Sort by:
Fri
27 May
13:30 - 14:30
Open
Opening ceremony
Room: Athena (Location: Acropolis, Number of seats: 730)
Fri
27 May
14:30 - 15:30
KN 1
Niels Peek Keynote session
Room: Athena (Location: Acropolis, Number of seats: 730)
Fri
27 May
16:00 - 17:30
Athena
Panel
Room: Athena (Location: Acropolis, Number of seats: 730)
Chair: Enea Parimbelli
Submission:
1
16:00 - 17:30
Explainability, Causability, Causality, Reliability: The many facets of ``good'' explanations in XAI for health (Enea Parimbelli, Arianna Dagliati, Brent Mittlestadt, Riccardo Guidotti, Niels Peek, Andreas Holzinger)
n.a.
Mr. Enea Parimbelli
University of Pavia
Assistant prof. of Biomedical engineering. Clinical decision support, AI in medicine and XAI
Ms. Dr. Arianna Dagliati
University of Pavia
Mr. Brent Mittlestadt
University of Oxford
Mr. Riccardo Guidotti
University of Pisa
Mr. Dr. Niels Peek
The University of Manchester
Mr. Andreas Holzinger
Medical University of Graz, Austria
Fri
27 May
16:00 - 17:30
Rh 10
Workshop
Room: Rhodes 10 (Location: Acropolis, Number of seats: 130)
Chair: Maria Hagglund
Submission:
1
16:00 - 17:30
Current State and Future Directions of Patients’ Access to their Electronic Health Records across Europe – an interactive workshop (Maria Hägglund, Isabella Scandurra, Anne Moen, Anna Kharko, Monika Johansen)
Patient portals are increasingly available to facilitate access to personal health information and used to facilitate communication between patients and healthcare professionals, as well as for performing administrative tasks, such as appointment bookings and prescription renewals. Patients are also increasingly provided with access to their electronic health records (EHRs), as a service to “look up and read” information, sometimes referred to as patient accessible EHRs (PAEHRs), or when referring specifically to the notes in the record, ‘open notes’, through portals. This is an important contribution to “Personal Health Data Spaces” to support citizen use of their data in line with the effort for setting up a European Health Data Space.
Research evidence reports positive outcomes among patients accessing their records, and the concerns expressed by healthcare professionals have not been realised. Patients who read their notes report better understanding of their care plans, feeling more in control of their care, doing a better job taking their medications improved communication with and trust in their clinicians and improved patient safety. Yet, implementation is far from straight forward, and many obstacles hinder or delay progress in uptake and use.
In this workshop, we want to discuss this topic based on results from a currently ongoing European survey that elicit to what extent patients across Europe can access their records online, and engage the participants in discussions of strategies to interact with the data and potentially added value in curating their information for personal use.
Ms. Associate Professor Maria Hägglund
Uppsala University
Fri
27 May
16:00 - 17:30
Rh 11
Session 1
Room: Rhodes 11 (Location: Acropolis, Number of seats: 130)
Chair: Parisis Gallos
Submissions:
1
16:00 - 16:15
Scaling AI projects for Radiology – Causes and Consequences (Gunnar Ellingsen, Line Silsand, Gro-Hilde Severinsen, Line Linstad)
Artificial intelligence (AI) for radiology has the potential to handle an ever-increasing volume of imaging examinations. However, the implementation of AI for clinical practice has not lived up to expectations. We suggest that a key problem with AI projects in radiology is that high expectations associated with new and unproven AI technology tend to scale the projects in ways that challenge their anchoring in local practice and their initial purpose of serving local needs. Empirically, we focus on the procurement of an AI solution for radiology practice at a large health trust in Norway where it was intended that AI technology would be used to process the screening of images more effectively. Theoretically, we draw on the information infrastructure literature, which is concerned with scaling innovative technologies from local settings, with a limited number of users, to broad-use contexts with many users.
Mr. Gunnar Ellingsen
UIT - The Arctic University of Norway
Ms. Dr. Line Silsand Researcher, Ph.D., RN.
Norwegian Centre for E-health Research
Line Silsand has a PhD in health science, and a particular interest in large-scale health information system. In addition, she has a bachelor's degree in nursing, and pedagogical education.
Ms. Dr Gro-Hilde Severinsen
2
16:15 - 16:30
ECG Classification Using Combination of Linear and Non-linear Features with Neural Network (Tetiana Biloborodova, Inna Skarga-Bandurova, Illia Skarha-Bandurov, Yelyzaveta Yevsieieva, Oleg Biloborodov)
In this paper, we present an approach to improve the accuracy and reliability of ECG classification. The proposed method combines features analysis of linear and non-linear ECG dynamics. Non-linear features are represented by complexity measures of assessment of ordinal network non-stationarity. We describe the basic concept of ECG partitioning and provide an experiment on PQRST complex data. The results demonstrate that the proposed technique effectively detects abnormalities via automatic feature extraction and improves the state-of-the-art detection performance on one of the standard collections of heartbeat signals, the ECG5000 dataset.
Ms. Tetiana Biloborodova
G.E. Pukhov Institute for Modelling in Energy Engineering
Ms. Inna Skarga-Bandurova
Oxford Brookes University
3
16:30 - 16:45
Dataset Comparison tool: utility and privacy (João Almeida, Ricardo Cruz-Correia, Pedro Pereira Rodrigues)
Synthetic data has been more and more used in the last few years. While its applications are various, measuring its utility and privacy is seldom an easy task. Since there are different methods of evaluating these issues, which are dependent on data types, use cases and purpose, a generic method for evaluating utility and privacy does not exist at the moment. So, we introduced a compilation of the most recent methods for evaluating privacy and utility into a single executable in order to create a report of the similarities and potential privacy breaches between two datasets, whether it is related to synthetic or not. We catalogued 24 different methods, from qualitative to quantitative, column-wise or table-wise evaluations. We hope this resource can help scientists and industries get a better grasp of the synthetic data they have and produce more easily and a better basis to create a new, more broad method for evaluating dataset similarities.
Mr. João Almeida
Faculty of Medicine of the University of Porto
4
16:45 - 17:00
Attitudes and Acceptance Towards Artificial Intelligence in Medical Care (Dana Holzner, Timo Apfelbacher, Wolfgang Rödle, Christina Schüttler, Hans-Ulrich Prokosch, Rafael Mikolajczyk, Sarah Negash, Nadja Kartschmit, Iryna Manuilova, Charlotte Buch, Jana Gundlack, Jan Christoph)
Background: Artificial intelligence (AI) in medicine is a very topical issue. As far as the attitudes and perspectives of the different stakeholders in healthcare are concerned, there is still much to be explored.

Objective: Our aim was to determine attitudes and aspects towards acceptance of AI applications from the perspective of physicians in university hospitals.

Methods: We conducted individual exploratory expert interviews. Low fidelity mockups were used to show interviewees potential application areas of AI in clinical care.

Results: In principle, physicians are open to the use of AI in medical care. However, they are critical of some aspects such as data protection or the lack of explainability of the systems.

Conclusion: Although some trends in attitudes e.g., on the challenges or benefits of using AI became clearer, it is necessary to conduct further research as intended by the subsequent PEAK project
Ms. Dana Holzner
Department of Medical Informatics, Friedrich-Alexander-Universität Erlangen-Nürnberg
5
17:00 - 17:15
Can Artificial Intelligence Enable the Transition to Electric Ambulances? (Emmanouil Rigas, Antonis Billis, Panagiotis Bamidis)
The electrification of the transportation sector is seen as a main pathway to reduce CO2 emissions and mitigate the earth's climate change. Currently, Electric Vehicles (EVs) are entering the market fast. Although EVs have not been used as ambulances yet, the transition to the new type of vehicle is a matter of time. Thus, in this paper we discuss a number of research questions related to the efficient deployment of electric ambulances, focusing on the Artificial Intelligence (AI) point of view and we propose a framework for developing online algorithms that schedule the charging of electric ambulances and their assignment to patients.
Mr. Dr. Emmanouil Rigas
Mr. Panagiotis Bamidis
Lab of Medical Physics and Digital Innovation, School of Medicine, Aristotle University of Thessaloniki (AUTH)
Fri
27 May
16:00 - 17:30
Rh 9-1
Session 2
Room: Rhodes 9-1 (Location: Acropolis, Number of seats: 120)
Chairs: Cyril Grouin and Akram REDJDAL
Submissions:
1
16:00 - 16:15
Pretrained Neural Networks Accurately Identify Cancer Recurrence in Medical Record (Helen Chen, Hussam Kaka, George Michalopoulos, Sujan Subendran, Kathleen Decker, Pascal Lambert, Marshall Pitz, Harminder Singh)
Cancer recurrence is the diagnosis of a second clinical episode of cancer after the first was considered cured. Identifying patients who had experienced cancer recurrence is an important task as it can be used to compare treatment effectiveness, measure recurrence-free survival, and plan and prioritize cancer control resources. We developed BERT-based natural language processing (NLP) contextual models for identifying cancer recurrence incidence and the recurrence time based on the records in progress notes. Using two datasets containing breast and colorectal cancer patients, we demonstrated the advantage of the contextual models over the traditional NLP models by overcoming the laborious and often unscalable tasks of composing keywords in a specific disease domain.
Ms. Dr. Helen Chen
University of Waterloo
2
16:15 - 16:30
Classification of Oncology Treatment Responses from French Radiology Reports with Supervised Machine Learning (Jean-Philippe Goldman, Luc Mottin, Jamil Zaghir, Daniel Keszthelyi, Belinda Lokaj, Hugues Turbé, Patrick Ruch, Julien Ehrsam, Christian Lovis, Julien Gobeil)
The present study shows first attempts to automatically classify oncology treatment responses on the basis of the textual conclusion sections of radiology reports according to the RECIST classification. After a robust and extended manual annotation of 543 conclusion sections (5-to-50-word long), and after the training of several machine learning techniques (from traditional machine learning to deep learning), the best results show an accuracy score of 0.90 for a two-class classification (non-progressive vs. progressive disease) and of 0.82 for a four-class classification (complete response, partial response, stable disease, progressive disease) both with Logistic Regression approach. Some innovative solutions are further suggested to improve these scores in the future.
Mr. Jean-Philippe Goldman
University of Geneva
3
16:30 - 16:45
Discovering key topics in emergency medical dispatch from free text dispatcher observations (Pablo Ferri, Carlos Sáez, Antonio Félix-De Castro, Purificación Sánchez-Cuesta, Juan M Garcia-Gomez)
The objective of this work was to discover key topics latent in free text dispatcher observations registered during emergency medical calls. We used a total of 1374931 independent retrospective cases from the Valencian emergency medical dispatch service in Spain, from 2014 to 2019. Text fields were preprocessed to reduce vocabulary size and filter noise, removing accent and punctuation marks, along with uninformative and infrequent words. Key topics were inferred from the multinomial probabilities over words conditioned on each topic from a Latent Dirichlet Allocation model, trained following an online mini-batch variational approach. The optimal number of topics was set analyzing the values of a topic coherence measure, based on the normalized pointwise mutual information, across multiple validation K-folds. Our results support the presence of 15 key topics latent in free text dispatcher observations, related with: ambulance request; chest pain and heart attack; respiratory distress; head falls and blows; fever, chills, vomiting and diarrhea; heart failure; syncope; limb injuries; public service body request; thoracic and abdominal pain; stroke and blood pressure abnormalities; pill intake; diabetes; bleeding; consciousness. The discovery of these topics implies the automatic characterization of a huge volume of complex unstructured data containing relevant information linked to emergency medical call incidents. Hence, results from this work could lead to the update of structured emergency triage algorithms to directly include this latent information in the triage process, resulting in a positive impact in patient wellbeing and health services sustainability.
Mr. Pablo Ferri
Universitat Politècnica de València
4
16:45 - 17:00
Clustering Nursing Sentences - Comparing Three Sentence Embedding Methods (Hans Moen, Henry Suhonen, Sanna Salanterä, Tapio Salakoski, Laura-Maria Peltonen)
In health sciences, high-quality text embeddings may augment qualitative data analysis of large amounts of text by enabling, e.g., searching and clustering of health information. This study aimed to evaluate three different sentence-level embedding methods in clustering sentences in nursing narratives from individual patients’ hospital care episodes. Two of these embeddings are generated from language models based on the BERT framework, and the third on the Sent2Vec method. These embedding methods were used to cluster sentences from 20 patient care episodes and the results were manually evaluated. Findings suggest that the best clusters were produced by the embeddings from a BERT model fine-tuned for the proxy task of predicting subject headings for nursing text.
Ms. Laura-Maria Peltonen
University of Turku
Mr. Hans Moen
Aalto University
5
17:00 - 17:15
Enriching UMLS-based phenotyping of rare diseases using deep-learning: evaluation on Jeune syndrome (Carole FAVIEZ, Marc VINCENT, Nicolas GARCELON, Caroline MICHOT, Genevieve BAUJAT, Valerie CORMIER-DAIRE, Sophie SAUNIER, Xiaoyi CHEN, Anita BURGUN)
The wide adoption of Electronic Health Records (EHR) in hospitals provides unique opportunities for high throughput phenotyping of patients. The phenotype extraction from narrative reports can be performed by using either dictionary-based or data-driven methods. We developed a hybrid pipeline using deep learning to enrich the UMLS Metathesaurus for automatic detection of phenotypes from EHRs. The pipeline was evaluated on a French database of patients with a rare disease characterized by skeletal abnormalities, Jeune syndrome. The results showed a 2.5-fold improvement regarding the number of detected skeletal abnormalities compared to the baseline extraction using the standard release of UMLS. Our method can help enrich the coverage of the UMLS and improve phenotyping, especially for languages other than English.
Ms. Carole FAVIEZ
INSERM
6
17:15 - 17:20
Deep Learning-based Brain Hemorrhage Detection in CT Reports (Gıyaseddin Bayrak, Muhammed Şakir TOPRAK, Murat Can GANİZ, Halife KODAZ, Ural KOÇ)
Radiology reports can potentially be used to detect critical cases that need immediate attention from physicians. We focus on detecting Brain Hemorrhage from Computed Tomography (CT) reports. We train a deep learning classifier and observe the effect of using different pre-trained word representations along with domain-specific fine-tuning. We have several contributions. Firstly, we report the results of a large-scale classification model for brain hemorrhage detection from Turkish radiology reports. Second, we show the effect of fine-tuning pre-trained language models using domain-specific data on the performance. We conclude that deep learning models can be used for detecting brain Hemorrhage with reasonable accuracy and fine-tuning language models using domain-specific data to improve classification performance.
Mr. Gıyaseddin Bayrak
Marmara University
Mr. Muhammed Şakir TOPRAK
Ministry of Health - Turkey
Mr. Murat Can GANİZ
Marmara University, Computer Engineering Department
7
17:20 - 17:25
Automatic Prediction of Semantic Labels for French Medical Terms (Thierry Hamon, Natalia Grabar)
We address the problem of semantic labeling of terms in two French medical corpora with the subset of the UMLS. We perform two experiments relying on the structure of words and terms, and on their context: (1) the semantic label of already identified terms is predicted ; (2) the terms are detected in raw texts and their semantic label is predicted. Our results show over 0.90 F-measure.
Mr. Thierry Hamon
LIMSI, CNRS & Université Paris 13
8
17:25 - 17:30
Artificial Intelligence Competencies in Postgraduate Medical Training in Germany (Louis Agha-Mir-Salim, Lina Mosch, Sophie Anne Ines Klopfenstein, Maximilian Markus Wunderlich, Nicolas Frey, Akira-Sebastian Poncette, Felix Balzer)
Routine medical care is to be transformed by the introduction of artificial intelligence (AI), requiring medical professionals to acquire a novel set of skills. We assessed the density of AI learning objectives and the availability of courses containing AI content in postgraduate medical education in Germany. The results reveal general paucity in AI learning objectives and content across (sub-)specialty training and continuing medical education (CME) in Germany. Innovative and regulatory solutions are needed to herald an era of physicians competent in navigating medical AI applications.
Mr. Louis Agha-Mir-Salim
Charité - Universitätsmedizin Berlin
Fri
27 May
16:00 - 17:30
Rh 9-2
Session 3
Room: Rhodes 9-2 (Location: Acropolis, Number of seats: 130)
Chairs: Felix Holl and Jacob Hofdijk
Submissions:
1
16:00 - 16:15
Understanding public priorities and perceptions of the use of linked healthcare data in South East England (Elizabeth Ford, Kathryn Stanley, Melanie Rees-Roberts, Sarah Giles, Katie Goddard, Jo Armes)
The counties of Kent, Surrey and Sussex (KSS) in South East England are creating anonymized, linked databases of healthcare records for audit, service planning and research for the first time. We consulted with 79 citizens from KSS in 5 deliberative focus groups, asking about perceived benefits and concerns regarding these new data assets. Participants hoped the linked datasets could be used for joining up care and information, improving efficiency, and improving healthcare provision, but were concerned about missing and inaccurate data, data breaches and hacking, use of data by profit-making organisations, and stigma and discrimination. Findings will be used to underpin governance and engagement strategies for integrated datasets in KSS.
Ms. Dr Elizabeth Ford
Brighton and Sussex Medical School
2
16:15 - 16:30
The rising importance of e-health in Norway (Rolf Wynn, Vicente Traver, Gunnar Ellingsen)
Drawing on three central sources of data on the development in e-health use in Norway (studies from the Norwegian Centre for e-Health Research, studies from Statistics Norway, and the Tromsø 7 Study), we describe the rising importance of e-health. Originally restricted to a limited use within the health services, in recent years the use of e-health has gained momentum both in the general population and within the traditional health services, as the Internet has offered easy access to health information as well as a range of other health-related services.
Mr. Rolf Wynn
University of Tromsø, Norway
3
16:30 - 16:45
Design for a Virtual peer-to-peer Knowledge to Action Platform for Type 2 Diabetes (Manmohan Mittal, Karim Keshavjee, Elham Golkhandan, Alessia Paglialonga, Aziz Guergachi, Marie Therese Lussier, Claude Richard, Laurette Dubed, Ian Zenleae, Robert Kyba, Warren Smokey Thomas)
Many patients with Type 2 Diabetes (T2D) have difficulty in controlling their disease despite the widespread availability of high-quality guidelines, T2D education programs, and primary care follow-up programs. Current diabetes education and treatment programs translate knowledge from bench to bedside well but underperform on the ‘last mile’ of converting that knowledge into action (KTA). Two innovations to the last-mile problem in the management of patients with T2D are introduced. 1) Design of a platform for peer-to-peer groups where patients can solve KTA problems together in a structured and psychologically safe environment using all the elements of the Action Cycle phase of the KTA framework. The platform uses Self-Determination Theory as the behavior change theory. 2) A novel patient segmentation method to enable the formation of groups of patients who have similar behavioral characteristics and therefore who are more likely to find common cause in the fight against diabetes.
Mr. Aziz Guergachi
Ted Rogers School of Management, Ryerson University
4
16:45 - 17:00
An agile approach to accelerate development and adoption of electronic Product Information standards (Catherine Chronaki, Anne Moen, Petter Hurlen, Giorgio Cangioli, Jens Kristian Villandsen, Giovanna Maria Ferarri, Craig Anderson)
The Medical Product Information found in most medication boxes offer a wealth of information, including terms of active ingredients, excipients, indications, dosage, route of administration, risks, and safety information. Digital health services that help patients, their care givers, and health professionals to manage medication, can be improved with tailored information based on user profile, the patient’s Electronic Health Record (EHR) summary, and Medicinal Product Information. The electronic Product information (ePI) comprises the summary of product characteristics, package leaflet, and product label. The European Medicines Agency released in 2021 the first version of the EU proof-of-concept ePI standard based on HL7 FHIR. The Gravitate-Health project uses this common standard as a springboard to implement a federated open-source platform and services that helps advance access, understanding, and adherence by providing trusted medicinal information in an interoperable and scalable way. In this paper, we present the agile technical approach and co-creation process to design, test, and progressively mature interoperability working with the HL7 Vulcan Accelerator and FHIR connectathons.
Ms. - Catherine Chronaki Secretary General
HL7 Foundation
Ms. Professor Anne Moen RN, PhD, FACMI, FIAHSI
University of Oslo
Mr. Dr. Petter Hurlen
Akershus University Hospital
Mr. Giorgio Cangioli
Mr. Craig Anderson
Pfizer
5
17:00 - 17:15
Collecting Data from a Mobile App and a Smartwatch Supports Treatment of Schizophrenia and Bipolar Disorder (Steinunn Gróa Sigurðardóttir, Anna Sigridur Islind, María Óskarsdóttir)
Mental disorders affect individuals and societies around the world neg-
atively, with the health-related burden of 32,4% out of the overall disease bur-
den. This large part of the overall burden underlines a growing need for innova-
tion to support the treatment of mental disorders like schizophrenia and bipolar
disorder. This empirical study features two groups of patients; a group of nine pa-
tients diagnosed with bipolar disorder and a group of twelve patients diagnosed
with schizophrenia. The patients in the study carry a smartwatch for six weeks,
continuously collecting data into a digital health platform. Additionally, they an-
swer five daily wellbeing questions in a mobile app. To supplement that data, they
also answer a questionnaire three times over the interval and at the end of the pe-
riod they attend a semi-structured interview. We offer four main aspects to consider
for PGHD in mental health: i) sharing data easily with healthcare professionals, ii)
being able to engage with your own PGHD, iii) the watch use can help the patients
regulate routine in their daily life, iv) tonality and phrasing.
Ms. Steinunn Gróa Sigurðardóttir
Reykjavik University, Department of Computer Science, Reykjavik, Iceland
Ms. Dr. Anna Sigridur Islind
Reykjavik University, School of Computer Science, Reykjavik, Iceland
6
17:15 - 17:30
iCompanion: Α Serious Games App for the Management of Frailty (Antonios Sykoutris, Angelina Kouroubali, Dimitrios Katehakis, Haridimos Kondylakis)
The term frailty is often used to describe a particular state of health, related to the ageing process, often experienced by older people. The most common indicators of frailty are weakness, fatigue, weight loss, low physical activity, poor balance, low gait speed, visual impairment and cognitive impairment. The objective of this work is the creation of a serious games mobile application to conduct elderly frailty assessments in an accurate and objective way using mobile phone capabilities. The proposed app includes three games (memory card, endless runner, and clicker) and three questionnaires, aiming towards the prediction of signs of memory and reflection deterioration, as well as endurance and strength. The games, when combined with a set of qualified questionnaires, can provide an efficient tool to support adults in identifying frailty symptoms and in some cases prevent further deterioration. At the same time the app can support older adults in improving physical and mental fitness, while gathering useful information about frailty.
Mr. Haridimos Kondylakis PhD
Foundation for Research and Technology - Hellas
Ms. Dr. Angelina Kouroubali
Foundation for Research and Technology Hellas (FORTH)
Sat
28 May
07:45 - 08:45
Athena
Workshop
Room: Athena (Location: Acropolis, Number of seats: 730)
Chair: Théophile Tiffet
Submission:
1
07:45 - 08:45
Deep learning from scratch (Théophile Tiffet, Cédric Bousquet, Christel Ducroz Gérardin)
Artificial intelligence, especially deep learning, achieved improved performances these last years, and has the potential to support medical practice. This is especially true for interpreting medical imagery, and extracting information from textual documents such as hospitalisation or surgery reports in the electronic health records. However, deep learning has a steep learning curve and may seem obscure at first. This workshop aims at demystifying how neural networks work. It is intended for, on the one hand those who are not familiar with the field but wish to have a first experience, and on the other hand, to participants who do not know the intermediate steps of the hidden layer computations performed automatically by deep learning frameworks such as “Pytorch” or “Keras”. We will reimplement a basic neural network from scratch, without using deep learning frameworks, and we will progressively make it more complex. Participants will learn what happens in the neural network, how it learns and makes predictions from the input data. In particular, we will see in detail the gradient descent algorithm. The updating of weights by backpropagation will be introduced. For a full explanation and implementation of it, additional notebooks will be provided at the end of the workshop.
Mr. Dr. Théophile Tiffet
University hospital of Saint-Etienne
Mr. Dr. Cédric Bousquet
Sorbonne Université, Inserm, université Paris 13, Laboratoire d’informatique médicale et d’ingénierie des connaissances en e-santé, LIMICS, F-75006 Paris, France
Ms. Christel Ducroz Gérardin
Sorbonne Université, Inserm, Institut Pierre Louis d'Epidémiologie et de Santé Publique, IPLESP, F-75012 Paris, France
Sat
28 May
07:45 - 08:45
EFMI NI GA
General Assembly NI
Room: Rhodes 10 (Location: Acropolis, Number of seats: 130)
Sat
28 May
07:45 - 08:45
Rh 11
Workshop
Room: Rhodes 11 (Location: Acropolis, Number of seats: 130)
Chair: Felix Holl
Submission:
1
07:45 - 08:45
Young EFMI: Building a Community of Students, Young Professionals and Scientists Within EFMI (Felix Holl, Ivana Ognjanovic)
yEFMI aims to support students, young professionals and scientists in their professional development and establish a network for peer-support with young EFMI members. The aim of this workshop is to improve the visibility of yEFMI and improve engagement of new members in yEFMI. Furthermore, the workshop is intended to identify future steps to improve yEFMI.
Mr. Felix Holl M.Sc.
Neu-Ulm University of Applied Sciences
Felix Holl is an PhD student in medical informatics at the University of Munich. He is a research associate at Neu-Ulm University of Applied Sciences in Germany and an affiliate at the Institute for Global Health Sciences at the University of California, San Francisco, USA. His research focuses on the development of evaluation methods for mHealth applications. He also has a research interest in the use of informatics in global health.
Ms. Ivana Ognjanovic
University Donja Gorica, Montenegro
Sat
28 May
07:45 - 08:45
Rh 9-1
Workshop
Room: Rhodes 9-1 (Location: Acropolis, Number of seats: 120)
Chair: Arriel Benis
Submission:
1
07:45 - 08:45
Improving communication in digital health using EFMI Medical Informatics Multilingual Ontology (Arriel Benis, Mihaela Crisan-Vida, Stefan Darmoni, Lacramioara Stoicu-Tivadar)
Digital health is a multidisciplinary field. The European Federation for Medical Informatics comprises more than 30 partner-countries affiliated societies. This workshop aims to discuss about EFMI Medical Informatics Multilingual Ontology (MIMO) which aims to improve the interregional cooperation and collaboration in the European healthcare informatics. The team who develop EFMI MIMO are members from European Federation of Medical Informatics (EFMI) Health Informatics for Interregional Cooperation Working Group. One of the Working Group goal is to improve and facilitates the communication between professionals in digital health domain and reduce misunderstanding
Mr. Dr. Arriel Benis PhD
Holon Institute of Technology
PhD in Medical Informatics and Artificial Intelligence Senior Lecturer and Head of the "Business Intelligence and Automation" (BIA) laboratory and Co-Head of the "Industrial Automation and Internet of Things" (IIoT) at HIT - Holon Institute of Technology, Israel Chair of the EFMI WG HIIC - Health Informatics for Interregional Cooperation Co-Chair of the EFMI WG ODH - One Digital Health Israeli / ILAMI representative at the EFMI council EFMI Executive Officer 2021-2023
Ms. Mihaela Crisan-Vida
University Politehnica Timișoara
Mr. Prof. Stefan Darmoni
Rouen University Hospital
Ms. Prof. Lacramioara Stoicu-Tivadar
University Politehnica Timisoara
Sat
28 May
07:45 - 08:45
Rh 9-2
Workshop
Room: Rhodes 9-2 (Location: Acropolis, Number of seats: 130)
Submission:
1
07:45 - 08:45
Participatory Definition of the Feature Extraction for Conducting retrospective studies (Antoine Lamer, Mathilde Fruchart, Marc Cuggia, Guillaume Bouzillé, Emmanuel Chazard)
The increasing implementation and use of electronic health records over the last few decades has made a significant volume of clinical data being available. Despite the many opportunities data reuse offers, its implementation presents many difficulties and primary data cannot be reused directly. A process of feature extraction transforms therefore raw data into usable information for reuse. This workshop aims at expanding and completing the current definition of feature extraction. Attendees will share their use cases related to their activity and fields. Organizers will gather use cases and synthetise dimensions and methods reported by participants. The workshop will provide the material for the development and evaluation of future feature extraction methods.
Mr. Antoine Lamer
INSERM CIC-IT Lille, Lille
Ms. Mathilde Fruchart
Lille University
Sat
28 May
09:00 - 10:30
Athena
Session 4
Room: Athena (Location: Acropolis, Number of seats: 730)
Chairs: Lucia Sacchi and Lacramioara Stoicu-Tivadar
Submissions:
1
09:00 - 09:15
User satisfaction with an AI system for chest x-ray analysis implemented in a hospital’s emergency setting (Diego A. Rabinovich, Candelaria Mosquera, Pierina Torrens, Martina Aineseder, Sonia Benitez)
The acceptance of artificial intelligence (AI) systems by health professionals is crucial to obtain a positive impact on the diagnosis pathway. We evaluated user satisfaction with an AI system for the automated detection of findings in chest x-rays, after five months of use at the Emergency Department. We collected quantitative and qualitative data to analyze the main aspects of user satisfaction, following the Technology Acceptance Model. We selected the intended users of the system as study participants: radiology residents and emergency physicians. We found that both groups of users shared a high satisfaction with the system’s ease of use, while their perception of output quality (i.e., diagnostic performance) differed notably. The perceived usefulness of the application yielded positive evaluations, focusing on its utility to confirm that no findings were omitted, and also presenting distinct patterns across the two groups of users. Our results highlight the importance of clearly differentiating the intended users of AI applications in clinical workflows, to enable the design of specific modifications that better suit their particular needs. This study confirmed that measuring user acceptance and recognizing the perception that professionals have of the AI system after daily use can provide important insights for future implementations.
Ms. Engineer Candelaria Mosquera
2
09:15 - 09:30
MISeval: a Metric Library for Medical Image Segmentation Evaluation (Dominik Müller, Dennis Hartmann, Philip Meyer, Florian Auer, Iñaki Soto-Rey, Frank Kramer)
Correct performance assessment is crucial for evaluating modern artificial intelligence algorithms in medicine like deep-learning based medical image segmentation models. However, there is no universal metric library in Python for standardized and reproducible evaluation. Thus, we propose our open-source publicly available Python package MISeval: a metric library for Medical Image Segmentation Evaluation. The implemented metrics can be intuitively used and easily integrated into any performance assessment pipeline. The package utilizes modern DevOps strategies to ensure functionality and stability. MISeval is available from PyPI (miseval) and GitHub: https://github.com/frankkramer-lab/miseval.
Mr. Dominik Müller
University of Augsburg
Mr. Florian Auer
University of Augsburg
3
09:30 - 09:45
Characterization of Type 2 Diabetes using Counterfactuals and Explainable AI (Marta Lenatti, Alberto Carlevaro, Karim Keshavjee, Aziz Guergachi, Alessia Paglialonga, Maurizio Mongelli)
Type 2 diabetes mellitus is a metabolic disorder of glucose management, whose prevalence is increasing inexorably worldwide. Adherence to therapies, along with a healthy lifestyle can help prevent the onset of disease. This preliminary study proposes the use of explainable artificial intelligence techniques with the aim of (i) characterizing diabetic patients through a set of easily interpretable rules and (ii) providing individualized recommendations for the prevention of the onset of the disease through the generation of counterfactual explanations, based on minimal variations of biomarkers routinely collected in primary care. The results of this preliminary study parallel findings from the literature as differences in biomarkers between patients with and without diabetes are observed for fasting blood sugar, body mass index, and high-density lipoprotein levels.
Ms. Marta Lenatti
National Research Council of Italy (CNR), Institute of Electronics, InformationEngineering and Telecommunications (IEIIT), Italy
4
09:45 - 10:00
Supporting AI-Explainability by Analyzing Feature Subsets in a Machine Learning Model (Lucas Plagwitz, Alexander Brenner, Michael Fujarski, Julian Varghese)
Machine learning algorithms become increasingly prevalent in the field of medicine, as they offer the ability to recognize patterns in complex medical data. Especially in this sensitive area, the active usage of a mostly black box is a controversial topic. We aim to highlight how an aggregated and systematic feature analysis of such models can be beneficial in the medical context. For this reason, we introduce a grouped version of the permutation importance analysis for evaluating the influence of entire feature subsets in a machine learning model. In this way, expert-defined subgroups can be evaluated in the decision-making process. Based on these results, new hypotheses can be formulated and examined.
Mr. Lucas Plagwitz
Institute of Medical Informatics, University of Münster, Germany
5
10:00 - 10:15
Using supervised machine learning to predict burnout in healthcare professionals (Karthik Adapa, Malvika Pillai, Meagan Foster, Lukasz Mazur, Nadia Charguia)
Burnout in healthcare professionals (HCPs) is a multi-factorial problem. There are limited studies utilizing machine learning approaches to predict HCPs’ burnout during the COVID-19 pandemic. A survey consisting of demographic characteristics and work system factors was administered to 450 HCPs during the pandemic (participation rate: 59.3%). Recursive feature elimination with random forest had the highest performance (area under the receiver operating curve of 0.81). The eight key features that best predicted burnout are excessive workload, inadequate staffing, administrative burden, professional relationships, organizational culture, values and expectations, intrinsic motivation, and work-life integration. These findings provide evidence for resource allocation and implementation of interventions to reduce HCPs’ burnout and improve the quality of care.
Mr. Dr Karthik Adapa
University of North Carolina at Chapel Hill
6
10:15 - 10:20
Phenotyping of Heart Failure with preserved ejection faction using health electronic records and echocardiography (Morgane Pierre-Jean, Erwan Donal, Marc Cuggia, Guillaume Bouzillé)
Patients suffering from heart failure (HF) symptoms and a normal left ventricular ejection fraction (LVEF 50%) present very different clinical phenotypes that could influence their survival. This study aims to identify phenotypes of this type of HF by using the medical information database from Rennes University Hospital Center. We present a preliminary work, where we explore the use of clinical variables from health electronic records (HER) in addition to echocardiography to identify several phenotypes of patients suffering from heart failure with preserved ejection fraction. The proposed methodology identifies 4 clusters with various characteristics (both clinical and echocardiographic) that are linked to survival (death, surgery, hospitalization). In the future, this work could be deployed as a tool for the physician to assess risks and contribute to support better care for patients.
Ms. Morgane Pierre-Jean
CHU of Rennes, LTSI University of Rennes1
7
10:20 - 10:25
Machine Learning in Medicine: To explain, or not to explain, that is the question (Cédric Bousquet, Diva Beltramin)
In 2022, the Medical Informatics Europe conference created a special topic called “Challenges of trustable AI and added-value on health” which was centered around the theme of eXplainable Artificial Intelligence. Unfortunately, two opposite views remain for biomedical applications of machine learning: accepting to use reliable but opaque models, vs. enforce models to be explainable. In this contribution we discuss these two opposite approaches and illustrate with examples the differences between them.
Mr. Dr. Cédric Bousquet
Sorbonne Université, Inserm, université Paris 13, Laboratoire d’informatique médicale et d’ingénierie des connaissances en e-santé, LIMICS, F-75006 Paris, France
Sat
28 May
09:00 - 10:30
Rh 10
Session 5
Room: Rhodes 10 (Location: Acropolis, Number of seats: 130)
Chairs: Georgi Chaltikyan and Alfred Winter
Submissions:
1
09:00 - 09:15
Minor and parental access to electronic health records: Differences across four countries (Josefin Hagström, Isabella Scandurra, Jonas Moll, Charlotte Blease, Barbara Haage, Iiris Hörhammer, Maria Hägglund)
An increasing number of countries are implementing patient access to electronic health records (EHR). However, EHR access for parents, children and adolescents presents ethical challenges of data integrity, and regulations vary across providers, regions, and countries. In the present study, we compare EHR access policy for parents, children and adolescents in four countries. Documentation from three areas: upper age limit of minors for which parents have access; age at which minors obtain access; and possibilities of access restriction and extension was collected from Sweden, Norway, Finland, and Estonia. Results showed that while all systems provided parents with automatic proxy access, age limits for its expiry differed. Furthermore, a lower minimum age than 18 for adolescent access was present in two of four countries. Differences between countries and potential implications for adolescents are discussed. We conclude that experiences of various approaches should be explored to promote the development of EHR regulations for parents, children and adolescents that increases safety, quality, and equality of care.
Ms. Josefin Hagström
Uppsala University
2
09:15 - 09:30
National integration components challenge the Epic implementation in Central Norway (Gunnar Ellingsen, Morten Hertzum, Bente Christensen, Rolf Wynn)
Electronic health record (EHR) suites cover a broad range of cross-sectoral use scenarios. Thereby, they streamline information flows but also require that healthcare professionals with diverse responsibilities must adapt to one and the same system. In the region of Central Norway, the EHR suite from Epic is being implemented at hospitals as well as in municipal healthcare. However, the 64 municipalities in the region are increasingly exploring the option of bypassing Epic by supplementing their existing systems with national integration components. These components provide integration and data exchange across systems for selected healthcare information. We discuss whether they are a viable alternative to Epic. The three components are the summary care record, the shared medication list, and the national welfare technology hub.
Mr. Gunnar Ellingsen
UIT - The Arctic University of Norway
Mr. Professor Morten Hertzum
University of Copenhagen
Ms. Bente Christensen
Nord Unviersity
Mr. Rolf Wynn
UiT The Arctic University of Norway (University of Tromsø)
3
09:30 - 09:45
Implementation of a Data Warehouse in Primary Care : First Analyses with Elderly Patients (Mathilde Fruchart, Paul Quindroit, Jean-Baptiste Beuscart, Haris Patel, Matthieu Calafiore, Antoine Lamer)
The implementation of clinical data warehouses has advanced in recent years. The standardization of clinical data in these warehouses has made it possible to carry out multicenter studies and to formalize the clinical vocabulary. However, there is limited insight into a patient's overall care pathway in the clinical domain. Regarding primary care data, the implementation of this type of warehouse in a routine way is hindered in particular by the analysis of textual data provided by general practitioners during patient consultations. In our study we collected primary care data for standardization in a data warehouse. The purpose of this analysis was to assess the feasibility of analyzing primary care data, and particularly to study the consultations and prescriptions of the elderly patient contained in our primary care data warehouse.
Keywords. Primary care, Data warehouse, Elderly patient, Data Reuse, Electronic Health Record
Ms. Mathilde Fruchart
Lille University
4
09:45 - 10:00
Patients' Access to their Psychiatric Records - A Comparison of Four countries (Annika Bärkås, Maria Hägglund, Jonas Moll, Åsa Cajander, Hanife Rexhepi, Iiris Hörhammer, Charlotte Blease, Isabella Scandurra)
Several Nordic and Baltic countries are forerunners in the digitalization of patient ehealth services and have since long implemented psychiatric records as parts of the ehealth services. There are country-specific differences in what clinical information is offered to patients concerning their online patient accessible psychiatric records. This study explores national differences in Sweden, Norway, Finland, and Estonia in patient access to their psychiatric records. Data was collected through a socio-technical data collection template developed during a workshop series and then analyzed in a cross-country comparison focusing on items related to psychiatry records online. The results show that psychiatric records online are offered to patients in all four countries, and provide the same functionality and similar psychiatry information. Overall, the conclusion is that experiences of various functionalities should be scrutinized to promote transparency of psychiatric records as part of the national eHealth services to increase equality of care and patient empowerment.
Ms. Annika Bärkås
Uppsala University
5
10:00 - 10:15
Human Factors of EHR Adoption in Saudi Primary Healthcare (bander alanazi, bander alanazi)
Electronic Health Records are rapidly gaining traction in healthcare with increased acceptance and adoption. However, there is limited understanding of factors influencing adoption in primary care. This paper investigates the human factors of EHR adoption in primary healthcare in Saudi Arabia. An online survey questionnaire was sent to all primary healthcare professionals in Riyadh city, Saudi Arabia. A 65.9% (1127/1710) response rate was obtained. The respondents demonstrated positive perceptions of EHRs in relation to the systems’ benefits. The perceptions were influenced by sociodemographic variables; hence, need consideration when implementing EHRs in primary care.
Mr. Dr bander alanazi
6
10:15 - 10:30
Assessment of Health Service Quality through Electronic Health Records – A Scoping Review (Hanna von Gerich, Laura-Maria Peltonen)
The World Health Organization defines, that high quality health services should be effective, safe, people-centered, timely, equitable, integrated, and effective. This requires systematic quality assessment. The aim of this scoping review was to explore how electronic health records (EHRs) have been used to assess quality of health services using the WHO criteria. A total of 4247 records were obtained whereof 8 studies were included in the review. Research showed that EHRs were used to evaluate safety, performance and care processes. EHRs were regarded as an applicable real-world data source, highlighting the importance of consistency and standardised terminologies. Use of EHR data is limited to its representation of the real world and current evaluation systems have limited quality criteria, diverse definitions and they use only structured data. Future research should explore possibilities of natural language processing methods and include narrative EHR information for a more a comprehensive view of service quality assessment
Ms. Hanna von Gerich
Ms. Laura-Maria Peltonen
University of Turku
Sat
28 May
09:00 - 10:30
Rh 11
Session 7
Room: Rhodes 11 (Location: Acropolis, Number of seats: 130)
Chairs: Ferdinand Dhombres and Christophe Gaudet-Blavignac
Submissions:
1
09:00 - 09:15
OntoBioStat: Supporting Causal Diagram Design and Analysis (Thibaut Pressat-Laffouilhère, Julien Grosjean, Jacques Bénichou, Stefan Darmoni, Lina F. Soualmia)
Suitable causal inference in biostatistics can be best achieved by knowledge representation thanks to causal diagrams or directed acyclic graphs. However, necessary and sufficient causes are not easily represented. Since existing ontologies do not fill this gap, we designed OntoBioStat in order to enable covariate selection support based on causal relation representations. OntoBioStat automatic ontological causal diagram construction and inferences are detailed in this study. OntoBioStat inferences are allowed by Semantic Web Rule Language rules and axioms. First, statements made by the users include outcome, exposure, covariate, and causal relation specification. Then, reasoning enable automatic construction using generic instances of Meta_Variable and Necessary_Variable classes. Finally, inferred classes highlighted potential bias such as confounder-like. Ontological causal diagram built with OntoBioStat was compared to a standard causal diagram (without OntoBioStat) in a theoretical study. It was found that confounding and bias were not completely identified by the standard causal diagram, and erroneous covariate sets were provided. Further research is needed in order to make OntoBioStat more usable.
Mr. Dr Thibaut Pressat-Laffouilhère
Rouen University Hospital
2
09:15 - 09:30
Mapping of ICD-O tuples to OncoTree codes using SNOMED CT post-coordination (Tessa Ohlsen, Valerie Kruse, Rosemarie Krupar, Alexandra Banach, Josef Ingenerf, Cora Marisa Drenkhahn)
Around 500,000 oncological diseases are diagnosed in Germany every year which are documented using the International Classification of Diseases for Oncology (ICD-O). Apart from this, another classification for oncology, OncoTree, is often used for the integration of new research findings in oncology. For this purpose, a semi-automatic mapping of ICD-O tuples to OncoTree codes was developed. The implementation uses a FHIR terminology server, pre-coordinated or post-coordinated SNOMED CT expressions, and subsumption testing. Various validations have been applied. The results were compared with reference data of scientific papers and manually evaluated by a senior pathologist, confirming the applicability of SNOMED CT in general and its post-coordinated expressions in particular as a viable intermediate mapping step. Resulting in an agreement of 84,00 % between the newly developed approach and the manual mapping, it becomes obvious that the present approach has the potential to be used in everyday medical practice.
Ms. Tessa Ohlsen
University of Lübeck, Institute for Medical Informatics, Lübeck, Germany
3
09:30 - 09:45
Deep SNOMED CT enabled large clinical database about COVID-19 (Christophe Gaudet-Blavignac, Julien Ehrsam, Hugues Turbé, Daniel Keszthelyi, Jamil Zaghir, Christian Lovis)
In spring 2020, as the COVID-19 pandemic is in its first wave in Europe, the University hospitals of Geneva (HUG) is tasked to take care of all Covid inpatients of the Geneva canton. It is a crisis with very little tools to support decision-taking authorities, and very little is known about the Covid disease. The need to know more, and fast, highlighted numerous challenges in the whole data pipeline processes. This paper describes the decisions taken and processes developed to build a unified database to support several secondary usages of clinical data, including governance and research. HUG had to answer to 5 major waves of COVID-19 patients since the beginning of 2020. In this context, a database for COVID-19 related data has been created to support the governance of the hospital in their answer to this crisis. The principles about this database were a) a clearly defined cohort; b) a clearly defined dataset and c) a clearly defined semantics. This approach resulted in more than 28 000 variables encoded in SNOMED CT and 1 540 human readable labels. It covers more than 216 000 patients and 590 000 inpatient stays. This database is used daily since the beginning of the pandemic to feed the “Predict” dashboards of HUG and prediction reports as well as several research projects.
Mr. Christophe Gaudet-Blavignac
Division of Medical Information Sciences, University Hospitals of Geneva and University of Geneva
4
09:45 - 10:00
Data element mapping in the data privacy era (Romain GRIFFIER, Sébastien COSSIN, Francois KONSCHELLE, Fleur MOUGIN, Vianney JOUHET)
Secondary use of health data is made difficult in part because of large semantic heterogeneity. Many efforts are being made to align local terminologies with international standards. With increasing concerns about data privacy, we focused here on the use of machine learning methods to align biological data elements using aggregated features that could be shared as open data. A 3-step methodology (features engineering, blocking strategy and supervised learning) was proposed. The first results, although modest, are encouraging for the future development of this approach.
Mr. Romain GRIFFIER
Bordeaux University Hospital
5
10:00 - 10:15
Temporal Medical Knowledge Representation using ontologies (Jacques Hilbey, Xavier Aimé, Jean Charlet)
Representing temporal information is a recurrent problem for biomedical ontologies. We propose a foundational ontology that combines the so-called three-dimensional and four-dimensional approaches in order to be able to track changes in an individual and to trace his or her medical history. This requires, on the one hand, associating with any representation of an individual the representation of his or her life course and, on the other hand, distinguishing the properties that characterize this individual from those that characterize his or her life course.
Mr. Jacques Hilbey
Sorbonne Université
6
10:15 - 10:20
WikiMeSH: Multi lingual MeSH translations via Wikipedia (Mikaël Dusenne, Kévin Billey, Florent Desgrippes, Arriel Benis, Stefan Darmoni, Julien Grosjean)
Objective: The aim of this paper is to propose an extended translation of the MeSH thesaurus based on Wikipedia pages. Methods: A mapping was realized between each MeSH descriptor (preferred terms and synonyms) and corresponding Wikipedia pages. Results: A tool called “WikiMeSH” has been developed. Among the top 20 languages of this study, seven have currently no MeSH translations: Arabic, Catalan, Farsi (Iran), Mandarin Chinese, Korean, Serbian, and Ukrainian. For these seven languages, WikiMeSH is proposing a translation for 47% for Arabic to 34% for Serbian. Conclusion: WikiMeSH is an interesting tool to translate the MeSH thesaurus and other health terminologies and ontologies based on a mapping to Wikipedia pages.
Mr. Dr. Arriel Benis PhD
Holon Institute of Technology
PhD in Medical Informatics and Artificial Intelligence Senior Lecturer and Head of the "Business Intelligence and Automation" (BIA) laboratory and Co-Head of the "Industrial Automation and Internet of Things" (IIoT) at HIT - Holon Institute of Technology, Israel Chair of the EFMI WG HIIC - Health Informatics for Interregional Cooperation Co-Chair of the EFMI WG ODH - One Digital Health Israeli / ILAMI representative at the EFMI council EFMI Executive Officer 2021-2023
Mr. Prof. Stefan Darmoni
Rouen University Hospital
7
10:20 - 10:25
An Ontology for Cardiothoracic Surgical Education and Clinical Data Analytics (Maryam Panahiazar, Yorick Chern, Ramon rojas, Omar Latin, Dexter Hadley, Ramin Beygui)
The development of an ontology facilitates the organization of the variety of concepts and relationships used to describe different terms in different resources. This ontology will facilitate the study in the different domains including computer-assisted surgery (CAS), surgical education, and analytics in electronic medical records (EMR).
Ms. Dr. Maryam Panahiazar
Institute for Computational Health Sciences, University of California, San Francisco
Sat
28 May
09:00 - 10:30
Rh 9-1
Session 8
Room: Rhodes 9-1 (Location: Acropolis, Number of seats: 120)
Chairs: Sylvia PELAYO and Anne Moen
Submissions:
1
09:00 - 09:15
Health Care Professionals´ Perspectives on the uses of Patient-generated Health Data (Sharon Guardado, Guido Giunti, Minna Isomursu)
Integration of digital self-management solutions into health care processes requires the involvement of health care professionals in the adoption and use of the solutions as part of the care pathway. We conducted 23 interviews with diverse profiles of health care professionals participating in the treatment of chronic patients in three different countries. Our results indicate that health care professionals appeared relatively motivated at the prospect of having access to patient-generated data. Nevertheless, they appeared less confident in weighing what types of data could be collected efficiently through mobile devices and how it could be presented in ways that would provide value to the care process. Our results identify four broad categories for how patient-generated health data could be useful: monitoring, prevention, research, and transparency of condition parameters.
Ms. MSc Sharon Guardado
University of Oulu
2
09:15 - 09:30
Personal health records an approach to answer: What works for whom in what circumstances? (Elisavet Andrikopoulou, Philip Scott)
National Health Service (NHS) policy suggests that increasing usage of electronic personal health records (PHR) by patients will result in cost savings and improved public health, especially for people with long-term conditions. PHR design features are inevitably important, since a good PHR design should make the users achieve their health goals effortlessly, which is understandable and usable. Three original theoretical models were developed using realist evaluation, one per long-term condition cohort, describing the interaction between the PHR design features and the patient and disease specific factors, to help determine what works for whom in what circumstances.
Ms. Dr Elisavet Andrikopoulou
University of Portsmouth
3
09:30 - 09:45
Patient values associated with an exergame supporting COPD treatment (Kira Oberschmidt, Marijke Broekhuis, Christiane Grünloh)
Exercise games (exergames) can help COPD patients stay active and prevent exacerbation. When evaluating such exergames, patient values are an important variable to take into account. In this study, seven COPD patients used an exergame technology at the physiotherapist for six months. Their values regarding treatment and the exergames were identified in interviews. Values were very stable throughout the study, and closely interconnected. Personal Guidance and Independence were important values. Additionally, patients sometimes held conflicting values that they prioritized differently at different times or based on specific events. As the study identified important values that appeared stable over a period of time, albeit with different prioritization, they are worth considering when designing new technology. However, values cannot be looked at in isolation because of the strong connection between values.
Ms. Kira Oberschmidt
Roessingh Research and Development
4
09:45 - 10:00
Mindful workarounds in Bar Code Medication Administration (Valentina Lichtner, Dawn Dowding)
Bar-Coded Medication Administration systems (BCMA) are often used with workarounds. These workarounds are usually judged against standard operating procedures or the use of the technology as ‘designers’ intended’. However, some workarounds may be reasonable and justified to prevent safety errors. In this conceptual paper, we clarify BCMA safety mechanisms and provide a framework to identify workarounds with BCMA that nullify the error prevention mechanisms inherent in the technology design and process. We also highlight the importance of understanding the purpose behind a nurse’s workaround in BCMA, focusing on the notion of mindful (thoughtful) workarounds that have the potential to improve patient safety.
Ms. Dr Valentina Lichtner
University of Leeds
5
10:00 - 10:15
How does mental workload influence the adoption of clinical information systems: An exploratory study (Lisanne Kremer, Ann-Kathrin Schwarz, Rainer Röhrig, Bernhard Breil)
Mental workload and technology acceptance are relevant factors that relate to use behavior and performance. Studies show a potential moderating effect of mental workload on predictors of technology acceptance. Aim of this study was the investigation of predictors of technology acceptance (UTAUT) related to clinical information systems and their relation to mental workload. This quasi-experimental study with 48 participants used the following measures: NASA TLX and UTAUT questionnaire. Participants had to perform three tasks on a clinical information system as well as four task-levels of the n-back task with increasing difficulty. Analyses show a high level of technology acceptance (M=3.82, SD=.76) and confirm performance expectancy as the most relevant predictor of behavioral intention (β=.48, p<.001). A linear regression showed that a high level of mental workload has an influence on performance expectancy (F1,46=8.438, p<.05). The study shows an influence of mental workload on acceptance, the strength and role of which (e.g. moderation) needs to be further investigated, especially in the context of other determinants.

Ms. Lisanne Kremer
Niederrhein University of Applied Sciences
6
10:15 - 10:30
Experiences Using Patient and Public Involvement in Digital Health Research for Multiple Sclerosis (Tiia Yrttiaho, Minna Isomursu, Guido Giunti)
Abstract. Patient and public involvement (PPI) is increasingly used for improving quality of the research. There are many barriers in translating PPI into practice, including lacking examples of good practices. Frameworks that have been developed in one setting do not readily transfer to other settings. In this paper, we examine the implementation of PPI in the context of a digital health research project that explores the design, development and use of mHealth for persons with Multiple Sclerosis taking an iterative user-centered design approach. Methods: Instrumental case study to describe the PPI process on a digital health research project. Results: Overall experience was positive. We found 3 roles for PPI involvement: strategic members; design and development partners; and expert members. Challenges lay on unclear PPI terminology; managing roles and expectations; and ensuring accessibility.
Ms. Tiia Yrttiaho
University of Oulu
Sat
28 May
09:00 - 10:30
Rh 9-2
Session 9
Room: Rhodes 9-2 (Location: Acropolis, Number of seats: 130)
Chair: Martin Staemmler
Submissions:
1
09:00 - 09:15
Towards a Clinical Decision Support System for helping medical students in emergency call centers (Edouard MICHOT, Jules WOO, Louis MOULINE, Catherine SINNAPPAN, Adrien BOUKOBZA, Florence CAMPEOTTO, Laurent DUPIC, Anita BURGUN, Benoît VIVIEN, Rosy TSOPRA)
In critical situations such as pandemic, medical students are often called to help in emergency call centers. However, they may encounter difficulties in phone triage because of a lack of medical skills. Here, we aim at developing a Clinical Decision Support System for helping medical students in phone call triage of pediatric patients. The system is based on the PAT (Pediatric Assessment Triangle) and local guidelines. It is composed of two interfaces. The first allows a quick assessment of severity signs, and the second provides recommendations and additional elements such as “elements to keep in mind” or “medical advice to give to patient”. The system was evaluated by 20 medical students, with two fictive clinical cases. 75% of them found the content useful and clear, and the navigation easy. 65% would feel more reassured to have this system in emergency call centers. Further works are planned to improve the system before implementation in real-life.
Ms. Dr Rosy TSOPRA
Université de Paris AP-HP
Mr. Edouard MICHOT
Université de Paris
2
09:15 - 09:30
An interactive interface for displaying recommendations on emergency phone triage in pediatrics (Claire DURCHON, Stephi Vanderlan, Alice JEGARD, Hasini SARAM, Marina FALCHI, Florence CAMPEOTTO, Laurent DUPIC, Anita BURGUN, Benoît VIVIEN, Rosy TSOPRA)
Emergency phone triage aims at identifying quickly patients with critical emergencies. Patient triage is not an easy task, especially in situations involving children, mostly due to the lack of training and the lack of clinical guidelines for children. To overcome these issues, we aim at designing and assessing an interactive interface for displaying recommendations on emergency phone triage in pediatrics. Four medical students formalized local guidelines written by the SAMU of Paris, into a decision tree and designed an interface according to usability principles. The navigation within the interface was designed to allow the identification of critical emergencies at the beginning of the decision process, and thus ensuring a quick response in case of critical emergencies. The interface was assessed by 10 medical doctors: they appreciated the ergonomics (e.g., intuitive colors), and found easy to navigate through the interface. Nine of them would like to use this interface during phone call triage. In the future, this interface will be improved and implemented in emergency call centers.
Ms. Dr Rosy TSOPRA
Université de Paris AP-HP
Ms. Claire DURCHON
Université de Paris
Ms. Stephi Vanderlan
Université de Paris
3
09:30 - 09:45
Optimization of performance by combining most sensitive and specific models in data science results in majority voting ensemble (Katoo Muylle, Pieter Cornu, Wilfried Cools, Kurt Barbé, Ronald Buyl, Sven Van Laere)
Ensemble modeling is an increasingly popular data science technique that combines the knowledge of multiple base learners to enhance predictive performance. In this paper, the idea was to increase predictive performance by holding out three algorithms when testing multiple classifiers: (a) the best overall performing algorithm (based on the harmonic mean of sensitivity and specificity (HMSS) of that algorithm); (b) the most sensitive model; and (c) the most specific model. This approach boils down to majority voting between the predictions of these three base learners. In this exemplary study, a case of identifying a prolonged QT interval after administering a drug-drug interaction with increased risk of QT prolongation (QT-DDI) is presented. Performance measures included accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). Overall performance was measured by calculating the HMSS. Results show an increase in all performance measure characteristics compared to the original best performing algorithm, except for specificity where performance remained stable. The presented approach is fairly simple and shows potential to increase predictive performance, even without adjusting the default cut-offs to differentiate between high and low risk cases. Future research should look at a way of combining all tested algorithms, instead of using only three. Similarly, this approach should be tested on a multiclass prediction problem.
Mr. Dr. Sven Van Laere
Vrije Universiteit Brussel
4
09:45 - 10:00
Implementation of a Clinical Trial Recruitment Support System Based on Fast Healthcare Interoperability Resources (FHIR) in a Cardiology Department (Clemens Scherer, Stephan Endres, Martin Orban, Stefan Kääb, Steffen Massberg, Alfred Winter, Matthias Löbe)
Clinical Trial Recruitment Support Systems can booster patient inclusion of clinical trials by automatically analyzing eligibility criteria based on electronic health records. However, missing interoperability has hindered introduction of those systems on a broader scale. Therefore, our aim was to develop a recruitment support system based on FHIR R4 and evaluate its usage and features in a cardiology department. Clinical conditions, anamnesis, examinations, allergies, medication, laboratory data and echocardiography results were imported as FHIR resources. Clinical trial information, eligibility criteria and recruitment status were recorded using the appropriate FHIR resources without extensions. Eligibility criteria linked by the logical operation “OR” were represented by using multiple FHIR Group resources for enrollment. The system was able to identify 52 of 55 patients included in four clinical trials. In conclusion, use of FHIR for defining eligibility criteria of clinical trials may facilitate interoperability and allow automatic screening for eligible patients at multiple sites of different healthcare providers in the future. Upcoming changes in FHIR should allow easier description of “OR”-linked eligibility criteria.
Mr. Dr. Clemens Scherer
Department of Medicine I University Hospital LMU Munich
5
10:00 - 10:15
Out-of-hospital cardiac arrest detection by machine learning based on phonetic characteristics of the caller's voice. (Sonia Rafi, Cedric Gangloff, Etienne Paulhet, Ollivier Grimault, Louis Soulat, Guillaume Bouzillé, Marc CUGGIA)
Introduction. Out-of-hospital cardiac arrest (OHCA) is a major public health issue. The prognosis is closely related to the time from collapse to return of spontaneous circulation. Resuscitation efforts are frequently initiated at the request of emergency call center professionals who are specifically trained to identify critical conditions over the phone. However, 25% of OHCAs are not recognized during the first call. Therefore, it would be interesting to develop automated computer systems to recognize OHCA on the phone. The aim of this study was to build and evaluate machine learning models for OHCA recognition based on the phonetic characteristics of the caller's voice. Methods. All patients for whom a call was done to the emergency call center of Rennes, France, between 01/01/2017 and 01/01/2019 were eligible. The predicted variable was OHCA presence. Predicting variables were collected by computer-automatized phonetic analysis of the call. They were based on the following voice parameters: fundamental frequency, formants, intensity, jitter, shimmer, harmonic to noise ratio, number of voice breaks, and number of periods. Three models were generated using binary logistic regression, random forest, and neural network. The area under the curve (AUC) was the primary outcome used to evaluate each model performance. Results. 820 patients were included in the study. The best model to predict OHCA was random forest (AUC=74.9, 95% CI=67.4-82.4). Conclusion. Machine learning models based on the acoustic characteristics of the caller’s voice can recognize OHCA. The integration of the acoustic parameters identified in this study will help to design decision-making support systems to improve OHCA detection over the phone.
Ms. Sonia Rafi
Univ Rennes, CHU Rennes, INSERM, LTSI-UMR 1099, F-35000
6
10:15 - 10:30
Support Vector Machine-based classifier of cross-correlated phoneme segments for Speech Sound Disorder Screening (Emilian Erman Mahmut, Stelian Nicola, Vasile Stoicu-Tivadar)
This paper presents a Support-Vector Machine (SVM) based method of classification of cross-correlated phoneme segments as part of the development of an automated Speech Sound Disorder (SSD) Screening tool. The pre-processing stage of the algorithm uses cross-correlation to segment the target phoneme and extracts data from the new homogeneously trimmed audio samples. Such data is then fed into the SVM-based classification script which currently achieves an accuracy of 97.5% on a dataset of 132 rows. Given the global context of an increasing trend in the incidence of Speech Sound Disorders (SSDs) amongst early-school aged children (5-6 years old), the constraints imposed by the new Corona virus pandemic, and the (consequent) shortage of professionally trained specialists, an automated screening tool would be of much assistance to Speech-Language Pathologists (SLPs).
Mr. PhD Student Emilian Erman Mahmut
Politehnica University Timisoara
BA English Philol. (University of Bucharest) MSc Speech Therapy in Communication Processes (University of Bucharest) PhD Student Computers & IT (Politehnica University Timisoara)
Sat
28 May
11:00 - 12:00
KN 2
Ted Shortliffe keynote
Room: Athena (Location: Acropolis, Number of seats: 730)
Sat
28 May
13:00 - 14:30
Athena
Panel
Room: Athena (Location: Acropolis, Number of seats: 730)
Chair: Vimla Patel
Submission:
1
13:00 - 14:30
The Current Status of AI in Medicine: Is it a Rebirth or an Evolution? (Vimla Patel, Edward H. Shortliffe, Riccardo Bellazzi, Niels Peek, Vimla Patel)
With an explosive current interest in the implications of AI in Medicine (AIM), it is appropriate to consider the forces that led to the novel methodologies that are beginning to move into standard practice. This panel brings together four informatics scientists with complementary expertise spanning many central topics in the field. One focus of the discussions will be to assess the extent to which the current activities and enthusiasm for the AIM field are a natural extension of what has gone before. Are the methods receiving emphasis today, such as data science and deep learning, part of a natural evolution in the discipline or a radically new set of methods and foci that suggest that AIM has been reborn? The panel will build on, and update, historic assessments of the state of AI in medicine. The first was a keynote presented at Artificial Intelligence in Medicine Europe (AIME) in Maastricht in 1991. The second was a reflective panel discussion that included three of the current speakers at AIME 2007 in Amsterdam. The resulting papers, published in Artificial Intelligence in Medicine in 1993) and 2009, serve as a starting point for the updated discussions and projections by the current panelists. Participants will focus on their personal areas of expertise, ranging from clinical decision making, machine learning, and knowledge representation to systems integration, translational bioinformatics, evaluation, and cognitive issues in both the modeling of expertise and the creation of acceptable systems.

Ms. Professor Vimla Patel
New York Academy of Medicine
Mr. Chair Emeritus & Adjunct Professor Edward H. Shortliffe MD, PhD
Columbia University Medical Center New York, NY 10032
Chair Emeritus and Adjunct Professor of Biomedical Informatics at Columbia University’s Vagelos College of Physicians and Surgeons. Editor Emeritus of the Journal of Biomedical Informatics (Elsevier). Editor of the Springer textbook Biomedical Informatics: Computer Applications in Health Care and Biomedicine, released in its 5th edition (2021). Former President/CEO of AMIA. Received ACM's Grace Murray Hopper Award (1976), ACMI Morris F. Collen Award (2006), IMIA's François Grémy Award of Excellence (2021).
Mr. Prof Riccardo Bellazzi
Università di Pavia, Italy
Mr. Dr. Niels Peek
The University of Manchester
Sat
28 May
13:00 - 14:30
Rh 10
Workshop
Room: Rhodes 10 (Location: Acropolis, Number of seats: 130)
Chair: Rada Hussein
Submission:
1
13:00 - 14:30
Towards European Health Data Space Eosystem: Health Data Scenarios (Rada Hussein, Fernando Martin-Sanchez, Lucas Scherdel)
The European Health Data Space (EHDS) aims to better exchange and access health data, including electronic health records, genomics data, patient registries, etc., across Europe. In this way, EHDS will support healthcare delivery (as a primary use of data) in addition to health research and health policy-making purposes (as a secondary use of data). To achieve this goal, we need to understand and establish the future health data ecosystem with all healthcare stakeholders. In this workshop, we will discuss health data challenges and practice future scenarios with the conference participants from different European member states.
Ms. Dr. Rada Hussein
Ludwig Boltzmann Institute for Digital Health and Prevention
Mr. Fernando Martin-Sanchez
National Institute of Health Carlos III
Mr. Lucas Scherdel
DayOne
Sat
28 May
13:00 - 14:30
Rh 11
Session 10
Room: Rhodes 11 (Location: Acropolis, Number of seats: 130)
Chairs: Jacques Hilbey and Ivana Ognjanovic
Submissions:
1
13:00 - 13:15
AP-HP Health Data Space (AHDS) to the test of the Covid-19 pandemic (Christel Daniel, Nicolas Paris, Olivier Pierre, Nicolas Griffon, Stéphane Bréant, Nina Orlova, Patricia Serre, Damien Leprovost, Stephane Denglos, Alexandre Mouchet, Julien Dubiel, Rafael Gozlan, Gilles Chatellier, Romain Bey, Claire Hassen-Khodja, Marie Frank, Marie-France Mamzer, Martin Hilka)
Sharing observational and interventional health data within a common data space enables university hospitals to leverage such data for biomedical discovery and moving towards a learning health system. Objective: To describe the AP-HP Health Data Space (AHDS) and the IT services supporting piloting, research, innovation and patient care. Methods: Built on three pillars – governance and ethics, technology and valorization – the AHDS and its major component, the Clinical Data Warehouse (CDW) have been developed since 2015. Results: Thanks to data protection measures and GDPR compliant patient information, the AP-HP CDW has been made available at scale to AP-HP both healthcare professionals and public or private partners in January 2017. Supported by an institutional secured and high-performance cloud and an ecosystem of tools, mostly open source, the AHDS integrates a large amount of massive healthcare data collected during care and research activities that can be leveraged for secondary use. As of December 2021, the AHDS operates the electronic data capture for almost +840 clinical trials sponsored by AP-HP, the CDW is enabling the processing of health data from more than 11 million patients and generated +200 secondary data marts from IRB authorized research projects. During the Covid-19 pandemic which has disrupted hospital activities, AHDS has had to evolve quickly to support administrative professionals and caregivers heavily involved in the reorganization of both patient care and biomedical research. Conclusion: The AP-HP Data Space is a key facilitator for data-driven evidence generation and making the health system more efficient and personalized.
Ms. Dr Christel Daniel MD, PhD
AP-HP Sorbonne Université, INSERM
Christel DANIEL, MD, PhD Pathologist, PhD in Medical Informatics, director of Data expertise & Clinical Research Informatics. Innovation & Data (I&D) - Direction of Information System, AP-HP, the largest hospital entity and first biomedical research center in Europe. Project manager at INSERM (LIMICS UMRS 1142: “Medical informatics and knowledge engineering fore-health”). Primary areas of research are clinical informatics, clinical research informatics, semantic interoperability, learning health system. Key topics include: knowledge representation, bio-medical terminologies and ontologies, terminology services for clinical information systems, clinical data warehouses, standard-based integration of clinical information systems and clinical research applications (HL7, CDISC). Co-editor of the Clinical Research Informatics section of the International Medical Informatics Association (IMIA) yearbook since 2013. Past co-chair of IHE Anatomic Pathology. Member of DICOM WG26, HL7 France, HL7 Pathology SIG and CDISC France.
2
13:15 - 13:30
When context matters for credible measurement of drug-drug interactions based on real-world data (Romain LELONG, Badisse Dahamna, Hélène Berthelot, Willy Duville, Catherine Letord, Julien Grosjean, Catherine Duclos)
The frequency of potential drug-drug interactions (DDI) in published studies on real world data considerably varies due to the methodological framework. Contextualization of DDI has a proven effect in limiting false positives. In this paper, we experimented with the application of various DDIs contexts elements to see their impact on the frequency of potential DDIs measured on the same set of prescription data collected in EDSaN, the clinical data warehouse of Rouen University Hospital. Depending on the context applied, the frequency of daily prescriptions with potential DDI ranged from 0.89% to 3.90%. Substance-level analysis accounted for 48% of false positives because it did not account for some drug-related attributes. Consideration of the patient’s context could eliminate up to an additional 29% of false positives.
Ms. Pr Catherine Duclos
3
13:30 - 13:45
Analysis of stroke assistance in covid-19 pandemic by process mining techniques (Gabrielle dos Santos Leandro, Daniella Yuri Miura, Juliana Safanelli, Rafaela Mantoan Borges, Cláudia Maria Cabral Moro)
Medical assistance to stroke patients must start as early as possible; however, several changes have impacted healthcare services during the Covid-19 pandemic. This research aimed to identify the stroke onset-to-door time during the Covid-19 pandemic considering the different paths a patient can take until receiving specialized care. It is a retrospective study based on process mining (PM) techniques applied to 221 electronic healthcare records of stroke patients during the pandemic. The results are two process models representing the patient's path and performance, from the onset of the first symptoms to admission to specialized care. PM techniques have discovered the patient journey in providing fast stroke assistance.
Ms. Ms. Gabrielle dos Santos Leandro
Pontifícia Universidade Católica do Paraná
4
13:45 - 14:00
Automated diagnosis of Autism Spectrum Disorder condition using shape based features extracted from brainstem (A. R. Jac Fredo, Vaibhav Jain, Priya Rani, Anandh K. Ramaniharan, Abirami S, Rakshit Mittal)
Alterations to the brainstem can hamper cognitive functioning, including audiovisual and behavioral disintegration, leading to individuals with Autism Spectrum Disorder (ASD) face challenges in social interaction. In this study, a process pipeline for the diagnosis of ASD has been proposed, based on geometrical and Zernike moments features, extracted from the brainstem of ASD subjects. The subjects considered for this study are obtained from publicly available data base ABIDE (300 ASD and 300 typically developing (TD)). Distance regularized level set (DRLSE) method has been used to segment the brainstem region from the midsagittal view of MRI data. Similarity measures were used to validate the segmented images against the ground truth images. Geometrical and Zernike moments features were extracted from the segmented images. The significant features were used to train Support vector machine (SVM) classifier to perform classification between ASD and TD subjects. The similarity results show high matching between DRLSE segmented brainstem and ground truth with high similarity index scores of Pearson Heron-II (PH II) = 0.9740 and Sokal and Sneath-II (SS II) = 0.9727. The SVM classifier achieved 70.53% accuracy to classify ASD and TD subjects. Thus, the process pipeline proposed in this study is able to achieve good accuracy in the classification of ASD subjects.
Mr. Dr. A. R. Jac Fredo
Indian Institute of Technology, BHU
Mr. Rakshit Mittal
Laboratory LTCI, Department INFRES, Telecom Paris, Institut Polytechnique de Paris
5
14:00 - 14:15
Accelerating high-dimensional temporal modelling using graphics processing units for pharmacovigilance signal detection on real-life data (Pierre Sabatier, Jean Feydy, Anne-Sophie Jannot)
Adverse drug reaction is a major public health issue. The increasing availability of medico-administrative databases offers major opportunities to detect real-life pharmacovigilance signals. We have recently adapted a pharmaco-epidemiological method to the large dimension, the WCE (Weigthed Cumulative Exposure) statistical model, which makes it possible to model the temporal relationship between the prescription of a drug and the appearance of a side effect without any a priori hypothesis. Unfortunately, this method faces a computational time problem. The objective of this paper is to describe the implementation of the WCE statistical model using Graphics Processing Unit (GPU) programming as a tool to obtain the spectrum of adverse drug reactions from medico-administrative databases. The process is divided into three steps: pre-processing of care pathways using the Python library Panda, calculation of temporal co-variables using the Python library "KeOps", estimation of the model parameters using the Python library "PyTorch" - standard in deep learning. Programming the WCE method by distributing the heaviest portions (notably spline calculation) on the GPU makes it possible to accelerate the time required for this method by 1000 times using a computer graphics card and up to 10,000 times with a GPU server. This implementation makes it possible to use WCE on all the drugs on the market to study their spectrum of adverse effects, to highlight new vigilance signals and thus to have a global vigilance tool on medico-administrative database. This is a proof of concept for the use of this technology in epidemiology.
Mr. Dr Pierre Sabatier
Hôpital européen Georges Pompidou
6
14:15 - 14:30
Analysis of saturation in the Emergency Department: a data-driven queuing model using machine learning (Adrien Wartelle, Farah Mourad-Chehade, Farouk Yalaoui, David Laplanche, Stéphane Sanchez)
Emergency department is a key component of the health system where the management of crowding situations is crucial to the well-being of patients. This study proposes a new machine learning methodology and a queuing network model to measure and optimize crowding through a congestion indicator, which indicates a real-time level saturation.
Mr. Adrien Wartelle
Sat
28 May
13:00 - 14:30
Rh 9-1
Session 11
Room: Rhodes 9-1 (Location: Acropolis, Number of seats: 120)
Chairs: Arriel Benis and Ronald Cornet
Submissions:
1
13:00 - 13:15
Pattern-Based Logical Definitions of Prenatal Disorders Grounded on Dispositions (Mirna El Ghosh, Fethi Ghazouani, Elise Akan, Jean Charlet, Ferdinand Dhombres)
Biomedical ontologies define concepts having biomedical significance and the semantic relations among them. Developing high-quality and reusable ontologies in the biomedical domain is a challenging task. Pattern-based ontology design is considered a promising approach to overcome the challenges. Ontology Design Patterns (ODPs) are reusable modeling solutions to facilitate ontology development. This study relies on ODPs to semantically enrich biomedical ontologies by assigning logical definitions to ontological entities. Specifically, pattern-based logical definitions grounded on dispositions are given to prenatal disorders. The proposed approach is performed under the supervision of fetal domain experts.
Ms. Dr. Mirna El Ghosh
INSERM
2
13:15 - 13:30
Proposal of Semantic Annotation for German Metadata using Bidirectional Recurrent Neural Networks (Hannes Ulrich, Hristina Uzunova, Heinz Handels, Josef Ingenerf)
The distributed nature of our digital healthcare and the rapid emergence of new data sources prevents a compelling overview and the joint use of new data. Data integration, e.g., with metadata and semantic annotations, is expected to overcome this challenge. In this paper, we present an approach to predict UMLS codes to given German metadata using recurrent neural networks. The augmentation of the training dataset using the Medical Subject Headings (MeSH), particularly the German translations, also improved the model accuracy. The model demonstrates robust performance with 75% accuracy and aims to show that increasingly sophisticated machine learning tools can already play a significant role in data integration.
Mr. Hannes Ulrich
Universität zu Lübeck
3
13:30 - 13:45
TerminoDiff – Detecting Semantic Differences in HL7 FHIR CodeSystems (Joshua Wiedekopf, Cora Marisa Drenkhahn, Lorenz Rosenau, Hannes Ulrich, Ann-Kristin Kock-Schoppenhauer, Josef Ingenerf)
While HL7 FHIR and its terminology package have seen a rapid uptake by the research community, in no small part due to the wide availability of tooling and resources, there are some areas where tool availability is still lacking. In particular, the comparison of terminological resources, which supports the work of terminologists and implementers alike, has not yet been sufficiently addressed. Hence, we present TerminoDiff, an application to semantically compare FHIR R4 CodeSystem resources. Our tool considers differences across all levels required, i.e. metadata and concept differences, as well as differences in the edge graph, and surfaces them in a visually digestible fashion.
Mr. Joshua Wiedekopf
University of Lübeck
I attained my degree MSc Medical Informatics at the University of Lübeck, I am currently employed as a research assistant at the Institute of Medical Informatics at the University of Lübeck.
4
13:45 - 14:00
A conceptual framework for representing events under public health surveillance (Anya Okhmatovskaia, David L. Buckeridge, Yannan Shen, Iris Ganser, Nicholas King, Nigel Collier, Zaiqiao Meng)
Information integration across multiple event-based surveillance (EBS) systems has been shown to improve global disease surveillance in experimental settings. In practice, however, integration does not occur due to the lack of a common conceptual framework for encoding data within EBS systems. We aim to address this gap by proposing a candidate conceptual framework for representing events and related concepts in the domain of public health surveillance
Ms. Anya Okhmatovskaia
McGill Clinical & Health Informatics, Department of Epidemiology and Biostatistics, McGill University
5
14:00 - 14:15
Data-driven Modeling of Randomized Controlled Trial Outcomes (Zhehuan Chen, Yilu Fang, Hao Liu, Chunhua Weng)
Anecdotally, 38.5% of clinical outcome descriptions in randomized controlled trial publications contain complex text. Existing terminologies are insufficient to standardize outcomes and their measures, temporal attributes, quantitative metrics, and other attributes. In this study, we analyzed the semantic patterns in the outcome text in a sample of COVID-19 trials and presented a data-driven method for modeling outcomes. We conclude that a data-driven knowledge representation can benefit natural language processing of outcome text from published clinical studies.
Ms. Professor Chunhua Weng
Columbia University
6
14:15 - 14:20
Evaluation and challenges of medical procedure data harmonization to SNOMED-CT for observational research (Ines Reinecke, Michael Kallfelz, Martin Sedlmayr, Joscha Siebel, Franziska Bathelt)
The relevance of health data research on real world data (RWD) is increasing. To prepare national RWD for international research, harmonization with standard terminologies is required. In this paper, we evaluate to what extent the German OPS vocabulary in OHDSI covers codes present in RWD and mappings to SNOMED-CT. The evaluation identified a mapping gap of 21.1% in the RWD set.
Ms. Ines Reinecke
Technische Universität Dresden
7
14:20 - 14:25
A Wide Database for a Multicenter Study on Pneumocystis jirovecii Pneumonia in Intensive Care Units (Gabriele Di Meco, Sara Mora, Daniele Roberto Giacobbe, Silvia Dettori, Ilias Karaiskos, Matteo Bassetti, Mauro Giacomini)
Pneumocystis jirovecii pneumonia (PJP) is an opportunistic fungal infection that may affect patients with immunosuppression. In order to improve the diagnosis accuracy for PJP, facilitating the collection of data across Europe to reliably assess the performance of diagnostic tests for PJP is essential to improve the care of critically ill patients developing this severe condition. Such large data can be collected thanks to the contribution of several European hospitals in the compilation of a dedicated electronic Case Report Form (eCRF). The main focus of this work is to create an interface with high ergonomics both in the compilation and in the subsequent validation of the records.
Ms. Sara Mora
8
14:25 - 14:30
An ADR database for clinical use – potential of and difficulties with the Summary of Product Characteristics (Birgit Eiermann, Daniel Rodrigues, Paul Cohen, Lars L Gustafsson)
Adverse drug reactions (ADRs) for all drugs in Europe are described in the legally approved Summary of Product Characteristics (SmPC). An overview of all ADRs of the patients’ drug list can support healthcare staff to link patient symptoms to possible ADRs.
We review the possibilities and challenges to extract ADR information from SmPCs and present the development of our semi-automated procedure for extraction of ADRs from the tabulated section of the SmPCs to create a database, named Bikt, which is regularly updated and used at point of care in Sweden.
The existence of five major table formats for ADRs used in the SmPCs required the development of different parsing scripts. Manual checks for correctness for all content has to be performed. The quality of extraction was investigated for all SmPCs by measuring precision, recall and F1 scores (i.e. the weighted harmonic mean of precision and recall) and compared with other methods published.
We conclude that it is possible to semi-automatically extract ADR information from SmPCs. However, clear technical and content guidelines and standards for ADR tables and terms from drug registration authorities would lead to improved extraction and usability of ADR information at point of care.
Ms. Ph.D. Birgit Eiermann
Sat
28 May
13:00 - 14:30
Rh 9-2
Session 12
Room: Rhodes 9-2 (Location: Acropolis, Number of seats: 130)
Chairs: Jan-David Liebe and Jacob Hofdijk
Submissions:
1
13:00 - 13:15
Making EHRs Trustable: A Quality Analysis of EHR-derived Datasets for COVID-19 Research (Miguel Pedrera Jiménez, Noelia Garcia Barrio, Paula Rubio Mayo, Guillermo Maestro de la Calle, Antonio Lalueza Blanco, Ana García Reyne, María José Zamorro Lorenci, Jaime Cruz Rojo, Víctor Quirós González, José María Aguado García, Juan Luis Cruz Bermúdez, José Luis Bernal Sobrino, Laura Merson, Carlos Lumbreras Bermejo, Pablo Serrano Balazote)
One approach to verifying the quality of research data obtained from EHRs is auditing how complete and correct the data are in comparison with those collected by manual and controlled methods. This study analyzed data quality of an EHR-derived dataset for COVID-19 research, obtained during the pandemic at Hospital Universitario 12 de Octubre. Data were extracted from EHRs and a manually collected research database, and then transformed into the ISARIC-WHO COVID-19 CRF model. Subsequently, a data analysis was performed, comparing both sources through this convergence model. More concepts and records were obtained from EHRs, and PPV (95% CI) was above 85% in most sections. In future studies, a more detailed analysis of data quality will be carried out.
Mr. Engineer Miguel Pedrera Jiménez
Hospital Universitario 12 de Octubre (Madrid, Spain)
Ms. Noelia Garcia Barrio
Hospital Universitario 12 de Octubre
2
13:15 - 13:30
Metadata Definition in Registries: What is a Data Element? (Jürgen Stausberg, Sonja Harkener, Markus Burgmer, Christoph Engel, Robert Finger, Carsten Heinz, Ekkehart Jenetzky, David Martin, Rüdiger Rupp, Martin Schönthaler, Christian Schuld, Barbara Suwelack, Jeannine Wegner)
Observational research benefits from a rich methodological foundation of registry development and operation published in international and national guidelines. Metadata management is an essential part of registry implementation based on concepts of data elements and value sets. The metadata from six German registries revealed vastly divergent interpretations of the concept of data elements. The different perspectives of research questions, data acquisition and data storage were all represented in the registries’ catalogs of data elements. Consequently, the whole life cycle of a registry needs to be accompanied by a catalog of data elements, which has to be continuously adapted to the changing perspectives. A standard for the representation of those metadata is still missing. The FAIR Guiding Principles introduce important methodological requirements, but the tools for their fulfillment in respect to the management of metadata are still in its infancy.
Mr. Jürgen Stausberg
University Duisburg-Essen, Faculty of Medicine
3
13:30 - 13:45
Establishing a data quality baseline in the AKTIN Emergency Department Data Registry – A secondary use perspective (Lucas Triefenbach, Ronny Otto, Jonas Bienzeisler, Alexander Kombeiz, Saskia Ehrentreich, Rainer Röhrig, Raphael W. Majeed)
Secondary use of clinical data is an increasing application that is affected by the data quality (DQ) of its source systems. Techniques such as audits and risk- based monitoring for controlling DQ often rely on source data verification (SDV). SDV requires access to data generating systems. We present an approach to a targeted SDV based on manual input and synthetic data that is applicable in low resource settings with restricted system access. We deployed the protocol in the DQ management of the AKTIN Emergency Department Data Registry. Our targeted approach has shown to be feasible to form a DQ baseline that can be used for different DQ monitoring processes such as the identification of different error sources.
Mr. Lucas Triefenbach
Institute of Medical Informatics, Medical Faculty, RWTH Aachen University, Aachen, Germany
4
13:45 - 14:00
Automated coding in case mix databases of bacterial infections based on antimicrobial susceptibility test results (Radia SPIGA, François-Elie Calvier, Anne Carricajo, Béatrice Trombert Paviot, Bruno Pozzetto, Cédric Bousquet)
Our objective was to improve the accuracy of bacteria and resistance coding in a hospital case mix database. Data sources consisted of 50,074 files on bacteriological susceptibility tests transmitted with the HPRIM protocol from laboratory management system to electronic health record of the University hospital of Saint Etienne in July 2017. An algorithm was implemented to detect susceptibility tests containing information corresponding to codes whose addition in the case mix database was susceptible to increase the severity level of a diagnosis related group. Among 132 hospital stays fulfilling the conditions, 27 were lacking bacteria and/or resistance codes, and the tariff was increased for 9 stays, with earnings of €54,612. Analyzing Antimicrobial susceptibility tests helps to improve clinical coding and optimize the financial gain.
Ms. Radia SPIGA
University of Saint Etienne, CHU, Department of public health and medical informatics, Saint Etienne, France
5
14:00 - 14:15
Towards a Computational Approach for the Assessment of Compliance of ALCOA+ Principles in Pharma Industry (Marta Durá, Ángel Sánchez-García, Carlos Sáez, Fátima Leal, Adriana E. Chis, Horacio González-Vélez, Juan M Garcia-Gomez)
The pharmaceutical industry is a data-intensive environment and a heavily-regulated sector, where exhaustive audits and inspections are performed to ensure the safety of drugs. In this context, processing and evaluating the data generated in the manufacturing lines is a relevant challenge since it requires compliance with pharma regulations. This work combines data integrity metrics and blockchain technology to evaluate the compliance-degree of ALCOA+ principles among different levels of drug manufacturing data. We propose the DIALCOA tool, a software to assess the compliance-degree for each ALCOA+ principle, based on the assessment of data from manufacturing batch reports and its different levels of information.
Ms. Marta Durá
Universitat Politècnica de València
6
14:15 - 14:30
Are People Ready to Report Digital Health Ethical Issues in Order to Contribute to their Resolution? (Magali ROBERT, Nathalie BAUDINIERE, Brigitte SEROUSSI)
Although guaranteed by the GDPR, transparency of health data processing may not be fully respected, leading citizens to mistrust eHealth and discard digital health services. Identifying and safeguarding ethics in eHealth services is thus important to promote their development. We conducted a survey to assess the extent of ethical issues induced by the use of digital health services, understand the efforts citizens would be willing to accept for reporting such issues, and evaluate citizens’ expectations regarding this reporting. Among 200 respondents, 36% reported having encountered ethical issues with the processing of their health data or with digital health services being poorly inclusive. Faced to ethical issues when using a digital health service, 49% of respondents were rather or very angry, and 33% felt rather or very dependent. Most respondents were ready to report digital health ethical issues if there is a feedback for each report.