Beljantsev, David Bellemare, Anna Beltrame, Beatriz Amorim Beltrão, Marine Beluze, Nicolas Benech, Lionel Eric Benjiman, Suzanne Bennett, Luís Bento, Jan-Erik Berdal, Lamis Berdeweel, Delphine Bergeaud, Hazel Bergin, Giulia Bertoli, Lorenzo Bertolino, Simon Bessis, Sybille Bevilcaqua, Karine Bezulier, Amar Bhatt, Krishna Bhavsar, Isabella Bianchi, Claudia Bianco, Sandra Bichoka, Farah Nadiah Bidin, Felwa Bin Humaid, Mohd Nazlin Bin Kamarudin, Muhannud Binnawara, Zeno Bisoffi, Patrick Biston, Laurent Bitker, Mustapha Bittaye, Jonathan Bitton, Pablo Blanco-Schweizer, Catherine Blier, Frank Bloos, Mathieu Blot, Filomena Boccia, Laetitia Bodenes, Debby Bogaert, Anne-Hélène Boivin, Ariel Bolanga, Isabela Bolaños, Pierre-Adrien Bolze, François Bompart, Aurelius Bonifasius, Joe Bonney, Diogo Borges, Raphaël Borie, Hans Martin Bosse, Elisabeth Botelho-Nevers, Lila Bouadma, Olivier Bouchaud, Sabelline Bouchez, Damien Bouhour, Kévin Bouiller, Laurence Bouillet, Camile Bouisse, Latsaniphone Bountthasavong, Anne-Sophie Boureau, John Bourke, Maude Bouscambert, Aurore Bousquet, Marielle Boyer-Besseyre, Maria Boylan, Fernando Augusto Bozza, Axelle Braconnier, Cynthia Braga, Timo Brandenburger, Filipa Brás Monteiro, Luca Brazzi, Dorothy Breen, Patrick Breen, David Brewster, Kathy Brickell, Tessa Broadley, Helen Brotherton, Alex Browne, Nicolas Brozzi, Sonja Hjellegjerde Brunvoll, Marjolein Brusse-Keizer, Petra Bryda, Nina Buchtele, Polina Bugaeva, Marielle Buisson, Danilo Buonsenso, Erlina Burhan, Donald Buri, Aidan Burrell, Ingrid G. Bustos, Denis Butnaru, André Cabie, Susana Cabral, Joana Cabrita, Eder Caceres, Cyril Cadoz, Rui Caetano Garcês, Kate Calligy, Jose Andres Calvache, João Camões, Valentine Campana, Paul Campbell, Josie Campisi, Cecilia Canepa, Mireia Cantero, Janice Caoili, Pauline Caraux-Paz, Sheila Cárcel, Filipa Cardoso, Filipe Cardoso, Nelson Cardoso, Sofia Cardoso, Simone Carelli, Nicolas Carlier, Thierry Carmoi, Gayle Carney, Inês Carqueja, Marie-Christine Carret, François Martin Carrier, Ida Carroll, Gail Carson, Maire-Laure Casanova, Mariana Cascão, Siobhan Casey, José Casimiro, Bailey Cassandra, Silvia Castañeda, Nidyanara Castanheira, Guylaine Castor-Alexandre, Ivo Castro, Ana Catarino, François-Xavier Catherine, Paolo Cattaneo, Roberta Cavalin, Giulio Giovanni Cavalli, Alexandros Cavayas, Adrian Ceccato, Masaneh Ceesay, Shelby Cerkovnik, Minerva Cervantes-Gonzalez, Muge Cevik, Anissa Chair, Catherine Chakveatze, Adrienne Chan, Meera Chand, Jean-Marc Chapplain, Charlotte Charpentier, Julie Chas, Muhammad Mobin Chaudry, Jonathan Samuel Chávez Iñiguez, Anjellica Chen, Yih-Sharng Chen, Léo Chenard, Matthew Pellan Cheng, Antoine Cheret, Thibault Chiarabini, Julian Chica, Suresh Kumar Chidambaram, Leong Chin Tho, Catherine Chirouze, Davide Chiumello, Sung-Min Cho, Bernard Cholley, Danoy Chommanam, Marie-Charlotte Chopin, Yock Ping Chow, Ting Soo Chow, Nathaniel Christy, Hiu Jian Chua, Jonathan Chua, Jose Pedro Cidade, José Miguel Cisneros Herreros, Anna Ciullo, Jennifer Clarke, Rolando Claure-Del Granado, Sara Clohisey, Cassidy Codan, Caitriona Cody, Jennifer Coles, Megan Coles, Gwenhaël Colin, Michael Collins, Pamela Combs, Jennifer Connolly, Marie Connor, Anne Conrad, Elaine Conway, Graham S. Cooke, Hugues Cordel, Amanda Corley, Sabine Cornelis, Alexander Daniel Cornet, Arianne Joy Corpuz, Andrea Cortegiani, Grégory Corvaisier, Camille Couffignal, Sandrine Couffin-Cadiergues, Roxane Courtois, Stéphanie Cousse, Juthaporn Cowan, Rachel Cregan, Gloria Crowl, Jonathan Crump, Claudina Cruz, Marc Csete, Ailbhe Cullen, Matthew Cummings, Gerard Curley, Elodie Curlier, Colleen Curran, Paula Custodio, Ana da Silva Filipe, Charlene Da Silveira, Al-Awwab Dabaliz, John Arne Dahl, Darren Dahly, Umberto D'Alessandro, Peter Daley, Zaina Dalloul, Heidi Dalton, Jo Dalton, Seamus Daly, Juliana Damas, Joycelyn Dame, Cammandji Damien, Nick Daneman, Jorge Dantas, Frédérick D'Aragon, Gillian de Loughry, Diego de Mendoza, Etienne De Montmollin, Rafael Freitas de Oliveira França, Ana Isabel de Pinho Oliveira, Rosanna De Rosa, Cristina De Rose, Thushan de Silva, Peter de Vries, Jillian Deacon, David Dean, Alexa Debard, Bianca DeBenedictis, Marie-Pierre Debray, Nathalie DeCastro, William Dechert, Romain Decours, Eve Defous, Isabelle Delacroix, Alexandre Delamou, Eric Delaveuve, Karen Delavigne, Nathalie M. Delfos, Ionna Deligiannis, Andrea Dell'Amore, Christelle Delmas, Pierre Delobel, Corine Delsing, Elisa Demonchy, Emmanuelle Denis, Dominique Deplanque, Pieter Depuydt, Diane Descamps, Mathilde Desvallées, Santi Dewayanti, Pathik Dhangar, Alpha Diallo, Souleymane Taran Diallo, Sylvain Diamantis, André Dias, Fernanda Dias Da Silva, Rodrigo Diaz, Juan Jose Diaz, Priscila Diaz, Bakary K. Dibba, Kévin Didier, Jean-Luc Diehl, Wim Dieperink, Jérôme Dimet, Vincent Dinot, Fara Diop, Alphonsine Diouf, Yael Dishon, Cedric Djadda, Félix Djossou, Annemarie B. Docherty, Helen Doherty, Arjen M. Dondorp, Christl A. Donnelly, Yoann Donohue, Sean Donohue, Peter Doran, Céline Dorival, Eric D'Ortenzio, Yash Doshi, Phouvieng Douangdala, James Joshua Douglas, Renee Douma, Nathalie Dournon, Joanne Downey, Mark Downing, Thomas Drake, Aoife Driscoll, Ibrahim Kwaku Duah, Claudio Duarte Fonseca, Vincent Dubee, François Dubos, Audrey Dubot-Pérès, Alexandre Ducancelle, Toni Duculan, Susanne Dudman, Abhijit Duggal, Paul Dunand, Mathilde Duplaix, Emanuele Durante-Mangoni, Lucian Durham, III, Bertrand Dussol, Juliette Duthoit, Xavier Duval, Anne Margarita Dyrhol-Riise, Sim Choon Ean, Ada Ebo, Marco Echeverria-Villalobos, Michael Edelstein, Siobhan Egan, Linn Margrete Eggesbø, Khadeja Ehzaz, Carla Eira, Mohammed El Sanharawi, Marwan El Sayed, Mohammed Elabid, Mohamed Bashir Elagili, Subbarao Elapavaluru, Mohammad Elbahnasawy, Sohail Elboshra, Brigitte Elharrar, Jacobien Ellerbroek, Merete Ellingjord-Dale, Hamida ELMagrahi, Mohammad Muatasm Elmubark, Loubna Elotmani, Lauren Eloundou, Philippine Eloy, Basma Elshaikhy, Tarek Elshazly, Wafa Elsokni, Aml Ahmed Eltayeb, Iqbal Elyazar, Zarief Kamel Emad, Hussein Embarek, Isabelle Enderle, Tomoyuki Endo, Gervais Eneli, Chan Chee Eng, Ilka Engelmann, Vincent Enouf, Olivier Epaulard, Haneen Esaadi, Mariano Esperatti, Hélène Esperou, Catarina Espírito Santo, Marina Esposito-Farese, Rachel Essaka, Lorinda Essuman, João Estevão, Manuel Etienne, Anna Greti Everding, Mirjam Evers, Isabelle Fabre, Marc Fabre, Ismaila Fadera, Asgad Osman Abdalla Fadlalla, Amna Faheem, Arabella Fahy, Cameron J. Fairfield, Zul Fakar, Komal Fareed, Pedro Faria, Ahmed Farooq, Hanan Fateena, Mohamed Fathi, Salem Fatima, Karine Faure, Raphaël Favory, Mohamed Fayed, Niamh Feely, Jorge Fernandes, Susana Fernandes, François-Xavier Ferrand, Eglantine Ferrand Devouge, Joana Ferrão, Mário Ferraz, Benigno Ferreira, Isabel Ferreira, Bernardo Ferreira, Sílvia Ferreira, Nicolas Ferriere, Céline Ficko, Claudia Figueiredo-Mello, William Finlayson, Thomas Flament, Tom Fletcher, Aline-Marie Florence, Letizia Lucia Florio, Brigid Flynn, Deirdre Flynn, Jean Foley, Victor Fomin, Tatiana Fonseca, Patricia Fontela, Karen Forrest, Simon Forsyth, Denise Foster, Giuseppe Foti, Berline Fotso, Erwan Fourn, Robert A. Fowler, Marianne Fraher, Diego Franch-Llasat, Christophe Fraser, John F. Fraser, Marcela Vieira Freire, Ana Freitas Ribeiro, Craig French, Caren Friedrich, Ricardo Fritz, Stéphanie Fry, Nora Fuentes, Masahiro Fukuda, G. Argin, Valérie Gaborieau, Rostane Gaci, Massimo Gagliardi, Jean-Charles Gagnard, Amandine Gagneux-Brunon, Abdou Gai, Sérgio Gaião, Linda Gail Skeie, Adham Mohamed Galal Mohamed Ramadan, Phil Gallagher, Carrol Gamble, Yasmin Gani, Arthur Garan, Rebekha Garcia, Julia Garcia-Diaz, Esteban Garcia-Gallo, Navya Garimella, Denis Garot, Valérie Garrait, Basanta Gauli, Anatoliy Gavrylov, Alexandre Gaymard, Johannes Gebauer, Eva Geraud, Louis Gerbaud Morlaes, Nuno Germano, Malak Ghemmeid, Praveen Kumar Ghisulal, Jade Ghosn, Marco Giani, Tristan Gigante, Elaine Gilroy, Guillermo Giordano, Michelle Girvan, Valérie Gissot, Gezy Giwangkancana, Daniel Glikman, Petr Glybochko, Eric Gnall, Geraldine Goco, François Goehringer, Siri Goepel, Jean-Christophe Goffard, Jin Yi Goh, Brigitta Golács, Jonathan Golob, Kyle Gomez, Joan Gómez-Junyent, Marie Gominet, Alicia Gonzalez, Patricia Gordon, Isabelle Gorenne, Laure Goubert, Cécile Goujard, Tiphaine Goulenok, Margarite Grable, Jeronimo Graf, Edward Wilson Grandin, Pascal Granier, Giacomo Grasselli, Lorenzo Grazioli, Christopher A. Green, Courtney Greene, William Greenhalf, Segolène Greffe, Domenico Luca Grieco, Matthew Griffee, Fiona Griffiths, Ioana Grigoras, Albert Groenendijk, Fassou Mathias Grovogui, Heidi Gruner, Yusing Gu, Jérémie Guedj, Martin Guego, Anne-Marie Guerguerian, Daniela Guerreiro, Romain Guery, Anne Guillaumot, Laurent Guilleminault, Maisa Guimarães de Castro, Thomas Guimard, Marieke Haalboom, Daniel Haber, Ali Hachemi, Abdurrahman Haddud, Nadir Hadri, Wael Hafez, Fakhir Raza Haidri, Fatima Mhd Rida Hajij, Sheeba Hakak, Adam Hall, Sophie Halpin, Shaher Hamdan, Abdelhafeez Hamdi, Jawad Hameed, Ansley Hamer, Raph L. Hamers, Rebecca Hamidfar, Bato Hammarström, Naomi Hammond, Terese Hammond, Lim Yuen Han, Matly Hanan, Rashan Haniffa, Kok Wei Hao, Hayley Hardwick, Ewen M. Harrison, Janet Harrison, Samuel Bernard Ekow Harrison, Alan Hartman, Sulieman Hasan, Mohammad Ali Nabil Hasan, Mohd Shahnaz Hasan, Junaid Hashmi, Madiha Hashmi, Amoni Hassan, Ebtisam Hassanin, Muhammad Hayat, Ailbhe Hayes, Leanne Hays, Jan Heerman, Lars Heggelund, Ahmed Helmi, Ross Hendry, Martina Hennessy, Aquiles Rodrigo Henriquez-Trujillo, Maxime Hentzien, Diana Hernandez, Andrew Hershey, Liv Hesstvedt, Astarini Hidayah, Eibhlin Higgins, Rupert Higgins, Samuel Hinton, Hiroaki Hiraiwa, Haider Hirkani, Hikombo Hitoto, Antonia Ho, Yi Bin Ho, Alexandre Hoctin, Isabelle Hoffmann, Wei Han Hoh, Oscar Hoiting, Rebecca Holt, Jan Cato Holter, Juan Pablo Horcajada, Ikram Houas, Mabrouka Houderi, Catherine L. Hough, Stuart Houltham, Jimmy Ming-Yang Hsu, Jean-Sébastien Hulot, Abby Hurd, Iqbal Hussain, Aliae Mohamed Hussein, Mahmood Hussein, Fatima Ibrahim, Bashir Ibran, Samreen Ijaz, M. Arfan Ikram, Carlos Cañada Illana, Patrick Imbert, Muhammad Imran Ansari, Rana Imran Sikander, Hugo Inácio, Carmen Infante Dominguez, Yun Sii Ing, Mariachiara Ippolito, Vera Irawany, Sarah Isgett, Tiago Isidoro, Nadiah Ismail, Margaux Isnard, Mette Stausland Istre, Junji Itai, Daniel Ivulich, Danielle Jaafar, Salma Jaafoura, Hamza Jaber, Julien Jabot, Clare Jackson, Abubacarr Jagne, Stéphane Jaureguiberry, Denise Jaworsky, Florence Jego, Anilawati Mat Jelani, Synne Jenum, Ruth Jimbo-Sotomayor, Ong Yiaw Joe, Ruth Noemí Jorge García, Silje Bakken Jørgensen, Cédric Joseph, Mark Joseph, Swosti Joshi, Mercé Jourdain, Philippe Jouvet, Anna Jung, Hanna Jung, Dafsah Juzar, Ouifiya Kafif, Florentia Kaguelidou, Neerusha Kaisbain, Thavamany Kaleesvran, Sabina Kali, Karl Trygve Kalleberg, Smaragdi Kalomoiri, Muhammad Aisar Ayadi Kamaluddin, Armand Saloun Kamano, Zul Amali Che Kamaruddin, Nadiah Kamarudin, Kavita Kamineni, Darshana Hewa Kandamby, Kong Yeow Kang, Darakhshan Kanwal, Dyah Kanyawati, Mohamed Karghul, Pratap Karpayah, Todd Karsies, Daisuke Kasugai, Kevin Katz, Christy Kay, Lamees Kayyali, Seán Keating, Pulak Kedia, Andrea Kelly, Aoife Kelly, Claire Kelly, Niamh Kelly, Sadie Kelly, Yvelynne Kelly, Maeve Kelsey, Kalynn Kennon, Sommay Keomany, Maeve Kernan, Younes Kerroumi, Sharma Keshav, Shams Khail, Sarah Khaled, Imrana Khalid, Antoine Khalil, Irfan Khan, Quratul Ain Khan, Sushil Khanal, Abid Khatak, Krish Kherajani, Michelle E. Kho, Denisa Khoo, Ryan Khoo, Saye Khoo, Muhammad Nasir Khoso, Amin Khuwaja, Khor How Kiat, Yuri Kida, Peter Kiiza, Beathe Kiland Granerud, Anders Benjamin Kildal, Jae Burm Kim, Antoine Kimmoun, Detlef Kindgen-Milles, Nobuya Kitamura, Eyrun Floerecke Kjetland Kjetland, Paul Klenerman, Rob Klont, Gry Kloumann Bekken, Stephen R. Knight, Robin Kobbe, Paa Kobina Forson, Chamira Kodippily, Malte Kohns Vasconcelos, Sabin Koirala, Mamoru Komatsu, Franklina Korkor Abebrese, Volkan Korten, Stephanie Kouba, Mohamed Lamine Kourouma, Karifa Kourouma, Arsène Kpangon, Karolina Krawczyk, Ali Kredan, Vinothini Krishnan, Sudhir Krishnan, Oksana Kruglova, Anneli Krund, Pei Xuan Kuan, Ashok Kumar, Deepali Kumar, Ganesh Kumar, Mukesh Kumar, Dinesh Kuriakose, Ethan Kurtzman, Demetrios Kutsogiannis, Galyna Kutsyna, Ama Kwakyewaa Bedu-Addo, Sylvie Kwedi, Konstantinos Kyriakoulis, Marie Lachatre, Marie Lacoste, John G. Laffey, Nadhem Lafhej, Marie Lagrange, Fabrice Laine, Olivier Lairez, Sanjay Lakhey, Marc Lambert, François Lamontagne, Marie Langelot-Richard, Vincent Langlois, Eka Yudha Lantang, Marina Lanza, Cédric Laouénan, Samira Laribi, Delphine Lariviere, Stéphane Lasry, Sakshi Lath, Naveed Latif, Youssef Latifeh, Odile Launay, Didier Laureillard, Yoan Lavie-Badie, Andy Law, Cassie Lawrence, Teresa Lawrence, Minh Le, Clément Le Bihan, Cyril Le Bris, Georges Le Falher, Lucie Le Fevre, Quentin Le Hingrat, Marion Le Maréchal, Soizic Le Mestre, Gwenaël Le Moal, Vincent Le Moing, Hervé Le Nagard, Ema Leal, Marta Leal Santos, Biing Horng Lee, Heng Gee Lee, Su Hwan Lee, Jennifer Lee, Todd C. Lee, Yi Lin Lee, Gary Leeming, Bénédicte Lefebvre, Laurent Lefebvre, Benjamin Lefèvre, Sylvie LeGac, Merili-Helen Lehiste, Jean-Daniel Lelievre, François Lellouche, Adrien Lemaignen, Véronique Lemee, Anthony Lemeur, Gretchen Lemmink, Ha Sha Lene, Jenny Lennon, Rafael León, Marc Leone, Tanel Lepik, Quentin Lepiller, François-Xavier Lescure, Olivier Lesens, Mathieu Lesouhaitier, Amy Lester-Grant, Andrew Letizia, Sophie Letrou, Bruno Levy, Yves Levy, Claire Levy-Marchal, Katarzyna Lewandowska, Erwan L'Her, Gianluigi Li Bassi, Janet Liang, Ali Liaquat, Geoffrey Liegeon, Kah Chuan Lim, Wei Shen Lim, Chantre Lima, Bruno Lina, Lim Lina, Andreas Lind, Maja Katherine Lingad, Guillaume Lingas, Sylvie Lion-Daolio, Keibun Liu, Marine Livrozet, Patricia Lizotte, Antonio Loforte, Navy Lolong, Leong Chee Loon, Diogo Lopes, Dalia Lopez-Colon, Anthony L. Loschner, Paul Loubet, Bouchra Loufti, Guillame Louis, Silvia Lourenco, Lara Lovelace-Macon, Lee Lee Low, Marije Lowik, Jia Shyi Loy, Jean Christophe Lucet, Carlos M. Luna, Olguta Lungu, Miles Lunn, Liem Luong, Nestor Luque, Dominique Luton, Olavi Maasikas, Moïse Machado, Sara Machado, Gabriel Macheda, Mustafa Magzoub, Rafael Mahieu, Sophie Mahy, Ana Raquel Maia, Lars S. Maier, Oumou Maiga Ascofare, Mylène Maillet, Thomas Maitre, Nimisha Abdul Majeed, Maximilian Malfertheiner, Nadia Malik, Paddy Mallon, Fernando Maltez, Denis Malvy, Victoria Manda, Laurent Mandelbrot, Frank Manetta, Julie Mankikian, Edmund Manning, Aldric Manuel, Veronika Maráczi, Ceila Maria Sant′Ana Malaque, Flávio Marino, Samuel Markowicz, Ana Marques, Catherine Marquis, Laura Marsh, Brian Marsh, Megan Marshal, John Marshall, Celina Turchi Martelli, Dori-Ann Martin, Emily Martin, Guillaume Martin-Blondel, Alessandra Martinelli, F. Eduardo Martinez, Martin Martinot, Alejandro Martín-Quiros, Ana Martins, João Martins, Nuno Martins, Caroline Martins Rego, Gennaro Martucci, Olga Martynenko, Eva Miranda Marwali, Marsilla Marzukie, David Maslove, Sabina Mason, Sobia Masood, Fatma Masoud, Moise Massoma, Palmer Masumbe, Mohd Basri Mat Nor, Moshe Matan, Henrique Mateus Fernandes, Meghena Mathew, Christina Mathew, Mathieu Mattei, Laurence Maulin, Juergen May, Javier Maynar, Mayfong Mayxay, Thierry Mazzoni, Lisa Mc Sweeney, Colin McArthur, Naina McCann, Peter McCanny, Aine McCarthy, Anne McCarthy, Colin McCloskey, Rachael McConnochie, Sherry McDermott, Sarah E. McDonald, Aine McElroy, Samuel McElwee, Natalie McEvoy, Allison McGeer, Kenneth A. McLean, Paul McNally, Bairbre McNicholas, Edel Meaney, Cécile Mear-Passard, Maggie Mechlin, Nastia Medombou, Omar Mehkri, Ferruccio Mele, Luis Melo, Kashif Ali Memon, João João Mendes, Ogechukwu Menkiti, Kusum Menon, France Mentré, Alexander J. Mentzer, Emmanuelle Mercier, Noémie Mercier, Antoine Merckx, Mayka Mergeay-Fabre, Blake Mergler, António Mesquita, Roberta Meta, Osama Metwally, Agnès Meybeck, Dan Meyer, Alison M. Meynert, Vanina Meysonnier, Mehdi Mezidi, Céline Michelanglei, Isabelle Michelet, Efstathia Mihelis, Vladislav Mihnovit, Duha Milad Abdullah, Jennene Miller, Hugo Miranda-Maldonado, Nor Arisah Misnan, Nik Nur Eliza Mohamed, Nouralsabah Mohamed, Tahira Jamal Mohamed, Alaa Mohamed Ads, Ahmed Reda Mohamed Elsayed Abdelhalim, Libya Mohammed, Shrouk Fawze Mohammed Mostafa, Manahil Omer Abdelrahman Mohammedahmed, Omer Abdullah Mohammedelhassan, Asma Moin, Walaa Mokhtar, Elena Molinos, Brenda Molloy, Mary Mone, Agostinho Monteiro, Claudia Montes, Giorgia Montrucchio, Sarah Moore, Shona C. Moore, Lina Morales Cely, Marwa Morgom, Lucia Moro, Catherine Motherway, Ana Motos, Hugo Mouquet, Clara Mouton Perrot, Julien Moyet, Suleiman Haitham Mualla, Mohamed Muftah, Aisha Kalsoom Mufti, Ng Yong Muh, Mo'nes Muhaisen, Dzawani Muhamad, Jimmy Mullaert, Fredrik Müller, Karl Erik Müller, Daniel Munblit, Syed Muneeb Ali, Nadeem Munir, Laveena Munshi, Aisling Murphy, Patrick Murray, Marlène Murris, Srinivas Murthy, Himed Musaab, Alamin Mustafa, Mus'ab Mustafa, Dana Mustafa, Himasha Muvindi, Dimitra Melia Myrodia, Farah Nadia Mohd-Hanafiah, Behzad Nadjm, Dave Nagpal, Alex Nagrebetsky, Blanka Nagybányai-Nagy, Herwin Nanda Boudoin, Mangala Narasimhan, Nageswaran Narayanan, Prashant Nasa, Rashid Nasim Khan, Ahmad Nasrallah, Adel Gerges Nassif Metri, Alasdair Nazerali-Maitland, Nadège Neant, Holger Neb, Nikita Nekliudov, Matthew Nelder, Erni Nelwan, Raul Neto, Emily Neumann, Wing Yiu Ng, Pauline Yeung Ng, Anthony Nghi, Duc Nguyen, Orna Ni Choileain, Niamh Ni Leathlobhair, Nerissa Niba, Alistair D. Nichol, Prompak Nitayavardhana, Stephanie Nonas, Nurul Amani Mohd Noordin, Nurul Faten Izzati Norharizam, Anita North, Alessandra Notari, Mahdad Noursadeghi, Adam Nowinski, Saad Nseir, Leonard Numfor, Nurnaningsih Nurnaningsih, Dwi Utomo Nusantara, Elsa Nyamankolly, Anders Benteson Nygaard, Fionnuala O. Brien, Annmarie O. Callaghan, Annmarie O'Callaghan, Giovanna Occhipinti, Derbrenn OConnor, Max O'Donnell, Lawrence Ofori-Boadu, Tawnya Ogston, Takayuki Ogura, Tak-Hyuk Oh, Sophie O'Halloran, Katie O'Hearn, Sally-Ann Ohene, João Oliveira, Larissa Oliveira, Piero L. Olliaro, Cinderella Omar Rageh Elnaggar, Alsarrah Ali Mohammed Omer, Pierre Ondobo, David S.Y. Ong, Jee Yan Ong, Wilna Oosthuyzen, Anne Opavsky, Peter Openshaw, Saijad Orakzai, Claudia Milena Orozco-Chamorro, Jamel Ortoleva, Mohamed Osama Elsayed Soliman, Javier Osatnik, Linda O'Shea, Miriam O'Sullivan, Eman Othman, Siti Zubaidah Othman, Nadia Ouamara, Rachida Ouissa, Micheal Owusu, Ama Akyampomaa Owusu-Asare, Eric Oziol, Maïder Pagadoy, Justine Pages, Amanda Palacios, Massimo Palmarini, Giovanna Panarello, Hem Paneru, Lai Hui Pang, Mauro Panigada, Nathalie Pansu, Aurélie Papadopoulos, Rachael Parke, Melissa Parker, Jérémie Pasquier, Bruno Pastene, Fabian Patauner, Drashti Patel, Mohan Dass Pathmanathan, Luís Patrão, Patricia Patricio, Lisa Patterson, Rajyabardhan Pattnaik, Christelle Paul, Mical Paul, Jorge Paulos, William A. Paxton, Jean-François Payen, Sandra L. Peake, Kalaiarasu Peariasamy, Giles J. Peek, Florent Peelman, Nathan Peiffer-Smadja, Vincent Peigne, Mare Pejkovska, Paolo Pelosi, Ithan D. Peltan, Rui Pereira, Daniel Perez, Thomas Perpoint, Antonio Pesenti, Vincent Pestre, Lenka Petrou, Michele Petrovic, Ventzislava Petrov-Sanchez, Frank Olav Pettersen, Gilles Peytavin, Richard Odame Philips, Ooyanong Phonemixay, Soulichanya Phoutthavong, Michael Piagnerelli, Walter Picard, Olivier Picone, Maria de Piero, Djura Piersma, Carlos Pimentel, Raquel Pinto, Catarina Pires, Lionel Piroth, Ayodhia Pitaloka, Chiara Piubelli, Riinu Pius, Simone Piva, Laurent Plantier, Hon Shen Png, Julien Poissy, Ryadh Pokeerbux, Sergio Poli, Georgios Pollakis, Diane Ponscarme, Diego Bastos Porto, Andra-Maris Post, Douwe F. Postma, Pedro Povoa, Diana Póvoas, Jeff Powis, Sofia Prapa, Viladeth Praphasiri, Sébastien Preau, Christian Prebensen, Jean-Charles Preiser, Anton Prinssen, Gamage Dona Dilanthi Priyadarshani, Lucia Proença, Sravya Pudota, Bambang Pujo Semedi, Mathew Pulicken, Peter Puplampu, Gregory Purcell, Luisa Quesada, Vilmaris Quinones-Cardona, Else Quist-Paulsen, Mohammed Quraishi, Fadi Qutishat, Maia Rabaa, Christian Rabaud, Ebenezer Rabindrarajan, Aldo Rafael, Marie Rafiq, Abdelrahman Ragab, Mutia Rahardjani, Arslan Rahat Ullah, Ahmad Kashfi Haji Ab Rahman, Rozanah Abd Rahman, Fernando Rainieri, Giri Shan Rajahram, Pratheema Ramachandran, Nagarajan Ramakrishnan, José Ramalho, Ahmad Afiq Ramli, Blandine Rammaert, Grazielle Viana Ramos, Asim Rana, Rajavardhan Rangappa, Ritika Ranjan, Christophe Rapp, Aasiyah Rashan, Thalha Rashan, Ghulam Rasheed, Menaldi Rasmin, Indrek Rätsep, Cornelius Rau, Tharmini Ravi, Ali Raza, Andre Real, Stanislas Rebaudet, Sarah Redl, Brenda Reeve, Attaur Rehman, Muhammad Osama Rehman Khalid, Dag Henrik Reikvam, Renato Reis, Jonathan Remppis, Martine Remy, Hongru Ren, Hanna Renk, Anne-Sophie Resseguier, Matthieu Revest, Oleksa Rewa, Maria Ines Ribeiro, Antonia Ricchiuto, David Richardson, Denise Richardson, Laurent Richier, Siti Nurul Atikah Ahmad Ridzuan, Ana L. Rios, Asgar Rishu, Patrick Rispal, Karine Risso, Maria Angelica Rivera Nuñez, Chiara Robba, André Roberto, Stephanie Roberts, Charles Roberts, David L. Robertson, Olivier Robineau, Anna Roca, Ferran Roche-Campo, Paola Rodari, Simão Rodeia, Bernhard Roessler, Claire Roger, Pierre-Marie Roger, Roberto Roncon-Albuquerque, Jr., Mélanie Roriz, Manuel Rosa-Calatrava, Michael Rose, Dorothea Rosenberger, Andrea Rossanese, Matteo Rossetti, Patrick Rossignol, Carine Roy, Benoît Roze, Desy Rusmawatiningtyas, Clark D. Russell, Maeve Ryan, Steffi Ryckaert, Aleksander Rygh Holten, Isabela Saba, Sairah Sadaf, Musharaf Sadat, Valla Sahraei, Abdurraouf Said, Nadia Saidani, Pranya Sakiyalak, Fodé Bangaly Sako, Moamen Salah, Ali Alaa Salah Eldin Mohamed Abbas, Nawal Salahuddin, Leonardo Salazar, Jodat Saleem, Mohammed Saleh Alyasiri, Talat Ahmed Abu Salem, Gabriele Sales, Charlotte Salmon Gandonniere, Hélène Salvator, Dana Samardali, Shaden Samardali, Yehia Samir Shaaban Aly Orabi, Emely Sanchez, Olivier Sanchez, Kizy Sanchez de Oliveira, Angel Sanchez-Miralles, Vanessa Sancho-Shimizu, Gyan Sandhu, Zulfiqar Sandhu, Pierre-François Sandrine, Oana Săndulescu, Marlene Santos, Shirley Sarfo-Mensah, Bruno Sarmento Banheiro, Iam Claire E. Sarmiento, Benjamine Sarton, Ankana Satya, Sree Satyapriya, Rumaisah Satyawati, Egle Saviciute, Yen Tsen Saw, Justin Schaffer, Tjard Schermer, Arnaud Scherpereel, Marion Schneider, János Schnur, Stephan Schroll, Michael Schwameis, Gary Schwartz, Janet T. Scott, James Scott-Brown, Nicholas Sedillot, Tamara Seitz, Jaganathan Selvanayagam, Mageswari Selvarajoo, Malcolm G. Semple, Rasidah Bt Senian, Eric Senneville, Claudia Sepulveda, Filipa Sequeira, Tânia Sequeira, Ary Serpa Neto, Ellen Shadowitz, Syamin Asyraf Shahidan, Hamza Shahla, Laila Shalabi, Haitam Shames, Anuraj Shankar, Shaikh Sharjeel, Pratima Sharma, Catherine A. Shaw, Victoria Shaw, John Robert Sheenan, Dr. Rajesh Mohan Shetty, Rohan Shetty, Mohiuddin Shiekh, Nobuaki Shime, Keiki Shimizu, Sally Shrapnel, Shubha Kalyan Shrestha, Pramesh Sundar Shrestha, Hoi Ping Shum, Nassima Si Mohammed, Ng Yong Siang, Moses Siaw-Frimpong, Jeanne Sibiude, Bountoy Sibounheuang, Nidhal Siddig, Atif Siddiqui, Maqsood Ahmed Siddiqui, Louise Sigfrid, Fatoumata Sillah, Piret Sillaots, Catarina Silva, Maria Joao Silva, Rogério Silva, Benedict Sim Lim Heng, Wai Ching Sin, Dario Sinatti, Mahendra Singh, Punam Singh, Pompini Agustina Sitompul, Karisha Sivam, Vegard Skogen, Sue Smith, Benjamin Smood, Coilin Smyth, Morgane Snacken, Dominic So, Tze Vee Soh, Lene Bergendal Solberg, Joshua Solomon, Tom Solomon, Emily Somers, Agnès Sommet, Myung Jin Song, Rima Song, Tae Song, Jack Song Chia, Arne Søraas, Albert Sotto, Edouard Soum, Ana Chora Sousa, Marta Sousa, Maria Sousa Uva, Vicente Souza-Dantas, Mamadou Saliou Sow, Alexandra Sperry, Elisabetta Spinuzza, B. P. Sanka Ruwan Sri Darshana, Shiranee Sriskandan, Sarah Stabler, Thomas Staudinger, Stephanie-Susanne Stecher, Trude Steinsvik, Ymkje Stienstra, Birgitte Stiksrud, Eva Stolz, Amy Stone, Anca Streinu-Cercel, Adrian Streinu-Cercel, Geoff Strong, Ami Stuart, David Stuart, Richa Su, Decy Subekti, Gabriel Suen, Jacky Y. Suen, Prasanth Sukumar, Asfia Sultana, Charlotte Summers, Dubravka Supic, Deepashankari Suppiah, Magdalena Surovcová, Atie Suwarti, Andrey Svistunov, Sarah Syahrin, Augustina Sylverken, Konstantinos Syrigos, Jaques Sztajnbok, Konstanty Szuldrzynski, Shirin Tabrizi, Fabio S. Taccone, Lysa Tagherset, Shahdattul Mawarni Taib, Sara Taleb, Cheikh Talla, Jelmer Talsma, Renaud Tamisier, Maria Lawrensia Tampubolon, Kim Keat Tan, Yan Chyi Tan, Hiroyuki Tanaka, Taku Tanaka, Hayato Taniguchi, Huda Taqdees, Arshad Taqi, Coralie Tardivon, Yousef Tarek Kamal Mostafa, Ali Tarhabat, Pierre Tattevin, M Azhari Taufik, Hassan Tawfik, Tze Yuan Tee, João Teixeira, Sofia Tejada, Marie-Capucine Tellier, Sze Kye Teoh, Vanessa Teotonio, François Téoulé, Olivier Terrier, Nicolas Terzi, Hubert Tessier-Grenier, Adrian Tey, Alif Adlan Mohd Thabit, Anand Thakur, Zhang Duan Tham, Suvintheran Thangavelu, Elmi Theron, Vincent Thibault, Simon-Djamel Thiberville, Benoît Thill, Jananee Thirumanickam, Niamh Thompson, Shaun Thompson, Emma C. Thomson, David Thomson, Mathew Thorpe, Surain Raaj Thanga Thurai, Ryan S. Thwaites, Paul Tierney, Vadim Tieroshyn, Peter S. Timashev, Jean-François Timsit, Noémie Tissot, Fiona Toal, Jordan Zhien Yang Toh, Maria Toki, Kristian Tonby, Sia Loong Tonnii, Marta Torre, Antoni Torres, Margarida Torres, Rosario Maria Torres Santos-Olmo, Hernando Torres-Zevallos, Aboubacar Tounkara, Michael Towers, Fodé Amara Traoré, Tony Trapani, Cécile Tromeur, Ioannis Trontzas, Tiffany Trouillon, Jeanne Truong, Christelle Tual, Sarah Tubiana, Helen Tuite, Alexis F. Turgeon, Jean-Marie Turmel, Lance C.W. Turtle, Anders Tveita, Pawel Twardowski, Makoto Uchiyama, PG Ishara Udayanga, Andrew Udy, Roman Ullrich, Alberto Uribe, Asad Usman, Effua Usuf, Timothy M. Uyeki, Cristinava Vajdovics, Piero Valentini, Luís Val-Flores, Stijn Van de Velde, Marcel van den Berge, Machteld van der Feltz, Job van der Palen, Paul van der Valk, Nicky Van Der Vekens, Peter Van der Voort, Sylvie Van Der Werf, Laura van Gulik, Jarne Van Hattem, Carolien van Netten, Ilonka van Veen, Noémie Vanel, Henk Vanoverschelde, Michael Varrone, Shoban Raj Vasudayan, Charline Vauchy, Pavan Kumar Vecham, Shaminee Veeran, Aurélie Veislinger, Sebastian Vencken, Sara Ventura, Annelies Verbon, José Ernesto Vidal, César Vieira, Deepak Vijayan, Judit Villar, Pierre-Marc Villeneuve, Andrea Villoldo, Gayatri Vishwanathan, Benoit Visseaux, Hannah Visser, Chiara Vitiello, Manivanh Vongsouvath, Harald Vonkeman, Fanny Vuotto, Suhaila Abdul Wahab, Noor Hidayu Wahab, Nadirah Abdul Wahid, Marina Wainstein, Laura Walsh, Wan Fadzlina Wan Muhd Shukeri, Chih-Hsien Wang, Steve Webb, Katharina Weil, Tan Pei Wen, Hassi Wesam, Sanne Wesselius, T. Eoin West, Murray Wham, Bryan Whelan, Nicole White, Paul Henri Wicky, Aurélie Wiedemann, Surya Otto Wijaya, Keith Wille, Sue Willems, Bailey Williams, Patricia J. Williams, Virginie Williams, Jessica Wittman, Calvin Wong, Xin Ci Wong, Yew Sing Wong, Teck Fung Wong, Natalie Wright, Lim Saio Xian, Ioannis Xynogalas, Siti Rohani Binti Mohd Yakop, Masaki Yamazaki, Elizabeth Yarad, Yazdan Yazdanpanah, Nicholas Yee Liang Hing, Abdelrahman Yehia Mahmoud Abdelaal, Cécile Yelnik, Chian Hui Yeoh, Stephanie Yerkovich, Touxiong Yiaye, Toshiki Yokoyama, Hodane Yonis, Obada Yousif, Saptadi Yuliarto, Akram Zaaqoq, Marion Zabbe, Gustavo E. Zabert, Kai Zacharowski, Masliza Zahid, Maram Zahran, Nor Zaila Binti Zaidan, Maria Zambon, Miguel Zambrano, Alberto Zanella, Nurul Zaynah, Hiba Zayyad, Alexander Zoufaly, and David Zucman

Subjects
COVID-19, Non-respiratory symptoms, Respiratory symptoms, Risk factors, Mortality, Science (General), Q1-390, Social sciences (General), H1-99
Abstract

Background: COVID-19 is primarily known as a respiratory illness; however, many patients present to hospital without respiratory symptoms. The association between non-respiratory presentations of COVID-19 and outcomes remains unclear. We investigated risk factors and clinical outcomes in patients with no respiratory symptoms (NRS) and respiratory symptoms (RS) at hospital admission. Methods: This study describes clinical features, physiological parameters, and outcomes of hospitalised COVID-19 patients, stratified by the presence or absence of respiratory symptoms at hospital admission. RS patients had one or more of: cough, shortness of breath, sore throat, runny nose or wheezing; while NRS patients did not. Results: Of 178,640 patients in the study, 86.4 % presented with RS, while 13.6 % had NRS. NRS patients were older (median age: NRS: 74 vs RS: 65) and less likely to be admitted to the ICU (NRS: 36.7 % vs RS: 37.5 %). NRS patients had a higher crude in-hospital case-fatality ratio (NRS 41.1 % vs. RS 32.0 %), but a lower risk of death after adjusting for confounders (HR 0.88 [0.83–0.93]). Conclusion: Approximately one in seven COVID-19 patients presented at hospital admission without respiratory symptoms. These patients were older, had lower ICU admission rates, and had a lower risk of in-hospital mortality after adjusting for confounders.

Published
2024
Full Text
View/download PDF

20. Privacy-Preserving Speech Representation Learning using Vector Quantization

Author

Champion, Pierre, Jouvet, Denis, and Larcher, Anthony

Subjects
Electrical Engineering and Systems Science - Audio and Speech Processing, Computer Science - Artificial Intelligence, Computer Science - Computation and Language, Computer Science - Cryptography and Security, Computer Science - Sound
Abstract

With the popularity of virtual assistants (e.g., Siri, Alexa), the use of speech recognition is now becoming more and more widespread.However, speech signals contain a lot of sensitive information, such as the speaker's identity, which raises privacy concerns.The presented experiments show that the representations extracted by the deep layers of speech recognition networks contain speaker information.This paper aims to produce an anonymous representation while preserving speech recognition performance.To this end, we propose to use vector quantization to constrain the representation space and induce the network to suppress the speaker identity.The choice of the quantization dictionary size allows to configure the trade-off between utility (speech recognition) and privacy (speaker identity concealment)., Comment: Journ{\'e}es d'{\'E}tudes sur la Parole - JEP2022, Jun 2022, {\^I}le de Noirmoutier, France

Published
2022

21. Failure of human rhombic lip differentiation underlies medulloblastoma formation

Author

Hendrikse, Liam D, Haldipur, Parthiv, Saulnier, Olivier, Millman, Jake, Sjoboen, Alexandria H, Erickson, Anders W, Ong, Winnie, Gordon, Victor, Coudière-Morrison, Ludivine, Mercier, Audrey L, Shokouhian, Mohammad, Suárez, Raúl A, Ly, Michelle, Borlase, Stephanie, Scott, David S, Vladoiu, Maria C, Farooq, Hamza, Sirbu, Olga, Nakashima, Takuma, Nambu, Shohei, Funakoshi, Yusuke, Bahcheli, Alec, Diaz-Mejia, J Javier, Golser, Joseph, Bach, Kathleen, Phuong-Bao, Tram, Skowron, Patryk, Wang, Evan Y, Kumar, Sachin A, Balin, Polina, Visvanathan, Abhirami, Lee, John JY, Ayoub, Ramy, Chen, Xin, Chen, Xiaodi, Mungall, Karen L, Luu, Betty, Bérubé, Pierre, Wang, Yu C, Pfister, Stefan M, Kim, Seung-Ki, Delattre, Olivier, Bourdeaut, Franck, Doz, François, Masliah-Planchon, Julien, Grajkowska, Wieslawa A, Loukides, James, Dirks, Peter, Fèvre-Montange, Michelle, Jouvet, Anne, French, Pim J, Kros, Johan M, Zitterbart, Karel, Bailey, Swneke D, Eberhart, Charles G, Rao, Amulya AN, Giannini, Caterina, Olson, James M, Garami, Miklós, Hauser, Peter, Phillips, Joanna J, Ra, Young S, de Torres, Carmen, Mora, Jaume, Li, Kay KW, Ng, Ho-Keung, Poon, Wai S, Pollack, Ian F, López-Aguilar, Enrique, Gillespie, G Yancey, Van Meter, Timothy E, Shofuda, Tomoko, Vibhakar, Rajeev, Thompson, Reid C, Cooper, Michael K, Rubin, Joshua B, Kumabe, Toshihiro, Jung, Shin, Lach, Boleslaw, Iolascon, Achille, Ferrucci, Veronica, de Antonellis, Pasqualino, Zollo, Massimo, Cinalli, Giuseppe, Robinson, Shenandoah, Stearns, Duncan S, Van Meir, Erwin G, Porrati, Paola, Finocchiaro, Gaetano, Massimino, Maura, Carlotti, Carlos G, Faria, Claudia C, Roussel, Martine F, Boop, Frederick, Chan, Jennifer A, Aldinger, Kimberly A, Razavi, Ferechte, Silvestri, Evelina, McLendon, Roger E, and Thompson, Eric M

Subjects
Stem Cell Research, Brain Disorders, Neurosciences, Rare Diseases, Stem Cell Research - Nonembryonic - Non-Human, Aetiology, 1.1 Normal biological development and functioning, Underpinning research, 2.1 Biological and endogenous factors, Cell Differentiation, Cell Lineage, Cerebellar Neoplasms, Cerebellum, Core Binding Factor alpha Subunits, Hedgehog Proteins, Histone Demethylases, Humans, Ki-67 Antigen, Medulloblastoma, Metencephalon, Muscle Proteins, Mutation, Otx Transcription Factors, Repressor Proteins, T-Box Domain Proteins, Transcription Factors, General Science & Technology
Abstract

Medulloblastoma (MB) comprises a group of heterogeneous paediatric embryonal neoplasms of the hindbrain with strong links to early development of the hindbrain1-4. Mutations that activate Sonic hedgehog signalling lead to Sonic hedgehog MB in the upper rhombic lip (RL) granule cell lineage5-8. By contrast, mutations that activate WNT signalling lead to WNT MB in the lower RL9,10. However, little is known about the more commonly occurring group 4 (G4) MB, which is thought to arise in the unipolar brush cell lineage3,4. Here we demonstrate that somatic mutations that cause G4 MB converge on the core binding factor alpha (CBFA) complex and mutually exclusive alterations that affect CBFA2T2, CBFA2T3, PRDM6, UTX and OTX2. CBFA2T2 is expressed early in the progenitor cells of the cerebellar RL subventricular zone in Homo sapiens, and G4 MB transcriptionally resembles these progenitors but are stalled in developmental time. Knockdown of OTX2 in model systems relieves this differentiation blockade, which allows MB cells to spontaneously proceed along normal developmental differentiation trajectories. The specific nature of the split human RL, which is destined to generate most of the neurons in the human brain, and its high level of susceptible EOMES+KI67+ unipolar brush cell progenitor cells probably predisposes our species to the development of G4 MB.

Published
2022

22. On the invertibility of a voice privacy system using embedding alignement

Author

Champion, Pierre, Thebaud, Thomas, Lan, Gaël Le, Larcher, Anthony, and Jouvet, Denis

Subjects
Electrical Engineering and Systems Science - Audio and Speech Processing, Computer Science - Cryptography and Security, Computer Science - Machine Learning, Computer Science - Sound
Abstract

This paper explores various attack scenarios on a voice anonymization system using embeddings alignment techniques. We use Wasserstein-Procrustes (an algorithm initially designed for unsupervised translation) or Procrustes analysis to match two sets of x-vectors, before and after voice anonymization, to mimic this transformation as a rotation function. We compute the optimal rotation and compare the results of this approximation to the official Voice Privacy Challenge results. We show that a complex system like the baseline of the Voice Privacy Challenge can be approximated by a rotation, estimated using a limited set of x-vectors. This paper studies the space of solutions for voice anonymization within the specific scope of rotations. Rotations being reversible, the proposed method can recover up to 62% of the speaker identities from anonymized embeddings.

Published
2021

23. Evaluating X-vector-based Speaker Anonymization under White-box Assessment

Author

Champion, Pierre, Jouvet, Denis, and Larcher, Anthony

Subjects
Computer Science - Sound, Computer Science - Artificial Intelligence, Computer Science - Cryptography and Security, Electrical Engineering and Systems Science - Audio and Speech Processing
Abstract

In the scenario of the Voice Privacy challenge, anonymization is achieved by converting all utterances from a source speaker to match the same target identity; this identity being randomly selected. In this context, an attacker with maximum knowledge about the anonymization system can not infer the target identity. This article proposed to constrain the target selection to a specific identity, i.e., removing the random selection of identity, to evaluate the extreme threat under a whitebox assessment (the attacker has complete knowledge about the system). Targeting a unique identity also allows us to investigate whether some target's identities are better than others to anonymize a given speaker.

Published
2021

24. Data Representation Structure to Support Clinical Decision-Making in the Pediatric Intensive Care Unit: Interview Study and Preliminary Decision Support Interface Design

Author

Najia Yakob, Sandrine Laliberté, Philippe Doyon-Poulin, Philippe Jouvet, and Rita Noumeir

Subjects
Medicine
Abstract

BackgroundClinical decision-making is a complex cognitive process that relies on the interpretation of a large variety of data from different sources and involves the use of knowledge bases and scientific recommendations. The representation of clinical data plays a key role in the speed and efficiency of its interpretation. In addition, the increasing use of clinical decision support systems (CDSSs) provides assistance to clinicians in their practice, allowing them to improve patient outcomes. In the pediatric intensive care unit (PICU), clinicians must process high volumes of data and deal with ever-growing workloads. As they use multiple systems daily to assess patients’ status and to adjust the health care plan, including electronic health records (EHR), clinical systems (eg, laboratory, imaging and pharmacy), and connected devices (eg, bedside monitors, mechanical ventilators, intravenous pumps, and syringes), clinicians rely mostly on their judgment and ability to trace relevant data for decision-making. In these circumstances, the lack of optimal data structure and adapted visual representation hinder clinician’s cognitive processes and clinical decision-making skills. ObjectiveIn this study, we designed a prototype to optimize the representation of clinical data collected from existing sources (eg, EHR, clinical systems, and devices) via a structure that supports the integration of a home-developed CDSS in the PICU. This study was based on analyzing end user needs and their clinical workflow. MethodsFirst, we observed clinical activities in a PICU to secure a better understanding of the workflow in terms of staff tasks and their use of EHR on a typical work shift. Second, we conducted interviews with 11 clinicians from different staff categories (eg, intensivists, fellows, nurses, and nurse practitioners) to compile their needs for decision support. Third, we structured the data to design a prototype that illustrates the proposed representation. We used a brain injury care scenario to validate the relevance of integrated data and the utility of main functionalities in a clinical context. Fourth, we held design meetings with 5 clinicians to present, revise, and adapt the prototype to meet their needs. ResultsWe created a structure with 3 levels of abstraction—unit level, patient level, and system level—to optimize clinical data representation and display for efficient patient assessment and to provide a flexible platform to host the internally developed CDSS. Subsequently, we designed a preliminary prototype based on this structure. ConclusionsThe data representation structure allows prioritizing patients via criticality indicators, assessing their conditions using a personalized dashboard, and monitoring their courses based on the evolution of clinical values. Further research is required to define and model the concepts of criticality, problem recognition, and evolution. Furthermore, feasibility tests will be conducted to ensure user satisfaction.

Published
2024
Full Text
View/download PDF

25. Pediatric Organ Dysfunction Information Update Mandate (PODIUM) Contemporary Organ Dysfunction Criteria: Executive Summary.

Author

Bembea, Melania, Agus, Michael, Akcan-Arikan, Ayse, Alexander, Peta, Basu, Rajit, Bennett, Tellen, Bohn, Desmond, Brandão, Leonardo, Brown, Ann-Marie, Carcillo, Joseph, Checchia, Paul, Cholette, Jill, Cheifetz, Ira, Cornell, Timothy, Doctor, Allan, Eckerle, Michelle, Erickson, Simon, Farris, Reid, Faustino, E, Fitzgerald, Julie, Fuhrman, Dana, Giuliano, John, Guilliams, Kristin, Gaies, Michael, Gorga, Stephen, Hall, Mark, Hanson, Sheila, Hartman, Mary, Hassinger, Amanda, Irving, Sharon, Jeffries, Howard, Jouvet, Philippe, Kannan, Sujatha, Karam, Oliver, Khemani, Robinder, Kissoon, Niranjan, Lacroix, Jacques, Laussen, Peter, Leclerc, Francis, Lee, Jan, Leteurtre, Stephane, Lobner, Katie, McKiernan, Patrick, Menon, Kusum, Monagle, Paul, Muszynski, Jennifer, Odetola, Folafoluwa, Parker, Robert, Pathan, Nazima, Pierce, Richard, Pineda, Jose, Prince, Jose, Robinson, Karen, Rowan, Courtney, Ryerson, Lindsay, Sanchez-Pinto, L, Schlapbach, Luregn, Selewski, David, Shekerdemian, Lara, Simon, Dennis, Smith, Lincoln, Squires, James, Squires, Robert, Sutherland, Scott, Ouellette, Yves, Spaeder, Michael, Srinivasan, Vijay, Steiner, Marie, Tasker, Robert, Thiagarajan, Ravi, Thomas, Neal, Tissieres, Pierre, Traube, Chani, Tucci, Marisa, Typpo, Katri, Wainwright, Mark, Ward, Shan, Watson, R, Weiss, Scott, Whitney, Jane, Willson, Doug, Wynn, James, Yehya, Nadir, and Zimmerman, Jerry

Subjects
Child, Critical Care, Critical Illness, Evidence-Based Medicine, Humans, Multiple Organ Failure, Organ Dysfunction Scores
Abstract

Prior criteria for organ dysfunction in critically ill children were based mainly on expert opinion. We convened the Pediatric Organ Dysfunction Information Update Mandate (PODIUM) expert panel to summarize data characterizing single and multiple organ dysfunction and to derive contemporary criteria for pediatric organ dysfunction. The panel was composed of 88 members representing 47 institutions and 7 countries. We conducted systematic reviews of the literature to derive evidence-based criteria for single organ dysfunction for neurologic, cardiovascular, respiratory, gastrointestinal, acute liver, renal, hematologic, coagulation, endocrine, endothelial, and immune system dysfunction. We searched PubMed and Embase from January 1992 to January 2020. Study identification was accomplished using a combination of medical subject headings terms and keywords related to concepts of pediatric organ dysfunction. Electronic searches were performed by medical librarians. Studies were eligible for inclusion if the authors reported original data collected in critically ill children; evaluated performance characteristics of scoring tools or clinical assessments for organ dysfunction; and assessed a patient-centered, clinically meaningful outcome. Data were abstracted from each included study into an electronic data extraction form. Risk of bias was assessed using the Quality in Prognosis Studies tool. Consensus was achieved for a final set of 43 criteria for pediatric organ dysfunction through iterative voting and discussion. Although the PODIUM criteria for organ dysfunction were limited by available evidence and will require validation, they provide a contemporary foundation for researchers to identify and study single and multiple organ dysfunction in critically ill children.

Published
2022

26. Machine Learning Based on Natural Language Processing to Detect Cardiac Failure in Clinical Narratives

Author

Le, Thanh-Dung, Noumeir, Rita, Rambaud, Jerome, Sans, Guillaume, and Jouvet, Philippe

Subjects
Computer Science - Computation and Language, Electrical Engineering and Systems Science - Signal Processing
Abstract

The purpose of the study presented herein is to develop a machine learning algorithm based on natural language processing that automatically detects whether a patient has a cardiac failure or a healthy condition by using physician notes in Research Data Warehouse at CHU Sainte Justine Hospital. First, a word representation learning technique was employed by using bag-of-word (BoW), term frequency inverse document frequency (TFIDF), and neural word embeddings (word2vec). Each representation technique aims to retain the words semantic and syntactic analysis in critical care data. It helps to enrich the mutual information for the word representation and leads to an advantage for further appropriate analysis steps. Second, a machine learning classifier was used to detect the patients condition for either cardiac failure or stable patient through the created word representation vector space from the previous step. This machine learning approach is based on a supervised binary classification algorithm, including logistic regression (LR), Gaussian Naive-Bayes (GaussianNB), and multilayer perceptron neural network (MLPNN). Technically, it mainly optimizes the empirical loss during training the classifiers. As a result, an automatic learning algorithm would be accomplished to draw a high classification performance, including accuracy (acc), precision (pre), recall (rec), and F1 score (f1). The results show that the combination of TFIDF and MLPNN always outperformed other combinations with all overall performance. In the case without any feature selection, the proposed framework yielded an overall classification performance with acc, pre, rec, and f1 of 84% and 82%, 85%, and 83%, respectively. Significantly, if the feature selection was well applied, the overall performance would finally improve up to 4% for each evaluation., Comment: Submitted to 2021 34th IEEE International Symposium on Computer-Based Medical Systems (CBMS)

Published
2021

27. Detecting of a Patient's Condition From Clinical Narratives Using Natural Language Representation

Author

Le, Thanh-Dung, Noumeir, Rita, Rambaud, Jerome, Sans, Guillaume, and Jouvet, Philippe

Subjects
Computer Science - Computation and Language, Electrical Engineering and Systems Science - Signal Processing
Abstract

The rapid progress in clinical data management systems and artificial intelligence approaches enable the era of personalized medicine. Intensive care units (ICUs) are the ideal clinical research environment for such development because they collect many clinical data and are highly computerized environments. We designed a retrospective clinical study on a prospective ICU database using clinical natural language to help in the early diagnosis of heart failure in critically ill children. The methodology consisted of empirical experiments of a learning algorithm to learn the hidden interpretation and presentation of the French clinical note data. This study included 1386 patients' clinical notes with 5444 single lines of notes. There were 1941 positive cases (36 % of total) and 3503 negative cases classified by two independent physicians using a standardized approach. The multilayer perceptron neural network outperforms other discriminative and generative classifiers. Consequently, the proposed framework yields an overall classification performance with 89 % accuracy, 88 % recall, and 89 % precision. This study successfully applied learning representation and machine learning algorithms to detect heart failure from clinical natural language in a single French institution. Further work is needed to use the same methodology in other institutions and other languages., Comment: Accepted for publication in IEEE Open Journal of Engineering in Medicine and Biology. arXiv admin note: text overlap with arXiv:2104.03934

Published
2021
Full Text
View/download PDF

28. Subglacial hydrology from high-resolution ice-flow simulations of the Rhine Glacier during the Last Glacial Maximum: a proxy for glacial erosion

Author

D. Cohen, G. Jouvet, T. Zwinger, A. Landgraf, and U. H. Fischer

Subjects
Geology, QE1-996.5
Abstract

At the Last Glacial Maximum (LGM), the Rhine Glacier complex (Rhine and Linth glaciers) formed large piedmont lobes extending north into the Swiss and German Alpine forelands. Numerous overdeepened valleys there were formed by repeated glaciations. A characteristic of these overdeepened valleys is their location close to the LGM ice margin, away from the Alps. Numerical models of ice flow of the Rhine Glacier indicate a poor fit between the sliding distance, a proxy for glacial erosion, and the location of these overdeepenings. Calculations of the hydraulic potential based on the computed time-dependent ice surface elevations of the Rhine Glacier lobe obtained from a high-resolution thermo-mechanically coupled Stokes flow model are used to estimate the location of subglacial water drainage routes. Results indicate that the subglacial water discharge is high and focused along glacial valleys and overdeepenings when water pressure is equal to the ice overburden pressure. These conditions are necessary for subglacial water to remove basal sediments, expose fresh bedrock, and favor further erosion by quarrying and abrasion. Knowledge of the location of paleo-subglacial water drainage routes may be useful to understand patterns of subglacial erosion beneath paleo-ice masses that do not otherwise relate to the sliding of ice. Comparison of the erosion pattern from subglacial meltwater with those from quarrying and abrasion shows the importance of subglacial water flow in the formation of distal overdeepenings in the Swiss lowlands.

Published
2023
Full Text
View/download PDF

30. Non-destructive detection of large molecules without mass limitation

Author

Poindron, Adrien, Pedregosa-Gutierrez, Jofre, Jouvet, Christophe, Knoop, Martina, and Champenois, Caroline

Subjects
Physics - Instrumentation and Detectors, Quantum Physics
Abstract

The problem for molecular identification knows many solutions which include mass spectrometers whose mass sensitivity depends on the performance of the detector involved. The purpose of this article is to show by means of molecular dynamics simulations, how a laser-cooled ion cloud, confined in a linear radio-frequency trap, can reach the ultimate sensitivity providing the detection of individual charged heavy molecular ions. In our simulations, we model the laser-cooled Ca + ions as two-level atoms, confined thanks to a set of constant and time oscillating electrical fields. A singly-charged molecular ion with a mass of 10 6 amu is propelled through the ion cloud. The induced change in the fluorescence rate of the lather is used as the detection signal. We show that this signal is due to a significant temperature variation triggered by the Coulombian repulsion and amplified by the radio-frequency heating induced by the trap itself. We identify the optimum initial energy for the molecular ion to be detected and furthermore, we characterize the performance of the detector for a large range of confinement voltages.

Published
2021
Full Text
View/download PDF

31. A Study of F0 Modification for X-Vector Based Speech Pseudonymization Across Gender

Author

Champion, Pierre, Jouvet, Denis, and Larcher, Anthony

Subjects
Electrical Engineering and Systems Science - Audio and Speech Processing, Computer Science - Cryptography and Security, Computer Science - Sound
Abstract

Speech pseudonymization aims at altering a speech signal to map the identifiable personal characteristics of a given speaker to another identity. In other words, it aims to hide the source speaker identity while preserving the intelligibility of the spoken content. This study takes place in the VoicePrivacy 2020 challenge framework, where the baseline system performs pseudonymization by modifying x-vector information to match a target speaker while keeping the fundamental frequency (F0) unchanged. We propose to alter other paralin-guistic features, here F0, and analyze the impact of this modification across gender. We found that the proposed F0 modification always improves pseudonymization We observed that both source and target speaker genders affect the performance gain when modifying the F0.

Published
2021

32. Observational constraints on the sensitivity of two calving glaciers to external forcings

Author

Andrea Kneib-Walter, Martin P. Lüthi, Martin Funk, Guillaume Jouvet, and Andreas Vieli

Subjects
Calving, glacier calving, glacier monitoring, ice velocity, remote sensing, Environmental sciences, GE1-350, Meteorology. Climatology, QC851-999
Abstract

Future mass loss projections of the Greenland ice sheet require understanding of the processes at a glacier terminus, especially of iceberg calving. We present detailed and high-rate terrestrial radar interferometer observations of Eqip Sermia and Bowdoin Glacier, two outlet glaciers in Greenland with comparable dimensions and investigate iceberg calving, surface elevation, velocity, strain rates and their links to air temperature, tides and topography. The results reveal that the two glaciers exhibit very different flow and calving behaviour on different timescales. Ice flow driven by a steep surface slope with several topographic steps leads to high velocities, areas of extension and intense crevassing, which triggers frequent but small calving events independent of local velocity gradients. In contrast, ice flow under smooth surface slopes leaves the ice relatively intact, such that sporadic large-scale calving events dominate, which initiate in areas with high shearing. Flow acceleration caused by enhanced meltwater input and tidal velocity variations were observed for terminus sections close to floatation. Firmly grounded terminus sections showed no tidal signal and a weak short-term reaction to air temperature. These results demonstrate reaction timescales to external forcings from hours to months, which are, however, strongly dependent on local terminus geometry.

Published
2023
Full Text
View/download PDF

34. Inversion of a Stokes glacier flow model emulated by deep learning

Author

Guillaume Jouvet

Subjects
Glacier flow, glacier mechanics, glacier modeling, ground-penetrating radar, ice dynamics, Environmental sciences, GE1-350, Meteorology. Climatology, QC851-999
Abstract

Data assimilation in high-order ice flow modeling is a challenging and computationally costly task, yet crucial to find ice thickness and ice flow parameter distributions that are consistent with ice flow mechanics and mass balance while best matching observations. Failing to find these distributions that are required as initial conditions leads to a disequilibrium between mass balance and ice flow, resulting in nonphysical transient effects in the prognostic model. Here we tackle this problem by inverting an emulator of the Stokes ice flow model based on deep learning. By substituting the ice flow equations using a convolutional neural network emulator, we simplify, make more robust and dramatically speed up the solving of the underlying optimization problem thanks to automatic differentiation, stochastic gradient methods and implementation of graphics processing unit (GPU). We demonstrate this process by simultaneously inferring the ice thickness distribution, ice flow parametrization and ice surface of ten of the largest glaciers in Switzerland. As a result, we obtain a high degree of assimilation while guaranteeing an equilibrium between mass-balance and ice flow mechanics. The code runs very efficiently (optimizing one large-size glacier at 100 m takes < 1 min on a laptop) while it is open-source and publicly available.

Published
2023
Full Text
View/download PDF

35. Adaptation of Autoencoder for Sparsity Reduction From Clinical Notes Representation Learning

Author

Thanh-Dung Le, Rita Noumeir, Jerome Rambaud, Guillaume Sans, and Philippe Jouvet

Subjects
Clinical natural language processing, cardiac failure, autoencoder, sparsity, Computer applications to medicine. Medical informatics, R858-859.7, Medical technology, R855-855.5
Abstract

When dealing with clinical text classification on a small dataset, recent studies have confirmed that a well-tuned multilayer perceptron outperforms other generative classifiers, including deep learning ones. To increase the performance of the neural network classifier, feature selection for the learning representation can effectively be used. However, most feature selection methods only estimate the degree of linear dependency between variables and select the best features based on univariate statistical tests. Furthermore, the sparsity of the feature space involved in the learning representation is ignored. Goal: Our aim is, therefore, to access an alternative approach to tackle the sparsity by compressing the clinical representation feature space, where limited French clinical notes can also be dealt with effectively. Methods: This study proposed an autoencoder learning algorithm to take advantage of sparsity reduction in clinical note representation. The motivation was to determine how to compress sparse, high-dimensional data by reducing the dimension of the clinical note representation feature space. The classification performance of the classifiers was then evaluated in the trained and compressed feature space. Results: The proposed approach provided overall performance gains of up to 3% for each test set evaluation. Finally, the classifier achieved 92% accuracy, 91% recall, 91% precision, and 91% f1-score in detecting the patient’s condition. Furthermore, the compression working mechanism and the autoencoder prediction process were demonstrated by applying the theoretic information bottleneck framework. Clinical and Translational Impact Statement— An autoencoder learning algorithm effectively tackles the problem of sparsity in the representation feature space from a small clinical narrative dataset. Significantly, it can learn the best representation of the training data because of its lossless compression capacity compared to other approaches. Consequently, its downstream classification ability can be significantly improved, which cannot be done using deep learning models.

Published
2023
Full Text
View/download PDF

36. Situation Awareness-Oriented Dashboard in ICUs in Support of Resource Management in Time of Pandemics

Author

Lise Boudreault, Maxence Hebert-Lavoie, Karine Ung, Celia Mahmoudhi, Quynh Phuong Vu, Philippe Jouvet, and Philippe Doyon-Poulin

Subjects
Clinical decision support system, clinical dashboard, critical care, intensive care unit, situation awareness, Computer applications to medicine. Medical informatics, R858-859.7, Medical technology, R855-855.5
Abstract

In a pediatric intensive care unit (PICU) of 32 beds, clinicians manage resources 24 hours a day, 7 days a week, from a large-screen dashboard implemented in 2017. This resource management dashboard efficiently replaces the handwriting information displayed on a whiteboard, offering a synthetic view of the bed’s layout and specific information on staff and equipment at bedside. However, in 2020 when COVID-19 hit, the resource management dashboard showed several limitations. Mainly, its visualization offered to the clinicians limited situation awareness (SA) to perceive, understand and predict the impacts on resource management and decision-making of an unusual flow of patients affected by the most severe form of coronavirus. To identify the SA requirements during a pandemic, we conducted goal-oriented interviews with 11 clinicians working in ICUs. The result is the design of an SA-oriented dashboard with 22 key indicators (KIs): 1 on the admission capacity, 15 at bedside and 6 displayed as statistics in the central area. We conducted a usability evaluation of the SA-oriented dashboard compared to the resource management dashboard with 6 clinicians. The results showed five usability improvements of the SA-oriented dashboard and five limitations. Our work contributes to new knowledge on the clinicians’ SA requirements to support resource management and decision-making in ICUs in times of pandemics.

Published
2023
Full Text
View/download PDF

38. Deep learning speeds up ice flow modelling by several orders of magnitude

Author

Guillaume Jouvet, Guillaume Cordonnier, Byungsoo Kim, Martin Lüthi, Andreas Vieli, and Andy Aschwanden

Subjects
Glacier flow, glacier modelling, ice dynamics, ice velocity, Environmental sciences, GE1-350, Meteorology. Climatology, QC851-999
Abstract

This paper introduces the Instructed Glacier Model (IGM) – a model that simulates ice dynamics, mass balance and its coupling to predict the evolution of glaciers, icefields or ice sheets. The novelty of IGM is that it models the ice flow by a Convolutional Neural Network, which is trained from data generated with hybrid SIA + SSA or Stokes ice flow models. By doing so, the most computationally demanding model component is substituted by a cheap emulator. Once trained with representative data, we demonstrate that IGM permits to model mountain glaciers up to 1000 × faster than Stokes ones on Central Processing Units (CPU) with fidelity levels above 90% in terms of ice flow solutions leading to nearly identical transient thickness evolution. Switching to the GPU often permits additional significant speed-ups, especially when emulating Stokes dynamics or/and modelling at high spatial resolution. IGM is an open-source Python code which deals with two-dimensional (2-D) gridded input and output data. Together with a companion library of trained ice flow emulators, IGM permits user-friendly, highly efficient and mechanically state-of-the-art glacier and icefields simulations.

Published
2022
Full Text
View/download PDF

39. Improving medication safety in a paediatric hospital: a mixed-methods evaluation of a newly implemented computerised provider order entry system

Author

Aude Motulsky, Tibor Schuster, Philippe Jouvet, Denis Lebel, Man Qing Liang, Maxime Thibault, and Marie-Pierre Moreault

Subjects
Computer applications to medicine. Medical informatics, R858-859.7
Abstract

Objectives Computerised provider order entry (CPOE) systems have been implemented around the world as a solution to reduce ordering and transcription errors. However, previous literature documented many challenges to attain this goal, especially in paediatric settings. The objectives of this study were to (1) analyse the impact of a paediatric CPOE system on medication safety and (2) suggest potential error prevention strategies.Methods A pre-post observational study was conducted at the pilot ward (n=60 beds) of a paediatric academic health centre through mixed methods. The implementation project and medication management workflows were described through active participation to the project management team, observation, discussions and analysis of related documents. Furthermore, using incident reports, the nature of each error and error rate was compared between the preperiod and postperiod.Results The global error rate was lower, but non-statistically significant, in the post implementation phase, which was mostly driven by a significant reduction in errors during order acknowledgement, transmission and transcription. Few errors occurred at the prescription step, and most errors occurred during medication administration. Furthermore, some errors could have been prevented using a CPOE in the pre-implementation period, and the CPOE led to few technology-related errors.Discussion and conclusion This study identified both intended and unintended effects of CPOE adoption through the entire medication management workflow. This study revealed the importance of simplifying the acknowledgement, transmission and transcribing steps through the implementation of a CPOE to reduce medication errors. Improving the usability of the electronic medication administration record could help further improve medication safety.

Published
2023
Full Text
View/download PDF

40. Hemodynamic assessment in children after cardiac surgery: A pilot study on the value of infrared thermography

Author

Armelle Bridier, Monisha Shcherbakova, Atsushi Kawaguchi, Nancy Poirier, Carla Said, Rita Noumeir, and Philippe Jouvet

Subjects
infrared thermography, cardiac surgery, pediatric, low cardiac output state, hemodynamic monitoring, Pediatrics, RJ1-570
Abstract

IntroductionLow cardiac output syndrome in the postoperative period after cardiac surgery leads to an increase in tissue oxygen extraction, assessed by the oxygen extraction ratio. Measurement of the oxygen extraction ratio requires blood gases to be taken. However, the temperature of the skin and various parts of the body is a direct result of blood flow distribution and can be monitored using infrared thermography. Thus, we conducted a prospective clinical study to evaluate the correlation between the thermal gradient obtained by infrared thermography and the oxygen extraction ratio in children at risk for low cardiac output after cardiac surgery.MethodsChildren aged 0 to 18 years, having undergone cardiac surgery with cardio-pulmonary bypass in a pediatric intensive care unit were included in the study. One to 4 thermal photos were taken per patient using the FLIR One Pro thermal imaging camera. The thermal gradient between the central temperature of the inner canthus of the eye and the peripheral temperature was compared to the concomitant oxygen extraction ratio calculated from blood gases.Results41 patients were included with a median age of 6 months (IQR 3–48) with median Risk Adjustment for Congenital Heart Surgery-1 score was 2 (IQR 2–3). Eighty nine thermal photos were analyzed. The median thermal gradient was 2.5 °C (IQR 1,01–4.04). The median oxygen extraction ratio was 35% (IQR 26–42). Nine patients had an oxygen extraction ratio ≥ 50%. A significant but weak correlation was found between the thermal gradient and the oxygen extraction ratio (Spearman's test p = 0.25, p = 0.016). Thermal gradient was not correlated with any other clinical or biologic markers of low cardiac output. Only young age was an independent factor associated with an increase in the thermal gradient.ConclusionIn this pilot study, which included mainly children without severe cardiac output decrease, a significant but weak correlation between thermal gradient by infrared thermography and oxygen extraction ratio after pediatric cardiac surgery was observed. Infrared thermography is a promising non-invasive technology that could be included in multimodal monitoring of postoperative cardiac surgery patients. However, a clinical trial including more severe children is needed.

Published
2023
Full Text
View/download PDF

41. Electron--Proton Decoupling in Excited-State Hydrogen Atom Transfer in the Gas Phase

Author

Miyazaki, Mitsuhiko, Ohara, Ryuhei, Daigoku, Kota, Hashimoto, Kenro, Woodward, Jonathan R., Dedonder, Claude, Jouvet, Christophe, and Fujii, Masaaki

Subjects
Physics - Atomic and Molecular Clusters
Abstract

Hydrogen-release by photoexcitation, excited-state- hydrogen-transfer (ESHT), is one of the important photo- chemical processes that occur in aromatic acids and is responsible for photoprotection of biomolecules. The mecha- nism is described by conversion of the initial state to a charge- separated state along the O(N)-H bond elongation, leading to dissociation. Thus ESHT is not a simple H-atom transfer in which a proton and a 1s electron move together. Here we show that the electron-transfer and the proton-motion are decoupled in gas-phase ESHT. We monitor electron and proton transfer independently by picosecond time-resolved near-infrared and infrared spectroscopy for isolated phenol--(ammonia)5, a benchmark molecular cluster. Electron transfer from phenol to ammonia occurred in less than 3 picoseconds, while the overall H-atom transfer took 15 picoseconds. The observed electron-proton decoupling will allow for a deeper understanding and control of of photochemistry in biomolecules

Published
2017
Full Text
View/download PDF

42. Electron-proton transfer mechanism of excited-state hydrogen transfer in phenol--(NH3)n (n = 3 and 5)

Author

Miyazaki, Mitsuhiko, Ohara, Ryuhei, Dedonder, Claude, Jouvet, Christophe, and Fujii, Masaaki

Subjects
Physics - Atomic and Molecular Clusters
Abstract

Excited state hydrogen transfer (ESHT) is responsible to various photochemical processes of aromatics including photoprotection of nuclear basis. Its mechanism is explained by the internal conversion from aromatic $\pi$$\pi$* to $\pi$$\sigma$* states via conical intersection. It means that the electron is transferred to a diffuse Rydberg like $\sigma$* orbital apart from the proton migration. This picture means the electron and the proton are not move together and its dynamics are different in principle. Here, we have applied the picosecond time-resolved near infrared (NIR) and infrared (IR) spectroscopies to the phenol--(NH 3) 5 cluster, the bench mark system of ESHT, and monitored the electron transfer and proton motion independently. The electron transfer monitored by the NIR transition rises within 3 ps while the overall H transfer detected by the IR absorption of NH vibration appears with the lifetime of $\approx$20 ps. It clearly proves that the electron motion and proton migration are decoupled. Such the difference of the time-evolutions between the NIR absorption and the IR transition has not been detected in the cluster with three ammonia molecules. We will report full of our observation together with theoretical calculations of potential energy surfaces of $\pi$$\pi$* and $\pi$$\sigma$* states, and will discuss the ESHT mechanism and its cluster size-dependence between n = 3 and 5. It is suggested that the presence and absence of a barrier in the proton transfer coordinate cause the different dynamics. 2, Comment: Chemistry - A European Journal, Wiley-VCH Verlag, 2017

Published
2017
Full Text
View/download PDF

43. Paediatric acute respiratory distress syndrome incidence and epidemiology (PARDIE): an international, observational study

Author

Khemani, Robinder G, Smith, Lincoln, Lopez-Fernandez, Yolanda M, Kwok, Jeni, Morzov, Rica, Klein, Margaret J, Yehya, Nadir, Willson, Douglas, Kneyber, Martin CJ, Lillie, Jon, Fernandez, Analia, Newth, Christopher JL, Jouvet, Philippe, Thomas, Neal J, Investigators, Pediatric Acute Respiratory Distress syndrome Incidence and Epidemiology, Abaleke, Eugenia, Ackerman, Kate G, Acuña, Carlos, Adu-Darko, Michelle, Affolter, Jeremy T, Agbeko, Rachel, Amoudi, Ahmed Al, Alahmadti, Ahmad, Aldairi, Nedaa, Alibrahim, Omar, Allen, Kiona, Allen, Christine, Al-Subu, Awni, Althabe, María, Alvear, Jimena, Anil, Ayse Berna, Anthony, Heather, Aramburo, Angela, Villanueva, David Arjona, Ashtari, Neda, Vera, Antonio Ávila, Baines, Paul, Bales, Melissa, Barr, Samantha, Barry, Dana, Baudin, Florent, Beca, John, Belfield, Holly, Beltramo, Fernando, Benken, Laura, Bhalla, Anoopindar, Blom, Andrea, Botta, Priscila, Bourgoin, Pierre, Brezmes, Marta, Briassoulis, George, Bridier, Armelle, Brierley, Joe, Sanagustin, Sonia Brio, Broden, Elizabeth, Butt, Warwick, Bysani, Kris, Camilo, Cristina, Camporesi, Anna, Campos-Miño, Santiago, Can, Fulya Kamit, Capocasa, Patricia, Caro, Daniel, Carroll, Christopher, Castellani, Pablo, Castillo, Andres E, Chen, Yang, Chima, Ranjit S, Chiusolo, Fabrizio, Cinquegrani, Karina, Coates, Bria, Coronado-Munoz, Alvaro, Cortéz, Ambar, Romero, Pablo Cruces, Cullimore, Melissa, Cvijanovich, Natalie, Dahmer, Mary K, Deep, Akash, Delzoppo, Carmel, Di Nardo, Matteo, Díaz, Franco, Dijkstra, Sandra, Dockery, W Keith, Dominguez, Troy E, Dumitrascu, Mariana, Dursun, Oguz, Dwarakanathan, Buvana, Elghuwael, Ismail, Emeriaud, Guillaume, Erickson, Simon, Español, Segundo Fernando, Estil, Jim Brian, Feather, Calandra, Feinstein, Yael, Fernández, Analía, Ferreyra, Marcela, Flori, Heidi, Fortini, Yanina Vanesa, Fortune, Peter-Marc, and French, Mary Ellen

Subjects
Biomedical and Clinical Sciences, Cardiovascular Medicine and Haematology, Clinical Sciences, Pediatric, Lung, Clinical Research, 4.1 Discovery and preclinical testing of markers and technologies, Detection, screening and diagnosis, Respiratory, Good Health and Well Being, Acute Lung Injury, Age Factors, Area Under Curve, Cause of Death, Child, Child, Preschool, Combined Modality Therapy, Cross-Sectional Studies, Female, Hospital Mortality, Humans, Intensive Care Units, Pediatric, Internationality, Kaplan-Meier Estimate, Male, Prognosis, Prospective Studies, Respiratory Distress Syndrome, Severity of Illness Index, Sex Factors, Survival Analysis, United States, Pediatric Acute Respiratory Distress syndrome Incidence and Epidemiology (PARDIE) Investigators, Pediatric Acute Lung Injury and Sepsis Investigators (PALISI) Network, Public Health and Health Services, Other Medical and Health Sciences, Cardiovascular medicine and haematology, Clinical sciences
Abstract

BackgroundPaediatric acute respiratory distress syndrome (PARDS) is associated with high mortality in children, but until recently no paediatric-specific diagnostic criteria existed. The Pediatric Acute Lung Injury Consensus Conference (PALICC) definition was developed to overcome limitations of the Berlin definition, which was designed and validated for adults. We aimed to determine the incidence and outcomes of children who meet the PALICC definition of PARDS.MethodsIn this international, prospective, cross-sectional, observational study, 145 paediatric intensive care units (PICUs) from 27 countries were recruited, and over a continuous 5 day period across 10 weeks all patients were screened for enrolment. Patients were included if they had a new diagnosis of PARDS that met PALICC criteria during the study week. Exclusion criteria included meeting PARDS criteria more than 24 h before screening, cyanotic heart disease, active perinatal lung disease, and preparation or recovery from a cardiac intervention. Data were collected on the PICU characteristics, patient demographics, and elements of PARDS (ie, PARDS risk factors, hypoxaemia severity metrics, type of ventilation), comorbidities, chest imaging, arterial blood gas measurements, and pulse oximetry. The primary outcome was PICU mortality. Secondary outcomes included 90 day mortality, duration of invasive mechanical and non-invasive ventilation, and cause of death.FindingsBetween May 9, 2016, and June 16, 2017, during the 10 study weeks, 23 280 patients were admitted to participating PICUs, of whom 744 (3·2%) were identified as having PARDS. 95% (708 of 744) of patients had complete data for analysis, with 17% (121 of 708; 95% CI 14-20) mortality, whereas only 32% (230 of 708) of patients met Berlin criteria with 27% (61 of 230) mortality. Based on hypoxaemia severity at PARDS diagnosis, mortality was similar among those who were non-invasively ventilated and with mild or moderate PARDS (10-15%), but higher for those with severe PARDS (33% [54 of 165; 95% CI 26-41]). 50% (80 of 160) of non-invasively ventilated patients with PARDS were subsequently intubated, with 25% (20 of 80; 95% CI 16-36) mortality. By use of PALICC PARDS definition, severity of PARDS at 6 h after initial diagnosis (area under the curve [AUC] 0·69, 95% CI 0·62-0·76) discriminates PICU mortality better than severity at PARDS diagnosis (AUC 0·64, 0·58-0·71), and outperforms Berlin severity groups at 6 h (0·64, 0·58-0·70; p=0·01).InterpretationThe PALICC definition identified more children as having PARDS than the Berlin definition, and PALICC PARDS severity groupings improved the stratification of mortality risk, particularly when applied 6 h after PARDS diagnosis. The PALICC PARDS framework should be considered for use in future epidemiological and therapeutic research among children with PARDS.FundingUniversity of Southern California Clinical Translational Science Institute, Sainte Justine Children's Hospital, University of Montreal, Canada, Réseau en Santé Respiratoire du Fonds de Recherche Quebec-Santé, and Children's Hospital Los Angeles, Department of Anesthesiology and Critical Care Medicine.

Published
2019

44. Antimicrobial treatment duration for uncomplicated bloodstream infections in critically ill children: a multicentre observational study

Author

Sandra Pong, Robert A. Fowler, Srinivas Murthy, Jeffrey M. Pernica, Elaine Gilfoyle, Patricia Fontela, Asgar H. Rishu, Nicholas Mitsakakis, James S. Hutchison, Michelle Science, Winnie Seto, Philippe Jouvet, Nick Daneman, and on behalf of the Canadian Critical Care Trials Group

Subjects
Antibacterial agents, Bacteremia, Critical care, Critical illness, Duration of therapy, Pediatric, Pediatrics, RJ1-570
Abstract

Abstract Background Bloodstream infections (BSIs) cause significant morbidity and mortality in critically ill children but treatment duration is understudied. We describe the durations of antimicrobial treatment that critically ill children receive and explore factors associated with treatment duration. Methods We conducted a retrospective observational cohort study in six pediatric intensive care units (PICUs) across Canada. Associations between treatment duration and patient-, infection- and pathogen-related characteristics were explored using multivariable regression analyses. Results Among 187 critically ill children with BSIs, the median duration of antimicrobial treatment was 15 (IQR 11–25) days. Median treatment durations were longer than two weeks for all subjects with known sources of infection: catheter-related 16 (IQR 11–24), respiratory 15 (IQR 11–26), intra-abdominal 20 (IQR 14–26), skin/soft tissue 17 (IQR 15–33), urinary 17 (IQR 15–35), central nervous system 33 (IQR 15–46) and other sources 29.5 (IQR 15–55) days. When sources of infection were unclear, the median duration was 13 (IQR 10–16) days. Treatment durations varied widely within and across PICUs. In multivariable linear regression, longer treatment durations were associated with severity of illness (+ 0.4 days longer [95% confidence interval (CI), 0.1 to 0.7, p = 0.007] per unit increase in PRISM-IV) and central nervous system infection (+ 17 days [95% CI, 6.7 to 27.4], p = 0.001). Age and pathogen type were not associated with treatment duration. Conclusions Most critically ill children with BSIs received at least two weeks of antimicrobial treatment. Further study is needed to determine whether shorter duration therapy would be effective for selected critically ill children.

Published
2022
Full Text
View/download PDF

47. Detecting of a Patient's Condition From Clinical Narratives Using Natural Language Representation

Author

Thanh-Dung Le, Rita Noumeir, Jerome Rambaud, Guillaume Sans, and Philippe Jouvet

Subjects
Clinical natural language processing, cardiac failure, machine learning, imbalance learning, feature selection, Computer applications to medicine. Medical informatics, R858-859.7, Medical technology, R855-855.5
Abstract

The rapid progress in clinical data management systems and artificial intelligence approaches enable the era of personalized medicine. Intensive care units (ICUs) are ideal clinical research environments for such development because they collect many clinical data and are highly computerized. Goal: We designed a retrospective clinical study on a prospective ICU database using clinical natural language to help in the early diagnosis of heart failure in critically ill children. Methods: The methodology consisted of empirical experiments of a learning algorithm to learn the hidden interpretation and presentation of the French clinical note data. This study included 1386 patients' clinical notes with 5444 single lines of notes. There were 1941 positive cases (36% of total) and 3503 negative cases classified by two independent physicians using a standardized approach. Results: The multilayer perceptron neural network outperforms other discriminative and generative classifiers. Consequently, the proposed framework yields an overall classification performance with 89% accuracy, 88% recall, and 89% precision. Conclusions: This study successfully applied learning representation and machine learning algorithms to detect heart failure in a single French institution from clinical natural language. Further work is needed to use the same methodology in other languages and institutions.

Published
2022
Full Text
View/download PDF

49. Efficient photochemistry of coronene: water complexes

Author

Noble, J. A., Jouvet, C., Aupetit, C., Moudens, A., and Mascetti, J.

Subjects
Astrophysics - Astrophysics of Galaxies, Astrophysics - Earth and Planetary Astrophysics
Abstract

The photochemistry of ices with polycyclic aromatic hydrocarbons (PAHs) has been extensively studied, but to date no investigation has been made of PAHs in interaction with low numbers (n < 4) of molecules of water. We performed photochemical matrix isolation studies of coronene:water complexes, probing the argon matrix with FTIR spectroscopy. We find that coronene readily reacts with water upon irradiation with a mercury vapour lamp to produce oxygenated PAH photoproducts, and we postulate a reaction mechanism via a charge transfer Rydberg state. This result suggests that oxygenated PAHs should be widely observed in regions of the ISM with sufficiently high water abundances, for example near the edges of molecular clouds where water molecules begin to form, but before icy layers are observed, that is at AV <3. In order to explain the low derived observational abundances of oxygenated PAHs, additional destruction routes must be invoked., Comment: 5 pages, 5 figures, accepted for publication in A&A

Published
2016
Full Text
View/download PDF

50. Hydrogen bonds vs. $\pi$-stacking interactions in the p-aminophenol...p-cresol dimer: an experimental and theoretical study

Author

Capello, Carolina, Hernández, Federico, Broquier, Michel, Dedonder-Lardeux, Claude, Jouvet, Christophe, and Pino, Gustavo

Subjects
Physics - Atomic and Molecular Clusters
Abstract

The gas phase structure and excited state lifetime of the p-aminophenol...p-cresol heterodimer have been investigated by REMPI and LIF spectroscopy with nanosecond laser pulses and pump-probe experiments with picosecond laser pulses as a model system to study the competition between p-p and H-bonding interactions in aromatic dimers. The excitation is a broad and unstructured band. The excitedstate of the heterodimer is long lived (2.5 +/- 0.5) ns with a very broad fluorescence spectrum red-shifted by 4000 cm^{-1} with respect to the excitation spectrum. Calculations at the MP2/RI-CC2 and DFT-oB97X-D levels indicate that hydrogen-bonded (HB) and p-stacked isomers are almost isoenergetic in the ground state while in the excited state only the p-stacked isomer exists. This suggests that the HB isomer cannot be excited due to negligible Franck-Condon factors and therefore the excitation spectrum is associated with the p-stacked isomer that reaches vibrationally excited states in the S1 state upon vertical excitation. The excited state structure is an exciplex responsible for the fluorescence of the complex. Finally,a comparison was performed between the p-stacked structure observed for the p-aminophenol...p-cresol heterodimer and the HB structure reported for the (p-cresol)2 homodimer indicating that the differences are due to different optical properties (oscillator strengths and Franck-Condon factors) of the isomers of both dimers and not to the interactions involved in the ground state

Published
2016
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results