Teaching Module to Train Students in the Participatory Design of Digital Technologies for Disability Assistance.

Teaching the principles of participatory design to students mainly interested in digital skills is important because user-centered approaches have become essential in the field of new technologies when we want to guarantee that a product will meet the needs of its end-users. Working with technologies dedicated to the disability assistance was considered to be the right application domain. In 2021, ESIEE Paris, a school training students to become engineers with digital skills, created and opened a new teaching module to learn how to follow the principles of participatory design to improve the quality of a project. This paper describes the organization and the conclusions of this experience after two editions of the module.

Design and Implementation of a Unique Patient Identification Model in Information Systems in Burkina Faso.

The implementation of a reliable identity process is the basis of any secure patient information sharing system. Indeed, each individual is unique and should be identified by a unique number (identifier). It is with these issues in mind that we have designed and implemented a unique patient identification method adapted to the context of Burkina Faso. The recommended method is inspired by the French method based on the work of the Group for the Modernization of the Hospital Information System (GMSIH) [1]. The developed model allows to assign a "Unique Identifier" (PatientID) to each patient from his profile of identification features (name, date of birth, gender,…). The patient ID is a sequence of 20 characters plus a security "key" of 2 characters. A reliability test of the model has been performed to take into account identity anomalies (duplicate, collision).

Ambulatory Peritoneal Dialysis Analysis Framework.

This paper presents the design of an autonomous tracking device to enhance understanding of ambulatory peritoneal dialysis. The resulting tool aims to serve as a framework for research analysis and a decision support for treatment adjustments in peritoneal dialysis.

Patient's Unique Identifier for Efficient and Secure Monitoring of Pregnant Women in Burkina Faso.

The objective of this work is to set up a device allowing to identify the pregnant woman in a univocal and reliable way during her pregnancy follow-up. This study is a continuation of a project to improve the electronic monitoring of pregnancy in pregnant women in Burkina Faso. The methodology is based on the scientific work of the GMSIH of France (1). The work has lead to the design and implementation of a model that allows to assign a "Unique Identifier" to each pregnant woman from her first prenatal visit. The Patient ID is developed from the person's identification trait profile. It consists of a sequence of 20 characters and a security "key" of 2 characters. After the design, a reliability test of the model was performed to take into account identity anomalies (duplicates, collisions).

Automatic Cyclic Alternating Pattern (CAP) analysis: Local and multi-trace approaches.

The Cyclic Alternating Pattern (CAP) is composed of cycles of two different electroencephalographic features: an activation A-phase followed by a B-phase representing the background activity. CAP is considered a physiological marker of sleep instability. Despite its informative nature, the clinical applications remain limited as CAP analysis is a time-consuming activity. In order to overcome this limit, several automatic detection methods were recently developed. In this paper, two new dimensions were investigated in the attempt to optimize novel, efficient and automatic detection algorithms: 1) many electroencephalographic leads were compared to identify the best local performance, and 2) the global contribution of the concurrent detection across several derivations to CAP identification. The developed algorithms were tested on 41 polysomnographic recordings from normal (n = 8) and pathological (n = 33) subjects. In comparison with the visual CAP analysis as the gold standard, the performance of each algorithm was evaluated. Locally, the detection on the F4-C4 derivation showed the best performance in comparison with all other leads, providing practical suggestions of electrode montage when a lean and minimally invasive approach is preferable. A further improvement in the detection was achieved by a multi-trace method, the Global Analysis-Common Events, to be applied when several recording derivations are available. Moreover, CAP time and CAP rate obtained with these algorithms positively correlated with the ones identified by the scorer. These preliminary findings support efficient automated ways for the evaluation of the sleep instability, generalizable to both normal and pathological subjects affected by different sleep disorders.

A Web Interface for Antibiotic Prescription Recommendations in Primary Care: User-Centered Design Approach.

BACKGROUND: Antibiotic misuse is a serious public health problem worldwide. National health authorities release clinical practice guidelines (CPGs) to guide general practitioners (GPs) in their choice of antibiotics. However, despite the large-scale dissemination of CPGs, GPs continue to prescribe antibiotics that are not recommended as first-line treatments. This nonadherence to recommendations may be due to GPs misunderstanding the CPGs. A web interface displaying antibiotic prescription recommendations and their justifications could help to improve the comprehensibility and readability of CPGs, thereby increasing the adoption of recommendations regarding antibiotic treatment. OBJECTIVE: This study aims to design and evaluate a web interface for antibiotic prescription displaying both the recommended antibiotics and their justifications in the form of antibiotic properties. METHODS: A web interface was designed according to the same principles as e-commerce interfaces and was assessed by 117 GPs. These GPs were asked to answer 17 questions relating to the usefulness, user-friendliness, and comprehensibility and readability of the interface, and their satisfaction with it. Responses were recorded on a 4-point Likert scale (ranging from "absolutely disagree" to "absolutely agree"). At the end of the evaluation, the GPs were allowed to provide optional, additional free comments. RESULTS: The antibiotic prescription web interface consists of three main sections: a clinical summary section, a filter section, and a recommended antibiotics section. The majority of GPs appreciated the clinical summary (90/117, 76.9%) and filter (98/117, 83.8%) sections, whereas 48.7% (57/117) of them reported difficulty reading some of the icons in the recommended antibiotics section. Overall, 82.9% (97/117) of GPs found the display of drug properties useful, and 65.8% (77/117) reported that the web interface improved their understanding of CPG recommendations. CONCLUSIONS: The web interface displaying antibiotic recommendations and their properties can help doctors understand the rationale underlying CPG recommendations regarding antibiotic treatment, but further improvements are required before its implementation into a clinical decision support system.

Bank of Digital Filters for Identification of Combined Drug Products in a French Public Database.

Drugs information systems, prescription support softwares, and drug decision support systems need to reason on drug properties. Combined pharmaceutical products need to be considered specifically because they may require a specific processing. Hence, they also need to be identified to automate the population of databases with up-to-date property values. We defined a set of digital filters designed for the identification of antibiotics in a public database. Four different filters are proposed, to be combined to extract the relevant information. Evaluation was conducted to combine filters and retrieve information about rand combined antibiotics with success. However, information provided in the structured files of the French drug database is limited; information provided in the HTML files suffers from a lack of quality. Hence, reuse of this data and this information should be performed very cautiously.

Data Fusion to Convert Drug Consumption Quantities into Defined Daily Doses.

Analysis of consumptions has proved a misuse of antibiotics, despite the existence of national requirements. To be able to compute drugs consumption quantities of highly heterogeneous drugs expressed in various doses units, the World Health Organization has defined a defined daily dose. A methodology has been also defined from previous work to compute manually the drugs consumptions in daily defined dose. We automated this methodology by using data fusion on data retrieved from different sources including a French public database and the World Health Organization website. Evaluation proved the efficiency of the approach, except for inconsistency cases. We identified these cases and proposed a solution to avoid them.

Automatic Sleep Stages Classification Combining Semantic Representation and Dynamic Expert System.

Interest in sleep has been growing in the last decades, considering its benefits for well-being, but also to diagnose sleep troubles. The gold standard to monitor sleep consists of recording the course of many physiological parameters during a whole night. The human interpretation of resulting curves is time consuming. We propose an automatic knowledge-based decision system to support sleep staging. This system handles temporal data, such as events, to combine and aggregate atomic data, so as to obtain high-abstraction-levels contextual decisions. The proposed system relies on a semantic reprentation of observations, and on contextual knowledge base obtained by formalizing clinical practice guidelines. Evaluated on a dataset composed of 131 full night polysomnographies, results are encouraging, but point out that further knowledge need to be integrated.

Modular Knowledge-Based Decision Support System Dedicated to a Cooperative Decision to Prevent Cardiovascular Diseases.

Despite the success of artificial intelligence solutions in the recent years, physicians are still reticent to use integrated functionalities to support their decision. Methods used to create these functionalities can be divided into two groups, each being associated to different questions. Data-based methods are seen as black boxes for which it is impossible to understand how the decision is taken; knowledge-based methods need to rely on formalized knowledge sources on the basis of evidence, which can be discussed and criticized by physicians for their use in real life. This paper presents a new modular decision support system for the prevention of cardiovascular diseases, based on knowledge and on cooperative decision between the patient and the physician. The decision support system is based on two layers: (i) the first layer is a knowledge-based module which generates automatically patient profile, and prevention strategies associated to the profile; (ii) the second layer is a dynamic collaborative graphic user interface which displayed information about the risks of treatment adherence failure, personalized motivation and follow-up strategies. In the future, we aim at assessing the platform in real life.

Building a Knowledge-Based Tool for Auto-Assessing the Cardiovascular Risk.

The prevention of cardiovascular diseases needs first to quantify the cardiovascular risk. To estimate this risk, French national health authorities provided clinical practice guidelines extending the existing European SCORE, which doesn't include all the cardiovascular risk factors (e.g. diabetes). Hence, French national clinical practice guidelines to quantify the cardiovascular risk is able to deal with more clinical situations than the SCORE. The goal of this paper is to formalize knowledge extracted from these guidelines and implement the rules so that they can be used into an auto-assessing tool of cardiovascular risk. Formalization followed five steps and was conducted under the guidance of medical experts. It resulted into a decision tree fed by eight decision variables. Evaluation of the accuracy of the decision tree showed 80% of agreement with an expert in medical informatics in predicting the cardiovascular risk level for 15 different clinical situations. Discrepancies correspond to the knowledge gaps within Clinical Practice Guidelines. We intend to extend the implementation of the decision tree to a complete tool, for allowing patient to auto-assess their cardiovascular risk. This tool will be integrated into a platform providing recommendations adapted to the calculated level of cardiovascular risk.

Using visual analytics for presenting comparative information on new drugs.

OBJECTIVE: When a new drug is marketed, physicians must decide whether they will consider it for their future practice. However, information about new drugs can be biased or hard to find. In this work, our objective was to study whether visual analytics could be used for comparing drug properties such as contraindications and adverse effects, and whether this visual comparison can help physicians to forge their own well-founded opinions about a new drug. MATERIALS AND METHODS: First, an ontology for comparative drug information was designed, based on the expectations expressed during focus groups comprised of physicians. Second, a prototype of a visual drug comparator website was developed. It implements several visualization methods: rainbow boxes (a new technique for overlapping set visualization), dynamic tables, bar charts and icons. Third, the website was evaluated by 22 GPs for four new drugs. We recorded the general satisfaction, the physician's decision whether to consider the new drug for future prescription, both before and after consulting the website, and their arguments to justify their choice. RESULTS: The prototype website permits the visual comparison of up to 10 drugs, including efficacy, contraindications, interactions, adverse effects, prices, dosage regimens,…All physicians found that the website allowed them to forge a well-founded opinion on the four new drugs. The physicians changed their decision about using a new drug in their future practice in 29 cases (out of 88) after consulting the website. DISCUSSION AND CONCLUSION: Visual analytics is a promising approach for presenting drug information and for comparing drugs. The visual comparison of drug properties allows physicians to forge their opinions on drugs. Since drug properties are available in reference texts, reviewed by public health agencies, it could contribute to the independent of drug information.

Parallel Design of Browsing Scheme and Data Model for Multi-Level Hierarchical Application Devoted to Management of Patient with Infectious Disease in Primary Care.

Many decision systems are based on a hierarchical approach, enriching the known context used to finally choose the right potential action. Designing the scheme for browsing the clinical guidelines is a task devoted to expert in infectious diseases. Designing the data model is a task devoted to the expert in data modeling. As a consequence, browsing scheme and data model generally differ in terms of abstraction levels. While the browsing scheme proposes to navigate into depth, the data model stays flat. We propose here a novel method to design in parallel the browsing scheme and the data model so that both of them reflect the different abstraction levels in decision process.

Automatic Extraction of Drug Adverse Effects from Product Characteristics (SPCs): A Text Versus Table Comparison.

BACKGROUND: Potential adverse effects (AEs) of drugs are described in their summary of product characteristics (SPCs), a textual document. Automatic extraction of AEs from SPCs is useful for detecting AEs and for building drug databases. However, this task is difficult because each AE is associated with a frequency that must be extracted and the presentation of AEs in SPCs is heterogeneous, consisting of plain text and tables in many different formats. METHODS: We propose a taxonomy for the presentation of AEs in SPCs. We set up natural language processing (NLP) and table parsing methods for extracting AEs from texts and tables of any format, and evaluate them on 10 SPCs. RESULTS: Automatic extraction performed better on tables than on texts. CONCLUSION: Tables should be recommended for the presentation of the AEs section of the SPCs.

Decision System Integrating Preferences to Support Sleep Staging.

Scoring sleep stages can be considered as a classification problem. Once the whole recording segmented into 30-seconds epochs, features, extracted from raw signals, are typically injected into machine learning algorithms in order to build a model able to assign a sleep stage, trying to mimic what experts have done on the training set. Such approaches ignore the advances in sleep medicine, in which guidelines have been published by the AASM, providing definitions and rules that should be followed to score sleep stages. In addition, these approaches are not able to solve conflict situations, in which criteria of different sleep stages are met. This work proposes a novel approach based on AASM guidelines. Rules are formalized integrating, for some of them, preferences allowing to support decision in conflict situations. Applied to a doubtful epoch, our approach has taken the appropriate decision.

Towards a Wireless Smart Polysomnograph Using Symbolic Fusion.

Polysomnography is the gold standard test for sleep disorders among which the Sleep Apnea Syndrome (SAS) is considered a public health issue because of the increase of the cardio-and cerebro-vascular risk it is associated with. However, the reliability of this test is questioned since sleep scoring is a time-consuming task performed by medical experts with a high inter- and intra-scorers variability, and because data are collected from 15 sensors distributed over a patient's body surface area, using a wired connection which may be a source of artefacts for the patient's sleep. We have used symbolic fusion to support the automated diagnosis of SAS on the basis of the international guidelines of the AASM for the scoring of sleep events. On a sample of 70 patients, and for the Apnea-Hypopnea Index, symbolic fusion performed at the level of sleep experts (97.1% of agreement). The next step is to confirm these preliminary results and move forward to a smart wireless polysomnograph.

Personalized sleep staging system using evolutionary algorithm and symbolic fusion.

This paper presents a novel system for automatic sleep staging based on evolutionary technique and symbolic intelligence. Proposed system mimics decision making process of clinical sleep staging using Symbolic Fusion and considers personal singularity with an adaptive thresholds setting up system using Evolutionary Algorithm. It proved to be an effective and promising system in personalizing sleep staging. This system can also be integrated with other medical systems to realize remote sleep monitoring or home-care.

Building a Semantic Interoperability Framework for Care and Research in Fibromuscular Dysplasia.

UNLABELLED: Identifying patients with Fibromuscular Dysplasia (FMD) at the international level will have considerable value for understanding the epidemiology, clinical manifestations and susceptible genes in this arterial disease, but also for identifying eligible patients in clinical trials or cohorts. We present a two-step methodology to create a general semantic interoperability framework allowing access and comparison of distributed data over various nations, languages, formats and databases. METHODS: The first step is to develop a pivot multidimensional model based on a core dataset to harmonize existing heterogeneous data sources. The second step is to align the model to additional data, semantically related to FMD and collected currently in various registries. We present the results of the first step that has been fully completed with the validation and implementation of the model in a dedicated information system (SIR-FMD). We discuss the current achievements for step 2 and the extensibility of the methodology in the context of other rare diseases.

Abstractive Summarization of Drug Dosage Regimens for Supporting Drug Comparison.

Complicated dosage regimens often reduce adherence to drug treatments. The ease-of-administration must thus be taken into account when prescribing. Given one drug, there exists often several dosage regimens. Hence, comparison to similar drugs is difficult. Simplifying and summarizing them appears to be a required task for supporting General Practitioners to find the drug with the simplest regimen for the patient. We propose a summarization in two steps: first prunes out all low-importance information, and second proceed to fusion of remaining information. Rules for pruning and fusion strategies were designed by an expert in drug models. Evaluation was conducted on a dataset of 169 drugs. The agreement rate was 27.2%. We demonstrate that applying rules leads to a result that is correct by a computational point of view, but the result is often meaningless for the GP. We conclude with recommendations for further work.

A new approach for cleansing geographical dataset using Levenshtein distance, prior knowledge and contextual information.

Epidemiological studies are necessary to take public health decisions. Their relevance depends on the quality of data. Doctors in continuous care collect a big amount of data that can be used for epidemiological purpose, but spatial data may be dirty; based on city names, the localization is imprecise, even more if it is misspelled. The only way to identify a city without ambiguity is to use its identifier, which can be retrieved by cleansing geographical textual data. In France, cities are organized in administrative zones called departments and some city names are shared by several cities in several departments. The clear identification of the department and the city name allows to deduce the city unique identifier and to make some spatial analysis such as epidemiological studies. In this paper, we propose a method to cleanse such data, using several steps. After having standardized the text to cleanse, we use the Levenshtein distance to generate a first set of propositions. Finally, the propositions are filtered, by removing the less likely candidates, so that it remains only one, which becomes the chosen city. Tested on a dataset of 9818 entries, we obtained 89.1% of concordance, whereas the standard Levenshtein distance obtained 70.5%. This demonstrates that our method has better results.