Search

International audience; Unstructured data in electronic health records, represented by clinical texts, are a vast source of healthcare information because they describe a patient's journey, including clinical findings, procedures, and information about the continuity of care. The publication of several studies on temporal relation extraction from clinical texts during the last decade and the realization of multiple shared tasks highlight the importance of this research theme. Therefore, we propose a review of temporal relation extraction in clinical texts. We analyzed 105 articles and verified that relations between events and document creation time, a coarse temporality type, were addressed with traditional machine learning–based models with few recent initiatives to push the state-of-the-art with deep learning–based models. For temporal relations between entities (event and temporal expressions) in the document, factors such as dataset imbalance because of candidate pair generation and task complexity directly affect the system's performance. The state-of-the-art resides on attention-based models, with contextualized word representations being fine-tuned for temporal relation extraction. However, further experiments and advances in the research topic are required until real-time clinical domain applications are released. Furthermore, most of the publications mainly reside on the same dataset, hindering the need for new annotation projects that provide datasets for different medical specialties, clinical text types, and even languages.

International audience; Automatic detection of negated content is often a prerequisite in information extraction systems, especially in the biomedical domain. This paper proposes our contributions on negation scope detection in French and Brazilian Portuguese. We introduce two corpora mainly built with excerpts from clinical trial protocols, describing the inclusion criteria of patients. The corpora have been manually annotated for marking up the negation cues and their scope. Then, we propose a recurrent neural networks approach to acquire the scopes. MOTS-CLÉS : négation, français, portugais brésilien, apprentissage supervisé, réseau de neurones.; La détection automatique de la négation fait souvent partie des pré-requis dans les systèmes d'extraction d'information, notamment dans le domaine biomédical. Cet article présente nos contributions concernant la détection de la portée de la négation en français et portugais brésilien. Nous présentons d'une part deux corpus principalement constitués d'extraits de protocoles d'essais cliniques en français et portugais brésilien, dédiés aux critères d'inclusion de patients. Les marqueurs de négation et leurs portées y ont été annotés manuellement. Nous présentons d'autre part une approche par réseau de neurones récurrents pour extraire les portées.

Introduction Based on the increasing usability of technology in healthcare, this paper discusses the use of an expert system (ES) to identify the sensory profile of patients starting Occupational Therapy, allowing the professional to make assertive decisions in establishing priorities in the therapeutic plan.Objective To develop a decision support system from the Infant/Toddler Sensory Profile.Method Structuring of an ES based on Infant/Toddler Sensory Profile, from terms translation into Portuguese, identification of variables and domain values involved, and construction of production rules.Results Twelve variables were registered for the construction of the ES, 6 of these were treated as goal-variables, 20 rules being built.Conclusion This ES is an important support to the occupational therapist in the decision-making process of treatment plans, determining priorities and respecting the sensory profile of each child. In addition, it must be noted that there is no equivalent system. Introdução Com a crescente usabilidade da tecnologia na área da saúde, este artigo aborda a utilização de um sistema especialista (SE) para identificar o perfil sensorial de pacientes a iniciarem o tratamento de Terapia Ocupacional, permitindo ao profissional tomar decisões assertivas no estabelecimento de prioridades no plano terapêutico.Objetivo Construir um sistema de apoio à decisão a partir do Infant/Toddler Sensory Profile.Método Estruturação de um SE baseado no Infant/Toddler Sensory Profile, a partir da tradução para o português dos termos contidos neste instrumento, identificação das variáveis e valores de domínio envolvidos; e a construção das respectivas regras de produção.Resultados Para a construção do SE foram cadastradas 12 variáveis, destas 6 foram tratadas como variáveis-objetivo, sendo construídas 20 regras.Conclusão O SE construído constitui apoio importante ao terapeuta ocupacional no processo de tomada de decisão sobre o plano terapêutico, determinando as prioridades e respeitando o perfil sensorial de cada criança. Além disso, é preciso salientar que não há um sistema equivalente.

International audience; Earthquake-triggered landslides have an increasing disastrous impact in seismic regions due to the fast growing urbanization and infrastructures. Just considering disasters from the last fifteen years, among which the 1999 Chi-Chi earthquake, the 2008 Wenchuan earthquake, and the 2011 Tohoku earthquake, these events generated tens of thousands of coseismic landslides. Those resulted in amazing death toll and considerable damages, affecting the regional landscape including its hydrological main features. Despite a strong impetus in research during past decades, knowledge on those geohazards is still fragmentary, while databases of high quality observational data are lacking. These phenomena call for further collaborative researches aiming eventually to enhance preparedness and crisis management. As one of the three SUPERSITE concept FP7 projects dealing with long term high level monitoring of major natural hazards at the European level, the MARSITE project gathers research groups in a comprehensive monitoring activity developed in the Sea of Marmara Region, one of the most densely populated parts of Europe and rated at high seismic risk level since the 1999 Izmit and Duzce devastating earthquakes. Besides the seismic threat, landslides in Turkey and in this region constitute an important source of loss. The 1999 Earthquake caused extensive landslides while tsunami effects were observed during the post-event surveys in several places along the coasts of the Izmit bay. The 6th Work Package of MARSITE project gathers 9 research groups to study earthquake-induced landslides focusing on two sub-regional areas of high interest. First, the Cekmece-Avcilar peninsula, located westwards of Istanbul, is a highly urbanized concentrated landslide prone area, showing high susceptibility to both rainfalls while affected by very significant seismic site effects. Second, the off-shore entrance of the Izmit Gulf, close to the termination of the surface rupture of the 1999 earthquake, that shows an important slump mass facing the Istanbul coastline. A multidisciplinary research program based on pre-existing studies has been designed with objectives and tasks linked to constrain and tackle progressively some challenging issues related to data integration, modeling, monitoring and mapping technologies. Concerning the on-shore area, this program includes the refined analysis of the seismic site response, the permanent multi-parameter ground monitoring of a representative unstable slope as well as the in-depth slope stability analysis based on the stress-strain dynamic numerical modelling approach. Hyperspectral and Dinsar imagery technologies are also deployed to complete inventory and observational information. The development of a dynamic GIS tool featuring capabilities to integrate and process very different types of data, and up-date susceptibility maps based on near to real-time rainfall-seismic shaking input, is currently undertaken. Moreover, the research is gaining high profit of a vast drilling program undertaken by the Istanbul Metropolitan Area, aiming to yield a detailed geological and geotechnical characterization of the slopes. Also included in the objectives is to test a landslide early warning system. As regards the selected off-shore area, high resolution geophysical marine surveys are being conducted to complete its geomorphological description to help in mapping possible incipient mass movements. This is especially expected to provide better-constrained input for both laboratory testing and numerical modeling of tsunami scenarios thank to a unique lab-scale tsunami channel.

O Programa de Aprendizagem de Ginecologia e Obstetrícia da PUC-PR adota a aprendizagem baseada em problemas como metodologia de aprendizagem. Neste programa, são executados diversos casos clínicos que compõem seu conteúdo. Para cada caso clínico, os professores definem uma semiologia principal e as respectivas semiologias secundárias, a fim de definir o conjunto de ações ou procedimentos que os alunos devem executar. Visando à inserção das tecnologias da informação e comunicação no aprendizado dos alunos, foi desenvolvido um sistema que permite obter, passo a passo, as ações e procedimentos executados pelos alunos. O objetivo geral deste trabalho é conceber um modelo matemático e computacional que permita obter índices de similaridade entre as semiologias definidas pelos professores e as executadas pelos alunos. The Undergraduate Course in Obstetrics and Gynecology at the Catholic University in Paraná State, Brazil, is based on problem-based learning. The program analyzes various clinical cases to form the course content. For each clinical case, the professors define the primary patient work-up protocol and the corresponding clinical/diagnostic techniques, in order to establish the set of measures or procedures that students are expected to perform. In order to include information and communication technologies in the students' learning process, developed a system that allows the students to follow a step-by-step approach to these measures and procedures. The overall objective of the current article was to design a mathematical and computational model that allows obtaining indices of similarity between the patient work-up approaches defined by the professors and those actually performed by students.