De-identification of French medical narratives

Gaudet-Blavignac, Christophe; Foufi, Vasiliki; Wehrli, Eric; Lovis, Christian

doi:10.4414/smi.34.00417

Cited by 4 publications

(4 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It also obtained good performance in terms of specificity and precision. This confirms the assumption that hybrid systems, combining advantages, may outperform other methods (Grouin and Névéol 2014;Névéol et al 2018;Gaudet-Blavignac et al 2018). These results will be instrumental in protecting personal data in secondary use of health data.…”

Section: Discussionsupporting

confidence: 80%

“…The self-attention mechanism has been newly applied to deidentification (Yang et al 2019;Fu et al 2020). Rule-based de-identification methods for clinical narratives previously developed in English have been adapted to French (Grouin and Névéol 2014;Névéol et al 2018;Gaudet-Blavignac et al 2018). Some of these works presume that hybrid systems would improve performance.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

De-identification of Emergency Medical Records in French: Survey and Comparison of State-of-the-Art Automated Systems

Bourdois

Avalos

Chenais

et al. 2021

FLAIRS

View full text Add to dashboard Cite

In France, structured data from emergency room (ER) visits are aggregated at the national level to build a syndromic surveillance system for several health events. For visits motivated by a traumatic event, information on the causes are stored in free-text clinical notes. To exploit these data, an automated de-identification system guaranteeing protection of privacy is required.In this study we review available de-identification tools to de-identify free-text clinical documents in French. A key point is how to overcome the resource barrier that hampers NLP applications in languages other than English. We compare rule-based, named entity recognition, new Transformer-based deep learning and hybrid systems using, when required, a fine-tuning set of 30,000 unlabeled clinical notes. The evaluation is performed on a test set of 3,000 manually annotated notes.Hybrid systems, combining capabilities in complementary tasks, show the best performance. This work is a first step in the foundation of a national surveillance system based on the exhaustive collection of ER visits reports for automated trauma monitoring.

show abstract

Section: Discussionsupporting

confidence: 80%

Section: Introductionmentioning

confidence: 99%

De-identification of Emergency Medical Records in French: Survey and Comparison of State-of-the-Art Automated Systems

Bourdois

Avalos

Chenais

et al. 2021

FLAIRS

View full text Add to dashboard Cite

show abstract

“…In both cases it was necessary to request EHRs from Dutch institutes, which are often not publicly available. In French, both [60] and [61] explored the possibilities of rule-based and CRF-based systems on data sets built by retrieving EHRs from French hospitals. In German, first [62] and then [63] developed rule-based techniques and machine learning, but they remain proof-of-concepts due to the lack of extensive data training.…”

Section: Clinical De-identification For Specific Languagesmentioning

confidence: 99%

A Novel COVID-19 Data Set and an Effective Deep Learning Approach for the De-Identification of Italian Medical Records

et al. 2021

View full text Add to dashboard Cite

In the last years, the need to de-identify privacy-sensitive information within Electronic Health Records (EHRs) has become increasingly felt and extremely relevant to encourage the sharing and publication of their content in accordance with the restrictions imposed by both national and supranational privacy authorities. In the field of Natural Language Processing (NLP), several deep learning techniques for Named Entity Recognition (NER) have been applied to face this issue, significantly improving the effectiveness in identifying sensitive information in EHRs written in English. However, the lack of data sets in other languages has strongly limited their applicability and performance evaluation. To this aim, a new de-identification data set in Italian has been developed in this work, starting from the 115 COVID-19 EHRs provided by the Italian Society of Radiology (SIRM): 65 were used for training and development, the remaining 50 were used for testing. The data set was labelled following the guidelines of the i2b2 2014 de-identification track. As additional contribution, combined with the best performing Bi-LSTM + CRF sequence labeling architecture, a stacked word representation form, not yet experimented for the Italian clinical de-identification scenario, has been tested, based both on a contextualized linguistic model to manage word polysemy and its morpho-syntactic variations and on sub-word embeddings to better capture latent syntactic and semantic similarities. Finally, other cutting-edge approaches were compared with the proposed model, which achieved the best performance highlighting the goodness of the promoted approach. INDEX TERMS clinical de-identification, contextualized embedding, deep learning, Italian language, named entity recognition

show abstract

“…Here, existing compliant data management platforms with dynamic consent management [176] or personal data cooperatives [177] could be used. Furthermore, novel deidentification approaches could help implement this NFR important for the patients [178].…”

Section: Privacymentioning

confidence: 99%

Factors Influencing Motivation and Engagement in Mobile Health Among Patients With Sickle Cell Disease in Low-Prevalence, High-Income Countries: Qualitative Exploration of Patient Requirements

Issom¹,

Henriksen²,

Woldaregay³

et al. 2020

JMIR Hum Factors

Self Cite

View full text Add to dashboard Cite

Background Sickle cell disease (SCD) is a hematological genetic disease affecting over 25 million people worldwide. The main clinical manifestations of SCD, hemolytic anemia and vaso-occlusion, lead to chronic pain and organ damages. With recent advances in childhood care, high-income countries have seen SCD drift from a disease of early childhood mortality to a neglected chronic disease of adulthood. In particular, coordinated, preventive, and comprehensive care for adults with SCD is largely underresourced. Consequently, patients are left to self-manage. Mobile health (mHealth) apps for chronic disease self-management are now flooding app stores. However, evidence remains unclear about their effectiveness, and the literature indicates low user engagement and poor adoption rates. Finally, few apps have been developed for people with SCD and none encompasses their numerous and complex self-care management needs. Objective This study aimed to identify factors that may influence the long-term engagement and user adoption of mHealth among the particularly isolated community of adult patients with SCD living in low-prevalence, high-income countries. Methods Semistructured interviews were conducted. Interviews were audiotaped, transcribed verbatim, and analyzed using thematic analysis. Analysis was informed by the Braun and Clarke framework and mapped to the COM-B model (capability, opportunity, motivation, and behavior). Results were classified into high-level functional requirements (FRs) and nonfunctional requirements (NFRs) to guide the development of future mHealth interventions. Results Overall, 6 males and 4 females were interviewed (aged between 21 and 55 years). Thirty FRs and 31 NFRs were extracted from the analysis. Most participants (8/10) were concerned about increasing their physical capabilities being able to stop pain symptoms quickly. Regarding the psychological capability aspects, all interviewees desired to receive trustworthy feedback on their self-care management practices. About their physical opportunities, most (7/10) expressed a strong desire to receive alerts when they would reach their own physiological limitations (ie, during physical activity). Concerning social opportunity, most (9/10) reported wanting to learn about the self-care practices of other patients. Relating to motivational aspects, many interviewees (6/10) stressed their need to learn how to avoid the symptoms and live as normal a life as possible. Finally, NFRs included inconspicuousness and customizability of user experience, automatic data collection, data shareability, and data privacy. Conclusions Our findings suggest that motivation and engagement with mHealth technologies among the studied population could be increased by providing features that clearly benefit them. Self-management support and self-care decision aid are patients’ major demands. As the complexity of SCD self-management requires a high cognitive load, pervasive health technologies such as wearable sensors, implantable devices, or inconspicuous conversational user interfaces should be explored to ease it. Some of the required technologies already exist but must be integrated, bundled, adapted, or improved to meet the specific needs of people with SCD.

show abstract

De-identification of French medical narratives

Cited by 4 publications

References 7 publications

De-identification of Emergency Medical Records in French: Survey and Comparison of State-of-the-Art Automated Systems

De-identification of Emergency Medical Records in French: Survey and Comparison of State-of-the-Art Automated Systems

A Novel COVID-19 Data Set and an Effective Deep Learning Approach for the De-Identification of Italian Medical Records

Factors Influencing Motivation and Engagement in Mobile Health Among Patients With Sickle Cell Disease in Low-Prevalence, High-Income Countries: Qualitative Exploration of Patient Requirements

Contact Info

Product

Resources

About