HEALTHINF 2024 Abstracts

Full Papers Short Papers
Full Papers
Paper Nr: 46
Title:

Comparison Between Graph Databases and RDF Engines for Modelling Epidemiological Investigation of Nosocomial Infections
Authors:

Lorena Pujante-Otalora, Manuel Campos, Jose M. Juarez and Maria-Esther Vidal
Abstract: We have evaluated the performance of property and knowledge graph databases in the context of spatiotemporal epidemiological investigation of an infection outbreak in a hospital. Specifically, we have chosen Neo4j as graph database, and GraphDB for knowledge graphs defined following RDF and its extension RDF*. We have defined a domain model describing a hospital layout and patient movements. For performance comparison, we have created ten graphs with different sizes based on MIMIC-III, implemented three epidemiological queries in Cypher, SPARQL and SPARQL* and defined three benchmarks that measure the execution time and main memory consumption of the three queries in each graph and database engine. Our research suggests that query complexity is a more determinant factor than graph size in the performance of the query executions. Neo4j presents better times and memory consumption than GraphDB for simple queries, but GraphDB is more efficient when traversing big subgraphs. Between RDF and RDF*, RDF* offers a more compact and human-friendly modelling and a better performance of the query execution.
Download
Paper Nr: 49
Title:

Sequential Networks for Predicting the Clinical Risk of Chronic Patients Using Drug Dispensation
Authors:

Daniel Hijosa-Guzmán, María Teresa Jurado-Camino, Pablo de Miguel-Bohoyo and Inmaculada Mora-Jiménez
Abstract: Chronic diseases are one of the leading causes of death worldwide, with diabetes, hypertension, congestive heart failure, and chronic obstructive pulmonary disease among the most common ones. In this sense, the extraction of clinical patterns from the data recorded in the Electronic Health Record is of great interest and motivates research in models to predict the temporal evolution of the patient’s health status. Predictive models would be of great help in the treatment of chronic patients to carry out preventive policies. Our approach considers the Gated Recurrent Unit neural network to extract temporal patterns of drug dispensation and to predict the progression of Chronic Conditions (CCs) towards a more complex health status. Real-world data linked to chronic patients of a Spanish hospital were considered, obtaining the most probable health status among a set of 10, including single dominant or moderate CCs, significant CCs in multiple organ systems, and dominant CC in three or more organ systems. Accuracy rates above 70% for single dominant or moderate CCs and nearly 50% for significant/dominant conditions across multiple organs were obtained. These results show the potential of sequential networks to predict the clinical risk of chronic patients and support clinical decision-making.
Download
Paper Nr: 56
Title:

e-Consent in Biomedical Research Registries: A GDPR-Compliant Approach Explored in the Context of the Australasian Diabetes Data Network
Authors:

Zhe Wang, Anthony Stell, Jean Paul Vera Soto and Richard O. Sinnott
Abstract: e-Consent - the digital capture of a patient’s consent to be involved in medical research - is a feature of biomedical research that is becoming increasingly prevalent with the advance of digital technology to support clinical/biomedical research and targeted registries. Although there have been many reviews of e-consent over the past decade - evaluating aspects such as informed consent, engagement, comprehension and data security - there remain unanswered questions about how e-consent fits in the context of recent data legislation and privacy demands such as the European General Data Protection Regulation (GDPR). This paper outlines key aspects of e-consent in the context of GDPR and the specific demands placed on biomedical registries used for diverse research objectives. We present a practical realisation of GDPR e-Consent in the context of the Australasian Diabetes Data Network (ADDN) – the national type-1 diabetes registry for Australia.
Download
Paper Nr: 59
Title:

Application of Formal Concept Analysis to Characterize Driving Behaviors and Socio-Cultural Factors Related to Driving
Authors:

Diogo Miranda, Luis Zárate and Mark Song
Abstract: This article addresses the global concern for road safety, where frequent accidents on roads and streets result in loss of human lives, severe injuries, and significant material damage, impacting not only the direct victims but also their families and society at large. To tackle this challenge, it is crucial to analyze the factors contributing to these accidents, particularly driver behaviors. This study investigates reckless behaviors such as speeding, less obvious influences such as personality traits and sociocultural factors. Using Formal Concept Analysis (FCA), the research examines a database containing information about Chinese drivers, aiming to provide valuable insights for accident prevention and the promotion of safer road behaviors. In summary, the article aims to deepen the understanding of factors related to traffic accidents with the goal of enhancing road and street safety.
Download
Paper Nr: 64
Title:

A Survey on Usability Evaluation in Digital Health and Potential Efficiency Issues
Authors:

Bilal Maqbool, Farzaneh Karegar and Sebastian Herold
Abstract: Context: Usability is a major factor in the acceptance of digital health (DH) solutions. Problem: Despite its importance, usability experts have expressed concerns about the insufficient attention given to usability evaluation in practice, indicating potential efficiency problems of common evaluation methods in the healthcare domain. Objectives: This research paper aimed to analyse industrial usability evaluation practices in digital health to identify potential threats to the efficiency of their application. Method: To this end, we conducted an online survey of 144 usability experts experienced in usability evaluations for digital health applications. The survey questions aimed to explore the prevalence of techniques applied, and the participants’ familiarity and perceptions regarding tools and techniques. Results: The prevalently applied techniques might impose efficiency problems in common scenarios in digital health. Participant recruitment is considered time-consuming and selecting the most appropriate evaluation method for a given context is perceived difficult. The results highlight a lack of utilisation of tools automating aspects of usability evaluation. Conclusions: A more widespread adoption of tools for automating usability evaluation activities seems desirable as well as guidelines for selecting evaluation techniques in a given context. We furthermore recommend to explore AI-based solutions to address the problem of involving targeted user groups that are difficult to access for usability evaluations.
Download
Paper Nr: 75
Title:

Synergizing Data Imputation and Electronic Health Records for Advancing Prostate Cancer Research: Challenges, and Practical Applications
Authors:

Abderrahim O. Batouche, Eugen Czeizler, Miika Koskinen, Tuomas Mirtti and Antti S. Rannikko
Abstract: The presence of detailed clinical information in electronic health record (EHR) systems presents promising prospects for enhancing patient care through automated retrieval techniques. Nevertheless, it is widely acknowledged that accessing data within EHRs is hindered by various methodological challenges. Specifically, the clinical notes stored in EHRs are composed in a narrative form, making them prone to ambiguous formulations and highly unstructured data presentations, while structured reports commonly suffer from missing and/or erroneous data entries. This inherent complexity poses significant challenges when attempting automated large-scale medical knowledge extraction tasks, necessitating the application of advanced tools, such as natural language processing (NLP), as well as data audit techniques. This work aims to address these obstacles by creating and validating a novel pipeline designed to extract relevant data pertaining to prostate cancer patients. The objective is to exploit the inherent redundancies available within the integrated structured and unstructured data entries within EHRs in order to generate comprehensive and reliable medical databases, ready to be used in advanced research studies. Additionally, the study explores potential opportunities arising from these data, offering valuable prospects for advancing research in prostate cancer.
Download
Paper Nr: 87
Title:

Leveraging Artificial Intelligence for Improved Hematologic Cancer Care: Early Diagnosis and Complications’ Prediction
Authors:

Yousra El Alaoui, Regina Padmanabhan, Adel Elomri, Halima El Omri and Abdelfatteh El Omri
Abstract: Today, medical artificial intelligence (AI) applications are being extensively utilized to enhance the outcomes of clinical diagnosis and overall patient care. This data-driven approach can be trained to account for individuals’ unique characteristics, medical history, ethnicity, and even genetic make-up to obtain accurately tailored treatment recommendations. Given the power of medical AI, the severe nature of hematological malignancies and the related constraints in terms of both time and cost, in this paper, we are investigating the importance of AI applications in hematology management, with an illustration of AI’s role in reducing pre-and post-diagnosis challenges. Insights discussed here are derived based on our experiments on clinical datasets from National Center for Cancer Care & Research (NCCCR), Qatar. Specifically, we developed AI models for blood cancer diagnosis as well as prediction of therapy-induced clinical complications in patients with hematological cancers to facilitate better hospital management and improved cancer care.
Download
Paper Nr: 93
Title:

Underdiagnosed Depression in Older Adults: Analysis of the National Health Survey and Other Aggregate Factors
Authors:

Eduardo H. S. Paraíso, Ariane C. B. da Silva and Cristiane N. Nobre
Abstract: According to the World Health Organization, the total number of people living with depression worldwide is more than 300 million, with depressive disorders ranked globally as the third leading cause of disability. Among older adults, depression is the most common mental illness. This study addressed the cultural stigma surrounding depression in older adults and investigated factors contributing to underdiagnosis and undertreatment. We used data from older adults participating in the National Health Survey (NHS). We applied machine learning algorithms to predict the disorder (Random Forest, Support Vector Machine, Logistic Regression, Gradient Boost, XGBoost, Decision Tree, and Multilayer Neural Network), carefully interpreting the result obtained. Through the interpretability of ML models, the study identified risk factors associated with depression, and using silhouette index and attribute comparison, we found evidence of potential individuals who, although undiagnosed, may be suffering or about to suffer from depression, requiring appropriate care and treatment. This study represents a significant advance in mitigating the impact of cultural stigma on mental health diagnoses in the older population in Brazil.
Download
Paper Nr: 108
Title:

Learning on Forecasting HIV Epidemic Based on Individuals' Contact Networks
Authors:

Chaoyue Sun, Yiyang Liu, Christina Parisi, Rebecca Fisk-Hoffman, Marco Salemi, Ruogu Fang, Brandi Danforth, Mattia Prosperi and Simone Marini
Abstract: Improving the diagnosis of HIV is a fundamental objective of the Ending the HIV Epidemic initiative, as it represents the initial step toward treatment and achieving undetectable status, thereby reducing transmission. To attain these objectives effectively, it is crucial to identify the groups most susceptible to HIV, allowing interventions to be tailored to their specific needs. In this study, we developed a predictive model designed to assess individual HIV risk within a high-risk contact network – predicting treatment or at-risk – leveraging surveillance data collected through routine HIV case interviews in Florida. Unique to our analysis, we explored the incorporation of behavioral network information with Graph Neural Networks to enhance the predictive capacity for identifying individuals within the treatment or intervention categories, when compared to models that mainly consider conventional HIV risk factors. Our deployed Graph Isomorphism Network achieved 77.3% and 73.2% balanced accuracy in inductive and transductive learning scenarios respectively, outperforming the traditional prediction algorithms that do not leverage the network structure. We then used our model to further investigate the importance of demographic and behavioral factors in the HIV risk prediction process. Our findings provide valuable insights for healthcare practitioners and policymakers in their efforts to combat HIV infection.
Download
Paper Nr: 120
Title:

The Beyond 5G (B5G) Era of Next-Generation Digital Networks: Preliminary Study of a Task-Technology Fit (TTF) Model for Remote Robotic Surgery Applications
Authors:

Maradona C. Gatara, Mjumo Mzyece and Sijo J. Parekattil
Abstract: The coming Beyond 5G (B5G) era could mark a paradigm shift towards user-centric Quality of Experience (QoE) centred network architectures. The infusion of QoE user requirements into network architectures will be crucial for future ultra-reliable, ultra-low latency haptic-enabled Internet applications. One such application will be the mission-critical use case of remote (tele-haptic) robotic surgery, signifying a transition towards skillset delivery networks that will augment user task performance experience. In extending traditional Quality of Service (QoS)-oriented networks to user focused QoE and with it, Quality of Task (QoT) components, human users in a global control loop (such as robotic surgeons) will be capable of true-to-life immersive remote task performance through the manipulation of objects in real-time, and of transcending geographical distance. In this preliminary study using data elicited from 20 practising robotic surgeons (n = 20), we examine the emergence of a future B5G network and haptic-enabled Internet of Skills (IoS) architecture, applied to the task-sensitive mission-critical use case of remote (tele-haptic) robotic surgery. We conceptualise and demonstrate the use of non-linear Task-Technology Fit (TTF) predictive modelling to empirically assess this futuristic use case, and in doing so, provide a novel QoE/QoT perspective of future B5G communication networks.
Download
Paper Nr: 121
Title:

Setting a PACS on FHIR
Authors:

Sébastien Jodogne
Abstract: FHIR has quickly emerged as the most important standard for the interoperability of clinical data. FHIR in-cludes support for medical imaging through its ImagingStudy resource that can be used to create mappings between FHIR entities and DICOM studies using the DICOMweb protocol. Unfortunately, there is currently a lack of openly available software implementation combining FHIR with DICOM, which hinders the development of new applications linking clinical data with imaging data. In this paper, the Orthanc server for medical imaging is associated with the HAPI framework in a consistent framework to propose an implementation of the ImagingStudy FHIR resource that is tightly coupled with the content of a DICOM server. The resulting framework is released as free and open-source software, with the goals of promoting support for medical imaging in FHIR, of sharing technological knowledge about medical interoperability, and of providing a test environment for developing new healthcare-related applications.
Download
Paper Nr: 122
Title:

Evaluating the Viability of Neural Networks for Analysing Electromyography Data in Home Rehabilitation: Estimating Foot Progression Angle
Authors:

Finn Siegel, Christian Buj, Ricarda Merfort, Andreas Hein and Frerk Müller-Von Aschwege
Abstract: Intramedullary (IM) nailing is a widely accepted treatment for femoral shaft fractures due to its good healing rate and rapid return to full weight bearing. However, a significant number of patients experience impairments years after treatment. One possible cause is a malrotation of the femur, resulting in altered foot progression angles (FPAs), which can lead to changes in gait or persistent pain. To gain a better understanding of compensation mechanisms and improve rehabilitation strategies, a continuous surface electromyography (EMG) measurement system worn on vastus lateralis (VL) and vastus medialis (VM) is proposed. To test the feasibility of this approach, a study is conducted with healthy participants (N=10) simulating different FPA. The EMG signal was recorded and analysed using a convolutional neural network (CNN). The feasibility study showed promising results, as the CNN could on average achieve a validation accuracy of 74% in classifying FPAs as normal, inward (-15°), or outward (+15°). These results show the potential of using EMG measurements from VL and VM to monitor changes in FPA during rehabilitation. This approach offers the opportunity to increase our understanding of compensatory mechanisms and improve rehabilitation outcomes following malrotation caused by IM nailing.
Download
Paper Nr: 124
Title:

Comparison of Different Data Augmentation Techniques for Improving Epileptic Seizure Detection Based on 3D Acceleration, Heart Rate and Temperature Data
Authors:

Maleyka Seyidova, Jasmin Henze, Arne Pelzer and Beate Rhein
Abstract: Epilepsy, characterized by recurrent seizures, poses a significant risk to an individual’s safety. To mitigate these risks, one approach is to use automated seizure detection systems based on Convolutional Neural Networks (CNN), which rely on large amounts of data to train effectively. However, real-world seizure data acquisition is challenging due to the short and infrequent nature of seizures, resulting in a data imbalance which complicates accurate seizure detection. In this paper, various data augmentation techniques were utilized to increase the amount of training data for CNN, aiming to investigate the potential of these techniques to enhance the performance of the seizure detection algorithm by providing more seizure data. For this purpose, two datasets, a unimodal (3D acceleration) and a multimodal dataset (3D acceleration, heart rate and temperature), were used. To evaluate the effect of the different augmentation techniques, a CNN trained without augmented data was used as a baseline. Experiments showed that data augmentation techniques improved the seizure detection by lowering the baseline’s false alarm rate while maintaining its high sensitivity. The best results were achieved with a combination of Rotation and Permutation in the multimodal dataset and Rotation, as well as Magnitude Warping, in the unimodal dataset.
Download
Paper Nr: 163
Title:

Enhancing Summarization Performance Through Transformer-Based Prompt Engineering in Automated Medical Reporting
Authors:

Daphne van Zandvoort, Laura Wiersema, Tom Huibers, Sandra van Dulmen and Sjaak Brinkkemper
Abstract: Customized medical prompts enable Large Language Models (LLM) to effectively address medical dialogue summarization. The process of medical reporting is often time-consuming for healthcare professionals. Imple-menting medical dialogue summarization techniques presents a viable solution to alleviate this time constraint by generating automated medical reports. The effectiveness of LLMs in this process is significantly influenced by the formulation of the prompt, which plays a crucial role in determining the quality and relevance of the generated reports. In this research, we used a combination of two distinct prompting strategies, known as shot prompting and pattern prompting to enhance the performance of automated medical reporting. The evaluation of the automated medical reports is carried out using the ROUGE score and a human evaluation with the help of an expert panel. The two-shot prompting approach in combination with scope and domain context outperforms other methods and achieves the highest score when compared to the human reference set by a general practitioner. However, the automated reports are approximately twice as long as the human references, due to the addition of both redundant and relevant statements that are added to the report.
Download
Paper Nr: 165
Title:

Chatting for Change: Insights into and Directions for Using Online Peer Support Groups to Interrupt Prolonged Workplace Sitting
Authors:

Ekaterina Uetova, Lucy Hederman, Dympna O’Sullivan, Robert Ross and Marily Oppezzo
Abstract: Prolonged sedentary behavior and insufficient physical activity increase the risk for non-communicable diseases. Online peer support groups, driven by the widespread use of mobile phones and social media, have gained popularity among people seeking health condition management advice. This position paper examines the role of online peer support groups within a behaviour change intervention, MOV’D (Move Often eVery Day), which promotes physical activity and reduces sedentary behavior in the workplace. We conducted a thematic analysis of post-study interviews from two randomized control trials to identify the benefits and limitations of online peer support groups and provide recommendations for improvement. We found that participation in online peer support groups contributes to a sense of belonging and accountability, helps to facilitate the exchange of knowledge and application of the intervention content, and serves as reminders encouraging physical activity throughout the day. However, participants do not always have enough time and cognitive resources to read all the messages and actively participate in the group chats. Individual differences also contribute to a decrease in overall chat activity, as the group chat does not always meet all participant’s preferences and needs.
Download
Paper Nr: 184
Title:

Aggregating Predicted Individual Hospital Length of Stay to Predict Bed Occupancy for Hospitals
Authors:

Mattis Hartwig, Simon Schiff, Sebastian Wolfrum and Ralf Möller
Abstract: This paper addresses the important issue of optimizing hospital bed management by integrating machine learning-based length of stay (LoS) predictions with bed occupancy forecasting. The study primarily utilizes the MIMIC-IV dataset to compare actual bed occupancy against predictions derived from estimated LoS. A novel approach is adopted to translate individual patient LoS predictions into bed occupancy forecasts for the entire hospital. Through various simulations, the paper evaluates the effects of different error margins and patterns in LoS predictions on bed occupancy forecasting accuracy. Key findings reveal that a more symmetric error distribution in LoS predictions significantly enhances the accuracy of bed occupancy forecasts compared to merely reducing the overall prediction error. The paper makes significant contributions to the field. The paper introduces a practical translation scheme from LoS prediction to bed occupancy, which is crucial for hospital administrators in resource planning and management. Also the paper illuminates how various improvements in state-of-the-art LoS prediction models can directly impact the accuracy of bed occupancy forecasts, thereby setting clear objectives for future machine learning research.
Download
Paper Nr: 189
Title:

Observational Study of a Digital Application to Detect Attachment in Dyads Using Markov Chains
Authors:

Sebastian Unger and Thomas Ostermann
Abstract: Attachment is a widely used term and basically refers to a strong emotional relationship that one person develops with another. It is often measured, for example, with the Adult Attachment Interview (AAI), one of the most popular tools, or the Child Attachment Interview (CAI), an adaption of the former. Even though these are two excellent tools for measuring attachment, they are labor-intensive and therefore not suitable for quick use without an adequate training period. Moreover, the mindset towards attachment has changed over time since the development of these tools, meaning that they can still be applied, but only in specific contexts. The digital application "IU" is intended to address these two issues by being easy to learn on the one hand and leaving plenty of freedom for measurement on the other. In this observational study, the interpersonal attachment of dyads captured by the app is interpreted as three-dimensional time series and analyzed based on a Markov chains. This approach shows how interpersonal attachment might be determined according to the homogeneity of the Markov chains, which could probably be improved by capturing other factors such as the interactions of dyads.
Download
Paper Nr: 215
Title:

Exploring the Design of Low-End Technology to Increase Patient Connectivity to Electronic Health Records
Authors:

Rens Kievit, Abdullahi Abubakar Kawu, Mirjam van Reisen, Dympna O’Sullivan and Lucy Hederman
Abstract: The tracking of the vitals of patients with long term health problems is essential for clinicians to determine proper care. Using Patient Generated Health Data (PGHD) communicated remotely allows patients to be monitored without requiring frequent hospital visits. Issues might arise when the communication of data digitally is difficult or impossible due to a lack of access to internet or a low level of digital literacy as is the case in many African countries. The VODAN-Africa project (van Reisen et al., 2021) started in 2020 and has greatly increased the capabilities of clinics in different countries in both Africa and Asia, but currently no systems are in place for the integration of external data from patients with long term health problems. In this article we outline our investigation into methods to increase the connectivity of patients with long term health problems with their clinics, and propose a solution in the form of a data pipeline prototype based on an Interactive Voice Response (IVR) system.
Download
Paper Nr: 220
Title:

Android App for Symptomatic Monitoring of Cervical Dystonia: Design and Usability Study
Authors:

Roland Stenger, Rica Schulze, Sebastian Löns, Tobias Bäumer and Sebastian Fudickar
Abstract: Movement disorders are characterized by paucity or excess of movement. Access to specialists is difficult for patients living in rural areas, making regular visits for symptom monitoring inconvenient. Asynchronous video recording represents a telemedicine approach with temporal freedom but holds the challenge that clinicians and patients can’t interact with each other and thus can’t correct errors, which can lead to a decrease in data quality. This article presents an android application (Move2Screen) that aims to enable asynchronous therapy monitoring and addresses the problem of missing interaction by an implemented video protocol for guided recording to capture visible symptoms in the example of cervical dystonia. The videos can be accessed by clinicians subsequent to an automated upload. A user study of the app was conducted, indicating a strong interest and acceptance rate with a high willingness to use the app. Furthermore, the app can be used to record a standardized data set, which allows a large number of patients to be reached without great effort by clinicians and also provides the possibility of a semi-automated video-based analysis of current symptoms and the longitudinal symptom progression.
Download
Paper Nr: 223
Title:

Automated Medical Text Simplification for Enhanced Patient Access
Authors:

Liliya Makhmutova, Giancarlo Salton, Fernando Perez-Tellez and Robert Ross
Abstract: Doctors and patients have significantly different mental models related to the medical domain; this can lead to different preferences in terminology used to describe the same concept, and in turn, makes medical text often difficult to understand for the average person. However, getting access to a good understanding of patient notes, medical history, and other health-related documents is crucial for patients’ recovery and sticking to a diet or medical procedures. Large language models (LLM) can be used to simplify and summarize text, yet there is no guarantee that the output will be correct and contain all the needed information. In this paper, we create and propose a new multi-modal medical text simplification dataset with pictorial explanations following along the aligned simplified and use it to evaluate the current state-of-the-art large language model (SOTA LLM) for the simplification task for the dataset and compare it to human-written texts. Our findings suggest that the current general-purpose LLMs are still not reliable enough for such in the medical sphere, though they may simplify texts quite well. The dataset and additional materials may be found at https://github.com/ LiliyaMakhmutova/medical texts simplification.
Download
Paper Nr: 230
Title:

Answering the Call to Go Beyond Accuracy: An Online Tool for the Multidimensional Assessment of Decision Support Systems
Authors:

Chiara Natali, Andrea Campagner and Federico Cabitza
Abstract: The research about, and use of, AI-based Decision Support Systems (DSS) has been steadily increasing in the recent years: however, tools and techniques to validate and evaluate these systems in an holistic manner are still largely lacking, especially in regard to their potential impact on actual human decision-making. This paper challenges the accuracy-centric paradigm in DSS evaluation by introducing the nuanced, multi-dimensional approach of the DSS Quality Assessment Tool. Developed at MUDI Lab (University of Milano-Bicocca), this free, open-source tool supports the quality assessment of AI-based decision support systems (DSS) along six different and complementary dimensions: model robustness, data similarity, calibration, utility, data reliability and impact on human decision making. Each dimension is analyzed for its relevance in the Medical AI domain, the metrics employed, and their visualizations, designed according to the principle of vague visualizations to promote cognitive engagement. Such a tool can be instrumental to foster a culture of continuous oversight, outcome monitoring, and reflective technology assessment.
Download
Paper Nr: 232
Title:

Unleashing the Potential of Reinforcement Learning for Personalizing Behavioral Transformations with Digital Therapeutics: A Systematic Literature Review
Authors:

Thure Georg Weimann and Carola Gißke
Abstract: Digital Therapeutics (DTx) are typically considered as patient-facing software applications delivering behavior change interventions to treat non-communicable diseases (e.g., cardiovascular diseases, obesity, diabetes). In recent years, they have successfully developed into a new pillar of care. A central promise of DTx is the idea of personalizing medical interventions to the needs and characteristics of the patient. The present literature review sheds light on using reinforcement learning, a subarea of machine learning, for personalizing DTx-delivered care pathways via self-learning software agents. Based on the analysis of 36 studies, the paper reviews the state of the art regarding the used algorithms, the objects of personalization, evaluation methods, and metrics. In sum, the results highlight the potential and could already demonstrate the medical efficacy. Implications for practice and future research are derived and discussed in order to bring self-learning DTx applications one step closer to everyday care.
Download
Paper Nr: 242
Title:

SynthCheck: A Dashboard for Synthetic Data Quality Assessment
Authors:

Gabriele Santangelo, Giovanna Nicora, Riccardo Bellazzi and Arianna Dagliati
Abstract: In recent years, synthetic data generation has become a topic of growing interest, especially in healthcare, where they can support the development of robust Artificial Intelligence (AI) tools. Additionally, synthetic data offer advantages such as easier sharing and consultation compared to original data, which are subject to patient privacy laws that have become increasingly rigorous in recent years. To ensure a safe use of synthetic data, it is necessary to assess their quality. Synthetic data quality evaluation is based on three properties: resemblance, utility, and privacy, that can be measured using different statistical approaches. Automatic evaluation of synthetic data quality can foster their safe usage within medical AI systems. For this reason, we have developed a dashboard application, in which users can perform a comprehensive quality assessment of their synthetic data. This is achieved through a user-friendly interface, providing easy access and intuitive functionalities for generating reports.
Download
Paper Nr: 250
Title:

A Type of EEG-ITNet for Motor Imagery EEG Signal Classification
Authors:

Maryam Khoshkhooy Titkanlou, Ehsan Monjezi and Roman Mouček
Abstract: The brain-computer interface (BCI) is an emerging technology that has the potential to revolutionize the world, with numerous applications ranging from healthcare to human augmentation. Electroencephalogram (EEG) motor imagery (MI) is among the most common BCI paradigms used extensively in healthcare applications such as rehabilitation. Recently, neural networks, particularly deep architectures, have received substantial attention for analyzing EEG signals (BCI applications). EEG-ITNet is a classification algorithm proposed to improve the classification accuracy of motor imagery EEG signals in a noninvasive brain-computer interface. The resulting EEG-ITNet classification accuracy and precision were 75.45% and 76.43%, using a motor imagery dataset of 29 healthy subjects, including males aged 21-26 and females aged 18-23. Three different methods have also been implemented to augment this dataset.
Download
Paper Nr: 262
Title:

Melanoma Classification Through Deep Ensemble Learning and Explainable AI
Authors:

Wadduwage Shanika Perera, Abm Islam, Van Vung Pham and Min Kyung An
Abstract: Melanoma is one of the most aggressive and deadliest skin cancers, leading to mortality if not detected and treated in the early stages. Artificial intelligence techniques have recently been developed to help dermatologists in the early detection of melanoma, and systems based on deep learning (DL) have been able to detect these lesions with high accuracy. However, the entire community must overcome the explainability limit to get the maximum benefit from DL for diagnostics in the healthcare domain. Because of the black box operation’s shortcomings in DL models’ decisions, there is a lack of reliability and trust in the outcomes. However, Explainable Artificial Intelligence (XAI) can solve this problem by interpreting the predictions of AI systems. This paper proposes a machine learning model using ensemble learning of three state-of-the-art deep transfer Learning networks, along with an approach to ensure the reliability of the predictions by utilizing XAI techniques to explain the basis of the predictions.
Download
Short Papers
Paper Nr: 14
Title:

Predictive Models of Ward Admissions from the Emergency Department
Authors:

Laiene Azkue, Jon Kerexeta, Jorge Sampedro, Moisés Espejo and Nekane Larburu
Abstract: The demand for emergency department (ED) care has increased significantly in recent years, mainly due to factors such as the increase in chronic diseases, aging population and urban population growth. The large influx of patients can lead to overcrowding and resource allocation problems, which impact the quality of care. A new tool to improve patient severity classification systems could improve ED care and avoid inappropriate admissions. Therefore, we propose the development of an artificial intelligence model to predict ED ward admissions. The proposed model uses electronic medical records from the Asunción Klinika in Spain and environmental data. Three models are created at different stages of ED: arrival model which predicts admission upon patient arrival, triage model which predicts admission after clinicians’ triage and the last one, laboratory model which make use of triage model data and laboratory analysis to estimate the risk among the most critical patients. The arrival model achieved an AUC of 0.801, the triage model achieved an AUC of 0.854, and the laboratory model achieved an AUC of 0.781. These models provide valuable information for efficient patient management and resource allocation in the ED, contributing to improved patient care and the adequacy of hospital admissions.
Download
Paper Nr: 25
Title:

Design and Implementation of a Software System for Surveillance of Antibiotics Concentrations in Wastewater
Authors:

Yousuf Al-Hakim, Kosmas Dragos, Kay Smarsly, Silvio Beier and Claudia Klümper
Abstract: Antibiotics are important drugs for treating infectious diseases. The extensive use of antibiotics for human, veterinary, and agricultural purposes has led to the permanent release of antibiotics into the environment, particularly into municipal wastewater. In turn, the widespread release of antibiotics into the environment has led to the emergence of antibiotic-resistant bacteria and antibiotic-resistant genes (collectively referred to as “antibiotic resistance”), which reduce the effectivity of antibiotic treatment. To counteract antibiotic resistance, surveillance of the release of antibiotics into the environment is necessary. Municipal wastewater surveillance may provide insights into the release of antibiotics into the environment. Current municipal wastewater surveillance systems, dedicated to antibiotics concentrations, rely on the ad-hoc use of third-party software, which may compromise the efficiency and user-friendliness of municipal wastewater surveillance systems. Designing software systems dedicated to the surveillance of antibiotics concentrations in municipal wastewater, based on well-established software design concepts, has received scarce research attention. In this study, a software system is proposed, which serves as a technological basis for the surveillance of the concentration of antibiotics in municipal wastewater in an efficient and user-friendly manner. The software system implements well-established software design concepts and is capable of conducting on-demand data analysis, as well as providing various user interfaces. The software system is validated using both data derived from simulations and real-world wastewater data recorded from a wastewater treatment plant. The results showcase the efficiency and user-friendliness of the proposed software system for the surveillance of antibiotics concentrations in municipal wastewater.
Download
Paper Nr: 34
Title:

Triadic Rules for Analysis of Productive and Well-Being Social in Activity-Based Working Environments
Authors:

Thiago H. C. Oliveira, Mark A. J. Song and Luis E. Zárate
Abstract: A longitudinal database records data and its variations over a period of time. The objective of this article is to use this resource, together with the Triadic Concept Analysis theory, to analyze and characterize how employees adapted and felt before, during and after the implementation of an activity-based work environment which is defined as a flexible work setting where employees have the autonomy to choose where they perform their tasks, seeking locations that offer optimal solutions in terms of social interaction, communication, and collaboration. The results seek to support the implementation of this concept, verifying how, and under what conditions, key points of employee experiences vary over time.
Download
Paper Nr: 37
Title:

Spatial-Temporal Visualization Tool for Hospital Support for Infection Spread and Outbreaks
Authors:

Denisse Kim, Manuel Campos, Bernardo Canovas-Segura and Jose M. Juarez
Abstract: Hospital-acquired infections (HAIs) are a major concern today, especially when related to multidrug-resistant bacteria, as they are associated with increases in healthcare costs, prolonged length of stay, and attributable mortality. Tracking the presence of these infections requires interweaving spatial-temporal information from patients and microbiological laboratory results. However, this is normally a manual process and the big amounts of daily clinical data makes it error-prone and time-consuming. In these processes, the temporal dimension is usually taken into account, but not the topology and spatial distribution of patients within a hospital building. Interactive Information Visualization can be used to bring together information from various data sources and to make these spatial-temporal relationships understandable to the human eye. We propose a new interactive visual tool for the exploration of infection spreads within hospitals. The tool presents several connected views to help analyze the epidemic situation of a hospital over time and understand the information contained in the epidemiological indicators.
Download
Paper Nr: 39
Title:

Evaluating Synthetic Data Generation Techniques for Medical Dataset
Authors:

Takayuki Miura, Eizen Kimura, Atsunori Ichikawa, Masanobu Kii and Juko Yamamoto
Abstract: Anticipation surrounds the use of real-world data for data analysis in medicine and healthcare, yet handling sensitive data demands ethical review and safety management, presenting bottlenecks in the swift progression of research. Consequently, numerous techniques have emerged for generating synthetic data, which preserves the features of the original data. Nonetheless, the quality of such synthetic data, particularly in the context of real-world data, has yet to be sufficiently examined. In this paper, we conduct experiments with a Diagonosis Procedure Combination (DPC) dataset to evaluate the quality of synthetic data generated by statistics-based, graphical model-based, and deep neural network-based methods. Further, we implement differential privacy for theoretical privacy protection and assess the resultant degradation of data quality. The findings indicate that a statistics-based method called Gaussian Copula and a graphical-model-based method called AIM yield high-quality synthetic data regarding statistical similarity and machine learning model performance. The paper also summarizes issues pertinent to the practical application of synthetic data derived from the experimental results.
Download
Paper Nr: 42
Title:

Monitoring Pain in Patients with Chronic Pain with a Wearable Wristband in Daily Life: A Pilot Study
Authors:

E. Pattyn, E. Vergaelen, E. Lutin, R. Van Stiphout, H. Davidoff, W. De Raedt and C. Van Hoof
Abstract: Chronic pain is a complex and personal condition that imposes a substantial burden on both individuals and society. Potentially, wearable technology could enable continuous monitoring of pain in real-world settings, offering insights into the complex relationship between physiological states and chronic pain. In this pilot study, we evaluated the practicability of collecting physiological data, from ten individuals with chronic pain and ten healthy controls, using wearable wristbands and digital pain diaries for one week in their everyday lives. Additionally, we trained various machine learning classifiers to classify pain levels and evaluated which feature modalities, e.g., heart rate-derived features, yielded the highest balanced accuracy. Our results demonstrated satisfactory data quantity, with wristband data being available for patients and controls approximately 92% to 82% of the time, and data quality, with high-quality physiology ranging from 80% to 72% for the respective groups. The median balanced accuracies in distinguishing pain intensity classes ranged between 0.27 and 0.40. Furthermore, we found that individual modalities did not outperform the combined modalities. Nonetheless, further research with larger sample sizes is necessary to elucidate these relationships and improve pain management strategies for individuals with chronic pain.
Download
Paper Nr: 45
Title:

GAN-Based Data Augmentation for Improving Biometric Authentication Using CWT Images of Blood Flow Sounds
Authors:

Natasha Sahare, Patricio Fuentealba, Rutuja Salvi, Anja Burmann and Jasmin Henze
Abstract: Biometric identification allows to secure sensitive information. Since existing biometric traits, such as finger-prings, voice, etc. are associated with different limitations, we exemplified the potential of blood flow sounds for biometric authentication in previous work. Therefore, we used measurements from seven different users acquired with a custom-built auscultation device to calculate the spectrograms of these signals for each cardiac cycle using continuous wavelet transform (CWT). The resulting spectral images were then used for training of a convolutional neural network (CNN). In this work, we repeated the same experiment with data from twelve users by adding more data from the original seven users and data from five more users. This lead to an imbalanced dataset, where the amount of available data for the new users was much smaller, e.g., U1 had more than 900 samples per side whereas the new user U9 had less than 100 samples per side. We experienced a lower performance for the new users, i.e. their sensitivity was 18-21% lower than the overall accuracy. Thus, we examined whether the augmentation of data leads to better results. This analysis was performed using generative adversarial networks (GANs). The newly generated data was then used for training of a CNN with several different settings, revealing the potential of GAN-based data augmentation for increasing the accuracy of biometric authentication using blood flow sounds.
Download
Paper Nr: 67
Title:

Bed Management System Implementation: Experimental Study
Authors:

Flannagán Noonan, Michelle Hanlon, Juncal Nogales, Ciarán Doyle, Eilish Broderick and Joseph Walsh
Abstract: Many hospitals today use bed management systems that are primarily manual and paper-based. This inhibits efficiency and informed decision making, as communication is constrained. Hence these systems are essentially memoryless as lessons learned reside with individuals but are lost to the organisation as a whole. Electronic systems that can capture and record checkpoints on the patient pathway allow that data to be analysed. This can help with improving efficiency and prediction, allowing “what if” scenarios to be examined with data to support it. This paper presents the outcome of developing a bed management system and deploying it in a hospital for a live trial over a period of approximately three months. It also highlights improvements suggested through system usage over the period of the deployment and presents a novel efficiency measure.
Download
Paper Nr: 70
Title:

Leveraging Health Informatics to Enhance Outpatient Chemotherapy Operations Management
Authors:

Majed Hadid, Adel Elomri and Regina Padmanabhan
Abstract: The rise in demand for cancer care services, particularly outpatient chemotherapy, highlights the importance of improving the management of outpatient chemotherapy operations (OCOM). Despite the numerous studies addressing OCOM issues, the existing literature has mostly focused on problem-driven research. In this study, we aimed to utilize data-driven research to identify opportunities for improvement and address research challenges. To achieve this goal, we collected extensive operational data from a large chemotherapy center and performed a thorough analysis. Our findings revealed four key research challenges, including the prediction of length of stay, change in patient drug posting weight, delay in appointment admission, and stochasticity in drug administration duration. To address these challenges, we developed two machine learning models to predict these outcomes, utilizing 15 features and highlighting the most important features. Our results showed an efficient performance in predicting the outcomes using the XGBoost model, emphasizing the potential of data-driven research in improving OCOM.
Download
Paper Nr: 76
Title:

Using Data Mining Techniques to Understand Patterns of Suicide and Reattempt Rates in Southern Brazil
Authors:

Caibe A. Pereira, Rômulo C. R. Peixoto, Manuella P. Kaster, Mateus Grellert and Jônata T. Carvalho
Abstract: Suicide is a multifactorial, complex condition and one of the leading global causes of death, with suicide attempt as the main risk factor. To this day, studies have shown relevant indicators that help identify people with risk of committing suicide, but the literature still lacks comprehensive studies that evaluate how different risk factors interact and ultimately affects the suicide risk. In this paper, we aimed to identify patterns in data from the Brazilian Unified Health System – SUS, from 2009 to 2020, of individual reports of suicide attempts and suicide deaths in the Brazilian Southern States, integrating those with a database of the healthcare infrastructure. We framed the problem as a classification task for each micro-region to predict suicide and reattempt rate as low, moderate, or high. We developed a pipeline for integrating, cleaning, and selecting the data, and trained and compared three machine learning models: Decision Tree, Random Forest, and XGBoost, with approximately 97% accuracy. The most important features for predicting suicide rates were the number of mental health units and clinics, and for both suicide and reattempts were the number of physicians and nurses available. This novel result brings valuable knowledge on possible directions for governmental investments in order to reduce suicide rates.
Download
Paper Nr: 79
Title:

Naïve Bayes as a Probabilistic Tool for Monitoring the Health Status of Chronic Patients
Authors:

Laura Teresa Martínez-Marquina, María Teresa Jurado-Camino, Isabel Caballero-López-Fando and Inmaculada Mora-Jiménez
Abstract: Chronic diseases have emerged as a pervasive global health concern, standing as a leading cause of mortality. Among these, prevalent conditions encompass diabetes, hypertension, congestive heart failure and chronic obstructive pulmonary disease. The large amount of data in Electronic Health Records is being exploited by machine learning schemes to design clinical decision support systems, usually of limited practical application because of lack of transparency. To overcome this issue and given the dynamic nature of the health-status over time, we propose here a patient health monitoring scheme based on a Näive Bayes approach because of its interpretability, minimal computational cost, and efficient handling of high-dimensional and unbalanced data. Our approach considers clinical codes (diagnosis and drugs) on real data collected by a Spanish hospital and provides a probability score for different chronic health-statuses. A gender-based approach has also been explored, exhibiting promising performance when there is a significant patient population for each sex. We conclude that pharmacological codes are more informative, although the best performance was obtained by using all the clinical codes and demographic features. Though a more exhaustive study on patient monitoring is necessary, the proposed NB scheme can be considered a proof of concept which has demonstrated to be a valuable tool and easily interpretable method.
Download
Paper Nr: 88
Title:

Antibiotic Prescriptions Before, During and after the Corona Pandemic in Schleswig-Holstein with Prescription Data from 2017 till 2023
Authors:

Reinhard Schuster, Timo Emcke, Vera Ries, Eva von Arnstedt and Mareike Burmester
Abstract: The ongoing COVID-19 pandemic threatens the health of humans, causes great economic losses and may disturb the stability of the societies and is a major challenge for physicians, politicians, scientists and many other groups. The article focuses on patients with antibiotic prescriptions and considers their risks in comparison to all patients. Time series are analyzed starting from the pre-Corona period till today. Mathematical analysis can be used to understand aspects of the dynamics of epidemics and to improve strategies, i. e. regarding effects of antibiotic stewardship programs or reaction to drug availability constraints.
Download
Paper Nr: 92
Title:

i-SART: An Intelligent Assistant for Safety Analysis in Radiation Therapy
Authors:

Natalia Silvis-Cividjian, Yijing Zhou, Anastasia Sarchosoglou and Evangelos Pappas
Abstract: Along with surgery and chemotherapy, radiation therapy (RT) is a very effective method to treat cancer. The process is safety-critical, involving complex machines, human operators and software. A proactive hazard analysis to predict what can go wrong in the process is therefore crucial. Failure Modes and Effect Analysis (FMEA) is one of the methods widely used for risk assessment in healthcare. Unfortunately, the available resources and FMEA expertise strongly vary across different RT organizations worldwide. This paper describes i-SART, an interactive web-application that aims to close the gap by bringing together best practices in conducting a sound RT-FMEA. Central is a database that at present contains approximately 420 FMs collected from existing risk assessments and cleaned from ambiguities and duplicates using NLP techniques. Innovative is that the database is designed to grow, due to both user input and generative AI algorithms. This is work in progress. First experiments demonstrated that using machine learning in building i-START is beneficial. However, further efforts will be needed to search for better solutions that do not require human judgment for validation. We expect to release soon a prototype of i-SART that hopefully will contribute to the global implementation and promotion of safe RT practices.
Download
Paper Nr: 95
Title:

Segmentation and Classification of Dental Caries in Cone Beam Tomography Images Using Machine Learning and Image Processing
Authors:

Luiz Guilherme Kasputis Zanini, Izabel Regina Fischer Rubira-Bullen and Fátima de Lourdes dos Santos Nunes
Abstract: Dental caries are caused by bacterial action that demineralizes tooth enamel and dentin. It is a serious threat to oral health and potentially leads to inflammation and tooth loss if not adequately treated. Cone Beam Computed Tomography (CBCT), a three-dimensional (3D) imaging technique used in dental diagnosis and surgical planning, can potentially contribute to detection of caries. This study aims at developing a computational method to segment and classify caries in CBCT images. The process involves data preparation, segmentation of caries regions, extraction of relevant features, feature selection, and training machine learning algorithms. We evaluated our method performance considering different stages of caries severity based on the International Caries Detection and Assessment System scale. The best results were achieved using a Gaussian filter with a multimodal threshold with a convex hull for the region of interest segmentation, feature selection via Random Forest, and classification using a model based on k-nearest neighbors algorithm. We achieved outcomes with an accuracy of 86.20%, a F1-score of 86.18%, and a sensitivity of 83.35% in multiclass classification. These results show that our approach contributes to the early segmentation and classification of dental caries, thereby improving oral health outcomes and treatment planning.
Download
Paper Nr: 106
Title:

Unpacking the Clinical Burden of Leukemia in GCC: Implications for Patient Care
Authors:

Hesham Ali Behary Aboelkhir, Yousra El Alaoui, Regina Padmanabhan, Adel Elomri, Halima El Omri and Abdelfatteh El Oomri
Abstract: Cancer constitutes a substantial global health challenge, which is poised to intensify primarily due to the growing elderly population globally. Leukemia, being a type of hematological cancer, presents unique diagnostic complexities compared to solid cancers, contributing to elevated levels of morbidity and mortality across various regions worldwide, resulting in a substantial clinical burden. Employing data sourced from the WHO Global Health Expenditure Database and the Institute for Health Metrics and Evaluation (IHME) Global Burden of Disease for the year 2019, this study undertakes an analysis of the prevalence, Years of Life Lost (YLLs), Years Lived with Disability (YLDs), Disability-Adjusted Life Years (DALYs), and healthcare expenditure in Gulf Cooperation Council (GCC) nations in comparison to the global figures.
Download
Paper Nr: 114
Title:

Stroke Prehospital Decision Support Systems Based on Artificial Intelligence: Grey Literature Scoping Review
Authors:

Hoor Jalo, Eunji Lee, Mattias Seth, Anna Bakidou, Minna Pikkarainen, Katarina Jood, Bengt Arne Sjöqvist and Stefan Candefjord
Abstract: Stroke is a leading cause of mortality and disability worldwide. Therefore, there is a growing interest in prehospital point-of-care stroke clinical decision support systems (CDSSs), which with improved precision can identify stroke and decrease the time to optimal treatment, thereby improving clinical outcomes. Artificial intelligence (AI) may be a route to improve CDSSs for clinical benefit. Deploying AI in the area of prehospital stroke care is still in its infancy. There are several existing systematic and scoping reviews summarizing the progress of AI methods for stroke assessment. None of these reviews include grey literature, which could be a valuable source of information, especially when analysing future research and development directions. This paper aims to use grey literature to investigate stroke assessment CDSSs based on AI. The study adheres to PRISMA guidelines and presents seven records showcasing promising technologies. These records included three clinical trials, two smartphone applications, one master thesis and one PhD dissertation, which identify electroencephalogram (EEG), video analysis and voice and facial recognition as potential data sources for early stroke identification. The integration of these technologies may offer the prospect of faster and more accurate CDSSs in the future.
Download
Paper Nr: 118
Title:

Automatic Scoring of Shulman's Clock Drawing Dementia Test
Authors:

Bianca Suermann, Miguel Schulz, Klaus Brinker and Markus Weih
Abstract: With dementia currently being one of the biggest healthcare challenges, an improvement in diagnosis represents a substantial improvement for the patients and medical experts. A frequently used diagnosis tool is the clock drawing test (CDT), cognitive short test typically conducted with pencil and paper and manually scored by a medical professional. This paper introduces a transparent approach for software-assisted scoring and screening of CDT, using a combination of deep learning elements and standard image recognition techniques. Unlike an end-to-end approach, our strategy involves dividing the process into distinct subprocesses. This division ensures that intermediate results are readily available throughout, establishing a robust and transparent foundation for the diagnostic process. A dataset containing 1236 CDT-scans is used for evaluating our algorithm’s ability to score the result into a category from 1 to 6 and the ability to classify pass or fail is assessed. Based on the results a component-wise software-assisted approach to CDT scoring seems to be a viable alternative to end-to-end systems.
Download
Paper Nr: 123
Title:

Assessment of the Relationship Between Attribute Coding and the Interpretability of Machine Learning Models: An Analysis in the Context of Children and Adolescents with Depression
Authors:

Ludmila B. S. Nascimento, Marcelo de S. Balbino, Maycoln L. M. Teodoro and Cristiane N. Nobre
Abstract: Depression is a global public health challenge that affects approximately 300 million people. Artificial Intelligence and Machine Learning have revolutionized the healthcare sector, allowing the development of models to diagnose depression. Tabular data, shared in healthcare, requires preprocessing, including encoding categorical attributes into numeric values, as many Machine Learning algorithms only support numeric data. This study aims to investigate different coding methods for non-ordinal nominal categorical attributes in a dataset related to depression in children and adolescents suffering from Major Depressive Disorder (MDD). The comparison results revealed that the XGBoost algorithm with the Hash Encoding, Customized One Hot, Frequency, and Dummy coding techniques were more effective for the analyzed data set. However, not all of these encodings are interpretable. These results provide significant insights, highlighting the importance of choosing appropriate coding methods to improve the accuracy of Machine Learning models and the interpretability of these models in healthcare.
Download
Paper Nr: 125
Title:

Covid-19 Impact on Standard Coding Systems Update
Authors:

Elena Cardillo, Maria Teresa Chiaravalloti and Erika Pasceri
Abstract: The outbreak of Covid-19 pandemic has sped up many healthcare processes and practices. Both stakeholders and standard organizations and authorities had to quickly implement new guidelines and codes to uniquely identify the disease and all the related healthcare data. The object of this work is to study the impact of the Covid-19 pandemic on clinical coding systems, in terms of updates and introduction of new specific codes for the identification of the SARS-CoV-2 virus, with the aim of allowing a better description of the disease and interoperability of the clinical data. The analysis is focused on ICD, SNOMED CT, LOINC, ATC as coding systems either included into the Italian EHR regulation or widely used internationally. Results show that coding systems that created a plenty of new codes for Covid-19 have: i) a flexible structure; ii) a speed process for updates; iii) a large user community for inputs. Others instead demonstrated in this circumstance that they are limited by hierarchical structures or excessively cumbersome updating processes, which conflict with the flexibility required to standards to represent the evolution of clinical knowledge. This is especially true in exceptional situation like the pandemic one.
Download
Paper Nr: 127
Title:

Skimming of Electronic Health Records Highlighted by an Interface Terminology Curated with Machine Learning Mining
Authors:

Mahshad Koohi H. Dehkordi, Navya Martin Kollapally, Yehoshua Perl, James Geller, Fadi P. Deek, Hao Liu, Vipina K. Keloth, Gai Elhanan and Andrew J. Einstein
Abstract: Clinical notes in Electronic Health Records (EHRs) contain large amounts of nuanced information. Healthcare professionals, e.g., clinicians, routinely review numerous EHR notes, further burdening their busy schedules. To capture the essential content of a note, they often quickly review its content, which can contribute to missing critical clinical information. Highlighting important content of EHRs enable clinicians to fast skim by reading only the highlighted words. Furthermore, effective highlighting of EHRs will support new research and interoperability. In this paper, we design a Cardiology Interface Terminology (CIT) dedicated for the application of highlighting cardiology EHRs to support their fast skimming. Once successful, Transfer Learning can be used to design an interface terminology for other specialties. In EHRs, we observe phrases of fine granularity containing SNOMED CT concepts. In our previous work, we extract such phrases from EHR notes to be considered as CIT concepts. This early CIT serves as training data for Machine Learning (ML) techniques, further enriching CIT and improving EHR highlighting. We describe the methodology and results of curating CIT with ML techniques. Furthermore, we introduce the coverage and breadth metrics for measuring the efficacy of highlighting EHRs, and discuss future improvements, enhancing the coverage of highlighted important content.
Download
Paper Nr: 129
Title:

Process-Aware Analysis of Treatment Paths in Heart Failure Patients: A Case Study
Authors:

Harry H. Beyel, Marlo Verket, Viki Peeva, Christian Rennert, Marco Pegoraro, Katharina Schütt, Wil M. P. van der Aalst and Nikolaus Marx
Abstract: Process mining in healthcare presents a range of challenges when working with different types of data within the healthcare domain. There is high diversity considering the variety of data collected from healthcare processes: operational processes given by claims data, a collection of events during surgery, data related to pre-operative and post-operative care, and high-level data collections based on regular ambulant visits with no apparent events. In this case study, a data set from the last category is analyzed. We apply process-mining techniques on sparse patient heart failure data and investigate whether an information gain towards several research questions is achievable. Here, available data are transformed into an event log format, and process discovery and conformance checking are applied. Additionally, patients are split into different cohorts based on comorbidities, such as diabetes and chronic kidney disease, and multiple statistics are compared between the cohorts. Conclusively, we apply decision mining to determine whether a patient will have a cardiovascular outcome and whether a patient will die.
Download
Paper Nr: 130
Title:

Exploring the Power of Triple Crown Process Modeling in Healthcare: Sepsis Case
Authors:

Camelia Maleki and Frederik Gailly
Abstract: Effective process modeling plays a pivotal role in optimizing patient care processes within the continually evolving healthcare landscape. This paper focuses on the application of the Triple Crown standard, which encompasses the Business Process Model and Notation (BPMN), Case Management Model and Notation (CMMN), and Decision Model and Notation (DMN), within the context of the sepsis diagnosis process. Through an in-depth exploration of this case study, the paper uncovers the immense potential of these standards in empowering healthcare practitioners to streamline workflows, enhance decision-making at critical junctures, and ensure the delivery of the highest quality care despite the diverse challenges inherent in patient care processes. By dissecting key dimensions such as flexibility, data and information flow, complexity management, and decision points, this study provides valuable insights into how the Triple Crown approach can significantly enhance patient care process models.
Download
Paper Nr: 131
Title:

Improving Self-Report Diaries: A Respondent-Centric Design Approach
Authors:

Rosaline Barendregt, Yngve Lamo and Barbara Wasson
Abstract: Medical self-report plays an indispensable role in healthcare, capturing vital subjective data from health and everyday life contexts. While considerable research has been dedicated to trialling self-report to ensure their clinical validity, there has been less focus on understanding user behaviour in self-reporting and on devising strategies to optimise the use of these self-report tools. The traditional approach to self-report design has been largely information-centric, relegating patients to the role of passive information providers. This can lead to a significant respondent burden due to the retrospective nature of the questions and the inherent challenges in data provision. Recently, the Respondent-centric Design (RxD) Framework has been suggested as an approach to bridge clinical needs and patient’s needs within the self-report design process. In this paper we report on the use of the RxD framework for redesigning a headache diary. Our experience on using RxD provides insight into its potential to reshape self-report design, RxD steered our focus during the redesign to consider respondent perspectives more thoroughly. The redesigned headache diary received positive feedback, both from users and experts, and the evaluation suggests improved adherence and higher respondent acceptance.
Download
Paper Nr: 132
Title:

Business Intelligence Enhancements to EDC for Clinical Trial Management
Authors:

Alessio Bottrighi, Elisa Gandini, Stefano Nera, Luca Piovesan, Erica Raina and Paolo Terenziani
Abstract: We describe our experience in enhancing Electronic Data Capture systems with Business Intelligence facilities, to provide additional decision support facilities. In particular, with our framework, we support analytical intelligent reporting, visualization and querying to improve managerial control in trial conduct. In this paper, we discuss a principled methodology, in which the analytical intelligent extension is based on an explicit conceptual modelling of a multi-dimensional view of the clinical trials. While our approach is general, we have developed it in the context of long-term cooperation with the Italian Lymphoma Foundation (FIL), managing dozens of clinical trials distributed in many national (Italian) and international institutes.
Download
Paper Nr: 133
Title:

2ViTA-B Cognitive: A Virtual Assistant for Cognitive Rehabilitation
Authors:

Nicoletta Balletti, Antonella Cascitelli, Patrizia Gabrieli, Aldo Lazich, Gianluca Maria Marcilli, Marco Notarantonio, Rocco Oliveto and Daniela Scognamiglio
Abstract: We present 2ViTA-B Cognitive, an advanced virtual assistant designed to effectively address emotional functioning disorders and enhance emotional well-being. The system is carefully crafted to engage with users and positively influence their emotions, supporting cognitive rehabilitation in both real and virtual environments. A controlled experiment has been conducted to evaluate the benefits of the proposed system. The results provide valuable insights into the potential benefits of immersive virtual reality interventions for improving emotional well-being and cognitive functions. These findings suggest promising avenues for advancements in therapeutic practices within this field.
Download
Paper Nr: 150
Title:

Safeguarding Ethical AI: Detecting Potentially Sensitive Data Re-Identification and Generation of Misleading or Abusive Content from Quantized Large Language Models
Authors:

Navya Martin Kollapally and James Geller
Abstract: Research on privacy-preserving Machine Learning (ML) is essential to prevent the re-identification of health data ensuring the confidentiality and security of sensitive patient information. In this era of unprecedented usage of large language models (LLMs), LLMs carry inherent risks when applied to sensitive data, especially as LLMs are trained on trillions of words from the internet, without a global standard for data selection. The lack of standardization in training LLMs poses a significant risk in the field of health informatics, potentially resulting in the inadvertent release of sensitive information, despite the availability of context-aware redaction of sensitive information. The research goal of this paper is to determine whether sensitive information could be re-identified from electronic health records during Natural Language Processing (NLP) tasks such as text classification without using any dedicated re-identification techniques. We performed zero and 8-shot learning with the quantized LLM models FLAN, Llama2, Mistral, and Vicuna for classifying social context data extracted from MIMIC-III. In this text classification task, our focus was on detecting potential sensitive data re-identification and the generation of misleading or abusive content during the fine-tuning and prompting stages of the process, along with evaluating the performance of the classification.
Download
Paper Nr: 151
Title:

The Impact of Class Weight Optimization on Improving Machine Learning Outcomes in Identifying COVID-19 Specific ECG Patterns
Authors:

Sara Khan, Walaa N. Ismail, Shada Alsalamah, Ebtesam Mohamed and Hessah A. Alsalamah
Abstract: The Covid-19 pandemic has resulted in 550 million cases and 6.3 million fatalities, with the virus severely affecting the lungs and cardiovascular system. A study utilizes a VGG16 model adapted for a 12-Lead ECG Image database to assess the disease’s impact on cardiovascular health. The research addresses the challenge of data imbalance by experimenting with different training approaches: using balanced datasets, imbalanced datasets, and class weight adjustments for imbalanced datasets. These models are designed for a three-class multiclass classification of ECG images: Abnormal, Covid-19, and Normal categories. Performance evaluations, including accuracy scores, confusion matrices, and classification reports, show promising results. The model trained on a balanced dataset achieved a 90% accuracy rate. When trained on an imbalanced dataset, the accuracy dropped to 82%. However, with class weight adjustments, the accuracy rebounded to 87%. The study proves that the adapted VGG16 model can effectively handle both balanced and imbalanced datasets. Further testing and enhancements can be carried out using additional datasets, making it a valuable tool for understanding the cardiovascular implications of Covid-19.
Download
Paper Nr: 155
Title:

Evaluation of "Speech System" and "Skill": An Interaction Paradigm for Speech Therapy
Authors:

Vita Santa Barletta, Miriana Calvano, Antonio Curci, Alessandro Pagano and Antonio Piccinno
Abstract: Speech therapy is the medical field in which speech impairments are treated. They concern the inability of individuals to adequately enunciate words, construct, elaborate and appropriate sentences when speaking, and overall lack linguistic skills. Although speech impairments can emerge throughout different stages of life, the most common period of time in which they are encountered is childhood. The professionals in this medical field use to treat these impairments with the employment of therapies that are carried out over an extended period of time. This research work aims at proposing and evaluating through a user study a new interactive paradigm that involves ”Speech System”, a web-application, and ”Skill”, a skill for Amazon Alexa. The objective consists in determining the practical feasibility of the solution and investigate the consequences that the use of technology brings to the world of speech therapy. The advantages and disadvantages of the interactive paradigm in question are explored and discussed to define the direction of the next steps in this field.
Download
Paper Nr: 158
Title:

Strategic Oversight Across Real-World Health Data Initiatives in a Complex Health Data Space: A Call for Collective Responsibility
Authors:

Lotte Geys and Liesbet M. Peeters
Abstract: Reusing real-world health data is useful, but challenging. Multiple initiatives exist and more are continuously arising to overcome these challenges, but the strategic oversight across these initiatives is lacking, which leads to a fragmented ecosystem. An overview of which initiatives that work on unlocking real-world health data, making this data accessible for research and/or innovation and/or policy and getting an idea about which aspect of the ecosystem the initiatives are working on would be very helpful. It could help in figuring out how initiatives can work in synergy in order that consortia can be formed more efficiently. We tried to create an overview, resulting in a static list, but have thereby run into many problems and difficulties and have noticed that the information is even more scattered than expected, and often ambiguous and unclear. This paper highlights the need for strategic oversight in our complex health data space, defines key challenges and focuses on solutions and strategies for overcoming these challenges, and aims to guide the future of health data research and innovation on a global scale, offering a valuable resource for stakeholders in the field.
Download
Paper Nr: 162
Title:

Comparative Experimentation of Accuracy Metrics in Automated Medical Reporting: The Case of Otitis Consultations
Authors:

Wouter Faber, Renske Eline Bootsma, Tom Huibers, Sandra van Dulmen and Sjaak Brinkkemper
Abstract: Generative Artificial Intelligence (AI) can be used to automatically generate medical reports based on transcripts of medical consultations. The aim is to reduce the administrative burden that healthcare professionals face. The accuracy of the generated reports needs to be established to ensure their correctness and usefulness. There are several metrics for measuring the accuracy of AI generated reports, but little work has been done towards the application of these metrics in medical reporting. A comparative experimentation of 10 accuracy metrics has been performed on AI generated medical reports against their corresponding General Practitioner’s (GP) medical reports concerning Otitis consultations. The number of missing, incorrect, and additional statements of the generated reports have been correlated with the metric scores. In addition, we introduce and define a Composite Accuracy Score which produces a single score for comparing the metrics within the field of automated medical reporting. Findings show that based on the correlation study and the Composite Accuracy Score, the ROUGE-L and Word Mover’s Distance metrics are the preferred metrics, which is not in line with previous work. These findings help determine the accuracy of an AI generated medical report, which aids the development of systems that generate medical reports for GPs to reduce the administrative burden.
Download
Paper Nr: 173
Title:

Towards Using Synthetic User Interaction Data in Digital Healthcare Usability Evaluation
Authors:

Bilal Maqbool, Laoa Jalal and Sebastian Herold
Abstract: Effective usability evaluation of user interface (UI) designs is essential. Particularly in digital healthcare, frequently involving relevant user groups in usability evaluations is not always possible or is ethically questionable. On the other hand, neglecting the perspectives of such groups can lead to UI designs that fail to be inclusive and adaptable. In this paper, we outline an initial idea to utilize artificial intelligence methods to simulate mobile user interface interactions of such user groups. The goal is to support software developers and designers with tools that show them how users of certain user groups might interact with a user interface under development and show potential issues before actual, more expensive usability evaluations are conducted. We present a study that employs synthetic representations of user interactions with UI elements based on a small sample of real interactions. This synthetic data was then used to train a classification model predicting whether real user interactions were from younger or elderly persons. The good performance of this model provides evidence that synthetic user interface interactions might be accurate enough to feed into imitation learning approaches, which, in turn, could be the foundation for the desired tool support.
Download
Paper Nr: 176
Title:

Approach and Method for Bayesian Network Modelling: The Case for Pregnancy Outcomes in England and Wales
Authors:

Scott McLachlan, Bridget J. Daley, Sam Saidi, Evangelia Kyrimi, Kudakwashe Dube, Crina Grossan, Martin Neil, Louise Rose and Norman E. Fenton
Abstract: For predicting and reasoning about outcomes of specific medical condition Bayesian Networks (BNs) can provide significant benefits over traditional statistical prediction models. However, developing appropriate and accurate BNs that incorporate key causal aspects of the condition is challenging and time-consuming. This work introduces a novel development approach, merging expert elicitation, literature knowledge, and national health statistics that enables such BNs to be developed efficiently. The approach is applied to build a BN for pregnancy complications and outcomes in England and Wales using 2021 data. The BN showed comparable predictive performance against logistic regression and nomograms, but additionally provides powerful support for decision-making and risk assessment across diverse pregnancy-related conditions and outcomes.
Download
Paper Nr: 177
Title:

A Concept for Daily Assessments During Nutrition Intake: Integrating Technology in the Nursing Process
Authors:

Sandra Hellmers, Tobias Krahn, Martina Hasseler and Andreas Hein
Abstract: The nursing process involves a cyclic sequence of functional and cognitive assessments and diagnosis, care planning, implementation, and evaluation of care. Ideally, this process should be performed regularly and documented in a standardized nursing language. However, due to the high workload of nurses, this approach is not systematically followed. Therefore, we developed a concept that enables a daily, technology-supported assessment during the activity of nutrition intake. For this purpose, we used camera-based body tracking to derive the hand and relevant object trajectories to analyze the movements regarding assistance needs and functional changes over time. We tested the approach of using generative AI to create training data sets. Our feasibility study has shown that trajectories can be derived and analyzed regarding assistance requirements. Although the quality is not yet satisfactory, generative AI can be used to create training data. Considering the rapid pace of further developments in generative AI, the approach seems to be promising. In conclusion, we believe that the technical support and documentation of the nursing process have the potential to increase the quality of care while reducing the workload of nurses.
Download
Paper Nr: 181
Title:

An Empirical Analysis of Undergraduate Information Systems Security Behaviors
Authors:

José A. García-Berná, Sofia Ouhbi, José L. Fernández-Alemán and Ana B. Sánchez-García
Abstract: The growing concern within healthcare organizations about the privacy of personal health data emphasizes how critical it is to address security and privacy issues, especially for nurses who handle sensitive data on a daily basis. In order to understand the habits and awareness of nursing degree students with regard to the protection of patients’ personal data, this study focuses on evaluating their security behavior. The purpose of the 21-item questionnaire was to provide insight into the data security practices of 95 fourth-year nursing students and 167 second-year nursing students. The findings indicated that students in their second year of study had more robust password practices than those in their fourth year, who in turn showed a propensity to click on potentially hazardous links more frequently. In light of the fact that nursing professionals will unavoidably work with large amounts of medical data in their future positions, the findings point to the necessity of raising awareness of and providing education on data protection.
Download
Paper Nr: 182
Title:

Quantified Health: A Feasibility Study on a Sensor-Based Feedback and Assistance System in Cardiology, Oncology and Orthopaedics
Authors:

Anne Grohnert, Michael John, Benny Häusler, Christian Giertz, Mirko Wolschke, Jana Liebach, Rona Reibis, Anne Klemmer, Lisa Konrad, Silke Kollath and Jan C. Zoellick
Abstract: This paper reports the results of the Quantified Health project that developed a complex, digitally supported intervention. The project provides insights into how a sensor-based system can be organizationally integrated into the existing workflows of everyday treatment. The concluding pilot study took place in three medical facilities and addressed patients of orthopaedic, oncologic and cardiologic diseases in an in- or outpatient therapeutic setting. As a study result from the user’s perspective, it is very appreciated to objectify the patient’s health related behavior. Care providers considered it positive that they received more data from patients’ everyday lives and that the improved data situation can lead to more sustainable care. On the other side, the time required to integrate a new digital application into the tightly scheduled daily treatment routine was perceived as a hindering factor. Nevertheless, the results of the study show that a more generic sensor-based assistance system could be used for different diseases and cross sectoral. Furthermore, the constant contact with therapists increases patients’ motivation to engage in health-preserving activities (self-regulation).
Download
Paper Nr: 185
Title:

Technology Support System and Review Process for a Decentralized Clinical Trial: Trials@Home, RADIAL DCT as Case Study
Authors:

Sten Hanke, Dimitrios Giannikopoulos, Hannes Hilberger, Theresa Weitlaner and Bernhard Neumayer
Abstract: Decentralized Clinical Trials (DCTs) revolutionize clinical research by leveraging digital technologies to decentralize various aspects inherent in the traditional clinical trial process like the need for patients’ physical presence. DCTs integrate virtual and remote elements for assessments, data collection, and monitoring, prioritizing convenience. However, the integration of diverse stakeholders and technologies poses challenges in delivering timely and effective solutions across all trial sites. Addressing this requires the establishment of a robust technology support system tailored to meet the unique demands of decentralization. This paper outlines the requirements for such a system and shares initial insights gained through the learning process. This system combines a Wiki-like knowledge base with a ticketing system for handling support requests, enabling the creation of topic-specific tickets and ensuring that queries are directed to the appropriate support agents swiftly. The implemented helpdesk system in the RADIAL study exemplifies how combining a comprehensive information resource with a responsive ticketing system not only streamlines supporting processes but also significantly enhances response efficiency and the overall user experience in DCTs. This integrated approach is pivotal in managing the complexities and dynamic nature of DCTs, ensuring that both patients and stakeholders benefit from the efficiency and adaptability of decentralized trials.
Download
Paper Nr: 186
Title:

Bridging Gaps in Fracture Rehabilitation: A Mobile Solution Proposal for Comprehensive Recovery
Authors:

Matthias Maszuhn, Frerk Müller-Von Aschwege, Felix Jansen, Andreas Hein, Hester Knol, David Snowdon, Michael Buschermöhle, Dominik Barth, Luisa Haag, Nadine Wohlers, Linda Rüde and Oliver Pieske
Abstract: This paper explores the prevalent challenges associated with musculoskeletal injuries across various demographics. It proposes the idea for a comprehensive mobile application designed to improve post-fracture aftercare by addressing existing gaps in information sharing, personalization, and remote care. Comprising three core components – recording and assessment of physiotherapy exercises, physical load measurement at the fracture, and a shared documentation tool for all participants involved in the aftercare process – the system aims to enhance patient compliance and improve recovery outcomes. The system will then be evaluated technically with healthy subjects to validate the system components. Subsequent usability evaluations will involve feedback from both healthy subjects and potential end-users, paving the way for planned clinical investigations with patients undergoing ankle fracture treatments to assess system efficacy, patient-reported outcomes, and compliance.
Download
Paper Nr: 192
Title:

Prediction of Heart Disease Severity Using Hierarchically-Structured Machine-Learning Models with Feature Space Reduction
Authors:

Ayami Kiuchi, Tomoya Fujita and Hayato Yamana
Abstract: Heart disease is the primary cause of death worldwide according to the 2019 statistics published by the World Health Organization (WHO), with roughly 8.9 million people dying annually. Predicting the likelihood and severity of this disease leads to earlier detection and helps reduce the workload of medical professionals. Previous studies have adopted a one-time classification that is insufficient to predict heart disease severity. This study proposes a novel classification method to enhance the prediction accuracy of heart disease by using: 1) a hierarchical binary-classification technique to classify the severity in order from the lowest level and 2) a data-preprocessing technique to transform continuous values into binary values based on medical knowledge and statistics information to decrease the feature space. An experimental evaluation of the heart-disease dataset from the UC Irvine (UCI) machine-learning repository confirms that the proposed method achieves the highest accuracy at 100% in predicting the presence of heart disease and at 93.13% in its severity level. In addition, the proposed method achieved 96.67%, 91.25%, 90.59%, and 93.64% accuracy for severity prediction in the Cleveland, Hungarian, Long-Beach-VA, and Switzerland datasets, respectively.
Download
Paper Nr: 193
Title:

Taking Behavioral Science to the next Level: Opportunities for the Use of Ontologies to Enable Artificial Intelligence-Driven Evidence Synthesis and Prediction
Authors:

Oscar Castro, Jacqueline Louise Mair, Florian von Wangenheim and Tobias Kowatsch
Abstract: Decades of research have created a vast archive of information on human behavior, with relevant new studies being published daily. Despite these advances, knowledge generated by behavioral science – the social and biological sciences concerned with the study of human behavior – is not efficiently translated for those who will apply it to benefit individuals and society. The gap between what is known and the capacity to act on that knowledge continues to widen as current evidence synthesis methods struggle to process a large, ever-growing body of evidence characterized by its complexity and lack of shared terminologies. The purpose of the present position paper is twofold: (i) to highlight the pitfalls of traditional evidence synthesis methods in supporting effective knowledge translation to applied settings, and (ii) to sketch a potential alternative evidence synthesis approach which leverages on the use of ontologies – formal systems for organizing knowledge – to enable a more effective, artificial intelligence-driven accumulation and implementation of knowledge. The paper concludes with future research directions across behavioral, computer, and information sciences to help realize such innovative approach to evidence synthesis, allowing behavioral science to advance at a faster pace.
Download
Paper Nr: 210
Title:

Architectural Design for Enhancing Remote Patient Monitoring in Heart Failure: A Case Study of the RETENTION Project
Authors:

Ourania Manta, Nikolaos Vasileiou, Olympia Giannakopoulou, Konstantinos Bromis, Ioannis Kouris, Maria Haritou, Lefteris Koumakis, George Spanoudakis, Irina E. Nicolae, C. Septimiu Nechifor, Miltiadis Kokkonidis, Michalis Vakalelis, Yorgos Goletsis, Maria Roumpi, Dimitrios I. Fotiadis, Heraklis Galanis, Panagiotis Dimitrakopoulos, George K. Matsopoulos and Dimitrios D. Koutsouris
Abstract: This paper introduces the RETENTION Platform, an integrated healthcare data management system meticulously crafted to support personalised interventions, thereby enhancing outcomes for heart failure (HF) patients. Comprising three fundamental components—the Global Insights Cloud (GIC), the Clinical Site Backend (CSB), and Patient Edge (PE)—the platform coordinates a sophisticated array of functions. The GIC facilitates data analysis and machine learning model training, while the CSB enables daily patient check-ups, data gathering, and intervention application. The Patient Edge enables continuous monitoring and feedback collection from patients. The system is deployed using virtual machines (VMs) and Docker containers on a cloud-based infrastructure. Integration and testing procedures are outlined to safeguard system functionality. This paper provides a comprehensive overview of the RETENTION Platform’s architecture and highlights its potential for improving healthcare delivery through personalised interventions.
Download
Paper Nr: 212
Title:

Apnea Hypopnea Index Estimation from Low-Granularity Overnight Oxymetry Data
Authors:

Nhung Huyen Hoang and Zilu Liang
Abstract: The challenge of detecting sleep disorders from consumer wearable sensors is attracting more and more researchers in the field. Sleep apnea has been the target of many sleep studies because this disorder has many health, physical, and mental consequences. Because obstruction in the airway is the direct cause of sleep apnea, overnight pulse oximetry provides valuable information to simplify the obstructive sleep apnea (OSA) screening. In this study, we aimed to estimate the apnea-hypopnea index (AHI) from consumer-grade low-granularity oximetry data. We used 5804 sleep records from the Sleep Heart Health Study (SHHS) dataset for training and testing six different regression models. The best model achieved an R-square of 0.64 ± 0.019 and ICC of 0.77 ± 0.015. The estimated AHI was further converted to 4 levels of severity (i.e., normal, mild, moderate, and severe). The macro F1-score, precision and recall were 0.576 ± 0.044, 65.16 ± 4.58 and 56.28 ± 3.42, respectively. Central tendency measure, sample entropy and zero crossing of the oximetry data are the most important features for AHI estimation. Differences between male and female groups indicate a promising direction to improve the models' performance.
Download
Paper Nr: 221
Title:

Perceptions of Humanoid Robots in Caregiving: A Study of Skilled Nursing Home and Long Term Care Administrators
Authors:

Rana Imtiaz and Arshia Khan
Abstract: As the aging population increases and the shortage of healthcare workers increases, the need to examine other means for caring for the aging population increases. One such means is the use of humanoid robots to care for social, emotional, and physical wellbeing of the people above 65. Understanding skilled and long term care nursing home administrators’ perspectives on humanoid robots in caregiving is crucial as their insights shape the implementation of robots and their potential impact on resident well-being and quality of life. This authors surveyed two hundred and sixty nine nursing homes executives to understand their perspectives on the use of humanoid robots in their nursing home facilities. The data was coded and results revealed that the executives were keen on exploring other avenues for care such as robotics that would enhance their nursing homes abilities to care for their residents. Qualitative analysis reveals diverse perspectives on integrating humanoid robots in nursing homes. While acknowledging benefits like improved engagement and staff support, concerns persist about costs, impacts on human interaction, and doubts about robot effectiveness. This highlights complex barriers—financial, technical, and human—and emphasizes the need for strategic implementation. It underscores the importance of thorough training, role clarity, and showcasing technology benefits to ensure efficiency and satisfaction among staff and residents.
Download
Paper Nr: 228
Title:

Evaluating Movement and Device-Specific DeepConvLSTM Performance in Wearable-Based Human Activity Recognition
Authors:

Gabriela Ciortuz, Hawzhin Hozhabr Pour and Sebastian Fudickar
Abstract: This article provides a comprehensive look at human activity recognition via three consumer devices with different body placements and a deep hybrid model containing CNN and LSTM layers. The used dataset consists of 53 activities recorded from the motion sensors (IMUs) of the three devices. Compared to the available human activity recognition datasets, this dataset holds the biggest number of classes, enabling us to provide an in-depth analysis of activity recognition for health-related assessments, as well as a comparison with other benchmark models such as a CNN and LSTM model. In addition, we categorize the activities into six movement groups and discuss their relevance for health-related assessments. Our results show that the hybrid model outperforms the benchmark models for all devices individually and all together. Furthermore, we show that the smartwatch could as a standalone consumer device classify activities in the six movement groups very well and for most of the use cases using a smartwatch would be practical.
Download
Paper Nr: 248
Title:

P2DS: A Holistic Approach to Psychiatric Disease Detection in Community Pharmacies
Authors:

André Dias, Tiago Dias, Eva Maia and Isabel Praça
Abstract: Health workers appear to have an increased risk of developing psychiatric diseases, namely Post-traumatic stress disorder (PTSD), Depression and Burnout, due to the nature of their job. In recent years, several approaches based on artificial intelligence have emerged, using facial expression, audio, text and physiological features to detect depression, stress and burnout. However, most of these solutions have limitations in their capacity to simultaneously detect multiple diseases, are not widely implemented in healthcare settings, and, in some cases, lack explainability. To address this challenge, we propose Psychiatric Disease Detection System (P2DS), a holistic rule-based system capable of detecting PTSD, Depression and Burnout in community pharmacists, combining emotion recognition, physiological and performance-related features. The set of rules developed to detect each disease is based on the most objective medical literature available, making the system explainable and suitable for healthcare environments.
Download
Paper Nr: 251
Title:

Com@Rehab: An Interactive and Personalised Rehabilitation Activity Based on Virtual Reality
Authors:

Carolina Silva, Claudia Quaresma, Raquel Silva, Sara Carvalho, Rute Costa, Miguel Fernández, Miguel Fonseca, Andreia Pinto de Sousa, Ana Londral and Micaela Fonseca
Abstract: Background: This work presents Com@Rehab, a patient-centred activity for individuals needing a physical rehabilitation approach and with specific loss of functionality, designed for the context of severe post-covid19 complications. Within this scope, this paper focuses on the description of the activity in virtual reality (VR), its components, the game design approach, and the results of an initial prototype testing in the laboratory aimed at evaluating the experience of the Com@Rehab system. Methods: The VR activity was customised according to patients’ clinical needs while replicating an activity of daily living. A prototype was tested by a group of 33 healthy individuals for a showering activity scenario. A questionnaire was developed within the scope of this project to test the efficiency of the technology that supports the VR activity, as well as to evaluate health literacy components. Results: Preliminary results showed that 94% of the participants recommended the experience, the performance of the various components of the system was successfully implemented, participants quickly adhered to the VR technology, and the user interface (UI) assistant functionality needs to be improved. Conclusion: The prototype test shows potential effectiveness in enhancing the rehabilitation experience and favourable usability, offering a promising path for advancing rehabilitative care. Further research is needed for validation in clinical settings. In addition, the Com@Rehab Communication Module improves human-human and human-machine communication while contributing to health literacy.
Download
Paper Nr: 260
Title:

Investigating the Impact of Ventilator Bundle Compliance Rates on Predicting ICU Patients with Risk for Hospital-Acquired Ventilator-Associated Pneumonia Infection in Saudi Arabia
Authors:

Ghaida S. Alsaab and Sarah A. Alkhodair
Abstract: Pneumonia is the most common infectious disease picked up in the Intensive Care Unit (ICU) and accounts for nearly 27% of all hospital infections—from 5% to 40% of ICU patients on mechanical ventilation risk getting infected by ventilator-associated pneumonia. Fortunately, by identifying patients more likely to contract pneumonia, up to 50% of ventilator-associated pneumonia infections can be avoided. To our knowledge, this is the first study that tackles the problem of identifying ICU patients with a high risk of developing ventilator-associated pneumonia in Saudi hospitals, considering the impact of ventilator bundle compliance rates on the predicted results. Five machine learning models were built using two real life datasets from the Health Electronic Surveillance Network (HESN) at the Saudi Ministry of Health. Results show that including ventilator bundle compliance rates data in the prediction process yields improved results in general; however, the extent of enhancement varies across models.
Download
Paper Nr: 261
Title:

Customer Identity Management in Health Insurance with Blockchain Technology: A Literature Review
Authors:

Matthias Pohl, Hannah Giegold, Christian Haertel, Daniel Staegemann and Klaus Turowski
Abstract: Customer identity management systems are an important part of the IT infrastructure of health insurance companies. However, the current systems face challenges due to the centralized system architecture, display disadvantages in identity verification, and pose security risks for customer data. Since blockchain systems are often mentioned as a solution, the goal of this paper is to examine how blockchain-based identity management can improve this particular process of identity management in the health insurance industry. Therefore, a systematic literature review was conducted, covering the challenges of centralized systems, a solution to the problem through decentralized systems, and possible designs and approaches of blockchain identity management systems. This revealed that current systems face problems in identity verification, authentication, user experience, data storage, data security, and data control. In addition to that, it was found that decentralized systems can solve many of those challenges. They facilitate the know-your-customer process for customers and companies, increase data security, create a trusting relationship between the customer and the company, and give customers control over their data. Thus, the use of a decentralized identity management system for the insurance industry is associated with advantages and has great potential to improve the current identity management process.
Download
Paper Nr: 264
Title:

Augmentation of Motor Imagery Data for Brain-Controlled Robot-Assisted Rehabilitation
Authors:

Roman Mouček, Jakub Kodera, Pavel Mautner and Jaroslav Průcha
Abstract: Brain-controlled robot-assisted rehabilitation is a promising approach in healthcare that can potentially and in parallel improve and partly automate the rehabilitation of motor apparatus and related brain structures responsible for movement. However, building a real-world rehabilitation system has many challenges and limitations. One of these challenges is the small size of the data that can be collected from the target group of people recovering from injured motor functions to train deep learning models recognizing motor imagery patterns. Therefore, the primary experiments with data augmentation and classification results over the collected and augmented dataset are presented.
Download
Paper Nr: 265
Title:

Evaluation of the Performance of Wearables’ Inertial Sensors for the Diagnosis of Resting Tremor in Parkinson’s Disease
Authors:

Carlos Polvorinos-Fernández, Luis Sigcha, Laura Pereira de Pablo, Luigi Borzí, Paulo Cardoso, Nelson Costa, Susana Costa, Juan Manuel López, Guillermo de Arcas and Ignacio Pavón
Abstract: Currently, objective monitoring of resting tremor in Parkinson’s disease (PD) involves wearable devices and machine learning. Smartwatches may present an affordable method for remote and unintrusive tremor monitoring. However, the development of optimized systems is necessary to perform accurate monitoring in free-living settings. In this study, the potential of inertial sensors to detect resting tremors is evaluated. A smartwatch was placed on the wrist of six subjects with PD during the execution of MDS-UPDRS motor tasks. Data were collected over eight weeks from triaxial accelerometer and gyroscope simultaneously and used to implement machine learning algorithms to detect resting tremor. The best performance (accuracy 97.0% in tremor detection) was achieved using accelerometer data analysed with a Random Forest classifier, while the gyroscope showed lower performance (93.0%). The results indicates that the use of the accelerometer in commercial smartwatches can offer effective results for detecting resting tremors, while reducing computational workload. These results show opportunities for the development of robust free-living tremor monitoring systems using commodity devices and algorithms using a single sensor.
Download
Paper Nr: 35
Title:

Hyperparameter Optimization Using Genetic Algorithm for Extracting Social Determinants of Health Text
Authors:

Navya Martin Kollapally and James Geller
Abstract: Clinical factors account only for a small portion, about 10-30%, of the controllable factors that affect an individual’s health outcomes. The remaining factors include where a person was born and raised, where he/she pursued their education, what their work and family environment is like, etc. These factors are collectively referred to as Social Determinants of Health (SDoH). Our research focuses on extracting sentences from clinical notes, using an SDoH ontology (called SOHO) to provide appropriate concepts. We utilize recent advancements in Deep Learning to optimize the hyperparameters of a Clinical BioBERT model for SDoH text. A genetic algorithm-based hyperparameter tuning regimen improved with principles of simulated annealing was implemented to identify optimal hyperparameter settings. To implement a complete classifier, we pipelined Clinical BioBERT with two subsequent linear layers and two dropout layers. The output predicts whether a text fragment describes an SDoH issue of the patient. The proposed model is compared with an existing optimization framework for both accuracy of identifying optimal parameters and execution time.
Download
Paper Nr: 41
Title:

Longitudinal Data Analysis Based on Triadic Rules to Describe of the Psychological Reactions During COVID 19 Pandemic
Authors:

Lincoln A. N. Coutinho, Mark A. J. Song and Luis E. Zárate
Abstract: Longitudinal studies are essential to understand the evolution of individuals’ psychological behaviors, especially in pandemic scenarios. The work proposes the application of the triadic analysis, derived from the theory of Formal Analysis of Concepts, to describe, through rules, a longitudinal database about the attitudes and reactions of individuals during COVID 19. As a main result, one can observe how the different factors considered in the study are related in different scenarios of the pandemic, showing degrees of stress related to the prevention of the disease.
Download
Paper Nr: 48
Title:

A Methodology Based on Subgroup Discovery to Generate Reduced Subgroup Sets for Patient Phenotyping
Authors:

Antonio Lopez-Martinez-Carrasco, Jose M. Juarez, Manuel Campos and Bernardo Canovas-Segura
Abstract: Subgroup Discovery (SD) is a supervised machine learning technique that mines a set of easily readable features of patients with a medical condition in the form of a subgroup set (called patient phenotype). However, using only the output obtained by a single execution of an SD algorithm hinders the discovery of the best phenotypes since it is difficult for clinicians to choose the most suitable algorithm, its best hyperparameters and the quality measure. Therefore, we propose a new phenotyping approach based on SD that evaluates the outcomes of different SD algorithms to obtain a final patient phenotype with a reduced dependency on the initial conditions of these executions and to ensure diversity in terms of coverage of the subgroups from this phenotype. For that, we first define the problem of mining a patient phenotype in the form of a reduced subgroup set and, after that, we propose a new 6-step methodology to tackle this problem. Moreover, we carry out experiments driven by this methodology and focused on the antibiotic resistance problem by using the MIMIC-III public database and the patients infected by an Enteroccous Sp. bacterium resistant to Vancomycin as a target. Finally, we obtain a phenotype formed of 7 subgroups.
Download
Paper Nr: 51
Title:

Comparision Through Architectures of Semantic Segmentation in Breast Ultrasound Images Across Differents Input Data Dimensions
Authors:

Clécio Elias Silva E. Silva, Salomão Machado Mafalda, Emili Silva Bezerra, Gustavo Moreira Oliveira de Castro, Paulo Chavez dos Santos Júnior and Ana Beatriz Alvarez
Abstract: Breast cancer is a problem that affects thousands of people every year, early diagnosis is important for the treatment of this disease. Deep learning methods shows impressive results in identification and segmentation of breast cancer task. This paper evaluates the impact of input size images on three semantic segmentation architectures applied to breast tumour ultrasound, in U-net, SegNet and DeepLabV3+. In order to (comprehensively) evaluate each architecture, 5-fold cross validation was carried out, thus reducing the impact of variations in validation and training sets. In addition, the performance of the analyzed architectures was measured using the IoU and Dice metrics. The results showed that the DeepLabV3+ architecture performed better than the others architectures in segmenting breast tumours, achieving an IoU of 0.70 and Dice of 0.73, with the input dimension of the images being 128×128.
Download
Paper Nr: 52
Title:

An Android App for Training New Doctors in Mechanical Ventilation
Authors:

Andrea Bombarda, Sara Noto Millefiori, Michela Penzo, Luca Novelli and Angelo Gargantini
Abstract: Mechanical ventilation is essential for critically ill patients, as recently demonstrated by the COVID-19 pandemic. The experience of physicians in correctly selecting ventilation parameters and values plays a crucial role in ensuring the best possible outcome. In order to aid physicians in setting up a mechanical ventilator, several brands have implemented in their products an adaptive ventilation mode called Adaptive Support Ventilation (ASV). This mode automatically selects pressure and respiratory rate to require the patient the minimum breathing effort possible. However, physicians are generally skeptical about adopting this ventilation mode, as they prefer to have all parameters under their control. Nevertheless, we believe that comprehending how ASV works is paramount important, to understanding the patterns used, and possibly exploiting them while manually setting mechanical ventilators. For this reason, in this paper, we present Ventilation App, an Android app for training new physicians in mechanical ventilation. It allows the simulation of a ventilation process for a patient unable to breathe and gives feedback to the user by exploiting the same operating principles of the ASV mode. Thanks to the feedback received by a collaborating physician, we believe that our app can be useful for allowing physicians-in-training to acquire proficiency in mechanical ventilation.
Download
Paper Nr: 83
Title:

Improving the Instance Selection Method for Better Detection of Depression in Children and Adolescents
Authors:

Ariane C. B. da Silva, Maycoln L. M. Teodoro and Cristiane N. Nobre
Abstract: Depression is the leading global cause of disability and often begins in adolescence, a critical period for developing depressive symptoms. Major depressive disorder in the early stages of life is common worldwide but challenging to diagnose. Identifying the most striking profiles of depression in children and adolescents could benefit the training and performance of Machine Learning models and thus help in the diagnosis. Instance Selection is one of the most applied methods for data reduction, allowing the most significant samples to represent them. This work seeks to improve the SI with the Ant Colony Optimization heuristic, introducing stochasticity control to better characterize profiles of children and adolescents with depression. The proposed technique increased the detection rate of individuals with high symptoms in all evaluated algorithms between 0.07 and 8.93 percentage points.
Download
Paper Nr: 96
Title:

Identification of Patient Ventilator Asynchrony in Physiological Data Through Integrating Machine-Learning
Authors:

Anthony J. Stell, Ernesto C. Caparo, Zhe Wang, Chenyang Wang, David J. Berlowitz, Mark E. Howard, Richard O. Sinnott and Uwe Aickelin
Abstract: Patient Ventilator Asynchrony (PVA) occurs where a mechanical ventilator aiding a patient’s breathing falls out of synchronisation with their breathing pattern. This de-synchronisation may cause patient distress and can lead to long-term negative clinical outcomes. Research into the causes and possible mitigations of PVA is currently conducted by clinical domain experts using manual methods, such as parsing entire sleep hypnograms visually, and identifying and tagging instances of PVA that they find. This process is very labour-intensive and can be error prone. This project aims to make this analysis more efficient, by using machine-learning approaches to automatically parse, classify, and suggest instances of PVA for ultimate confirmation by domain experts. The solution has been developed based on a retrospective dataset of intervention and control patients that were recruited to a non-invasive ventilation study. This achieves a specificity metric of over 90%. This paper describes the process of integrating the output of the machine learning into the bedside clinical monitoring system for production use in anticipation of a future clinical trial.
Download
Paper Nr: 109
Title:

Machine Learning-Based Qualitative Analysis of Human Gait Through Video Features
Authors:

Nicoletta Balletti, Roberto Zinni, Marco Russodivito, Gennaro Laudato, Simone Scalabrino and Rocco Oliveto
Abstract: Strokes constitute a major cause of both mortality and disability, carrying significant economic implications for healthcare systems. Evaluating the quality of gait in post-stroke patients during rehabilitation is essential for providing effective care. The Dynamic Gait Index (DGI) is a valuable metric for evaluating gait quality. However, the assessment of such an index typically requires invasive tests or specialized sensors. In this paper, we introduce a machine learning-based approach for estimating DGI exclusively from video recordings. Our research encompasses a comprehensive set of experiments, including data preprocessing, feature selection, and the application of various machine learning algorithms. To ensure the robustness of our findings, we employ the Leave 1 Subject Out (L1SO) cross-validation method. Our results underscore the challenge of accurately estimating DGI using solely video data. We achieved an R-squared (R2 ) value of only 0.19 and a mean absolute error (MAE) of 2.2. Notably, we observed that our approach yielded notably poorer results for a specific subset of three patients. Upon excluding this subset, the R2 increased to 0.30, and the MAE improved to 1.9. This observation suggests that incorporating patient-specific features into the model may hold the key to enhancing its overall accuracy.
Download
Paper Nr: 116
Title:

A Systematic Analysis of Depression-Related Discourse Within Facebook: A Comparison Between Brazilian and American Communities
Authors:

Silas Lima Filho, Mônica Ferreira da Silva and Jonice Oliveira
Abstract: Identifying the symptoms of a depressive disorder can help potential sufferers seek professional help, increasing their chances of recovery. This article presents the operationalization of systems and tools to systematize the analysis process using data from depression-related communities within Facebook. We discuss how we can utilize the data to understand details about depression and the discourse surrounding the disorder through textual analysis using LIWC. The results show a low correlation between textual analysis and the features of social media interaction. This study, through a systematic use of data collection and analysis tools, aims to provide explanatory insights into messages discussing the topic of depression.
Download
Paper Nr: 188
Title:

Evaluation of the Role of the Informatician after Changes in the Legislative Landscape of Women’s Reproductive Health
Authors:

Rebecca A. Meehan and Julaine Clunis
Abstract: In the United States, on June 24, 2022, the Supreme Court removed constitutional rights for abortion in the Dobbs v. Jackson Women’s Health Organization decision, which had been precedent for almost 50 years. Given these legal changes, how do health informaticists continue to use data and information to improve health and healthcare, when that data may, quite plausibly, diminish the quality of care for girls and women. This position paper discusses three strategies for health informaticians to improve health and healthcare in light of these recent legislation changes: 1) education and training of patients and stakeholders on limitations of HIPAA and on the importance of maintaining privacy and personal health information; 2) strengthening the protection of personal health information for women’s reproductive care by re-categorizing it as ‘sensitive’ information, similar to behavioral and mental health data; and 3) clarify medical conditions by evaluating medical vocabularies and coding structures that accurately reflect the clinical realities of reproductive care.
Download
Paper Nr: 195
Title:

Image and Text Feature Based Multimodal Learning for Multi-Label Classification of Radiology Images in Biomedical Literature
Authors:

Md. Rakibul Hasan, Md Rafsan Jani and Md Mahmudur Rahman
Abstract: Biomedical images are crucial for diagnosing and planning treatments, as well as advancing scientific understanding of various ailments. To effectively highlight regions of interest (RoIs) and convey medical concepts, annotation markers like arrows, letters, or symbols are employed. However, annotating these images with appropriate medical labels poses a significant challenge. In this study, we propose a framework that leverages multimodal input features, including text/label features and visual features, to facilitate accurate annotation of biomedical images with multiple labels. Our approach integrates state-of-the-art models such as ResNet50 and Vision Transformers (ViT) to extract informative features from the images. Additionally, we employ Generative Pre-trained Distilled-GPT2 (Transformer based Natural Language Processing architecture) to extract textual features, leveraging their natural language understanding capabilities. This combination of image and text modalities allows for a more comprehensive representation of the biomedical data, leading to improved annotation accuracy. By combining the features extracted from both image and text modalities, we trained a simplified Convolutional Neural Network (CNN) based multi-classifier to learn the image-text relations and predict multi-labels for multi-modal radiology images. We used ImageCLEFmedical 2022 and 2023 datasets to demonstrate the effectiveness of our framework. This dataset likely contains a diverse range of biomedical images, enabling the evaluation of the framework’s performance under realistic conditions. We have achieved promising results with the F1 score of 0.508. Our proposed framework exhibits potential performance in annotating biomedical images with multiple labels, contributing to improved image understanding and analysis in the medical image processing domain.
Download
Paper Nr: 204
Title:

Counting Red Blood Cells in Thin Blood Films: A Comparative Study
Authors:

Alan Klinger Sousa Alves and Bruno Motta de Carvalho
Abstract: Malaria is a disease caused by a parasite that is transmitted to humans through the bites of infected mosquitoes. There were an estimated 247 million cases of malaria in 2021, with an estimated number of 619,000 deaths. One of the tasks in diagnosing malaria and prescribing the correct course of treatment is the computation of parasitemia, that indicates the level of infection. The parasitemia can be computed by counting the number of infected Red Blood Cells (RBCs) per µL or the percentage of infected RBCs in 500 to 2,000 RBCs, depending on the used protocol. This work aims to test several techniques for segmenting and counting red blood cells on thin blood films. Popular methods such as Otsu, Watershed, Hough Transform, combinations of Otsu with Hough Transform and convolutional neural networks such as U-Net, Mask R-CNN and YOLO v8 were used. The results obtained were compared with two other published works, the Malaria App and Cell Pose. As a result, one of our implemented methods obtained a higher F1 score than previous works, especially in the scenario where there are clumps or overlapping cells. The methods with the best results were YOLO v8 and Mask R-CNN.
Download
Paper Nr: 205
Title:

Evaluation of Orthogonal Vector Projection Method in ST Algorithm for Generating Differential Diagnoses of Chest Pain: A Pilot Study
Authors:

Irosh Fernando, Luke Nepia, Hoang Mai Khanh Do and Edward Holmes
Abstract: Diagnosing chest pain can be a challenging process with potential misdiagnoses causing significant morbidity and mortality, and the associated healthcare cost and burden. As a potential solution to increase the diagnostic accuracy and rule out non-life-threatening conditions, we have evaluated the method known as orthogonal vector projection which is a part of the Select and Test (ST) algorithm for medical diagnosis, as a pilot study. Using a knowledgebase consisting of 12 diagnoses and 43 clinical features, we have evaluated 47 clinical cases by comparing the diagnosis given by a senior clinician to the diagnosis arrived by the orthogonal vector projection method.
Download
Paper Nr: 207
Title:

Decoding Autism Diagnosis: A Journey Towards Transparency with XAI in ML Models
Authors:

Shivani Pandya and Swati Jain
Abstract: Autism Spectrum Disorder (ASD) is a developmental condition that manifests within the first three years of life. Despite the strides made in developing accurate autism classification models, particularly utilizing datasets like AQ-10, the lack of interpretability in these models poses a significant challenge. In response to this concern, we employ eXplainable Artificial Intelligence (XAI) techniques, specifically Local Interpretable Model-agnostic Explanations (LIME) and Shapley Additive exPlanations (SHAP), to enhance transparency. Our primary aim, following the commendable accuracy achieved with the AQ-10 dataset, is to demystify the black-box nature of machine learning models used for autism classification. LIME provides locally faithful explanations, offering a more nuanced understanding of predictions, while SHAP quantifies the contribution of each feature to the model’s output. Through instance-based analyses, we leverage these XAI techniques to delve into the decision-making processes of the model at an individual level. Integrating LIME and SHAP not only elevates the model’s trustworthiness but also helps a deeper comprehension of the factors influencing autism classification. Our results underscore the efficacy of these techniques in unraveling the intricacies of the model’s decisions, shedding light on relevant features and their impact on classification outcomes. This research contributes to bridging the gap between accuracy and interpretability in machine learning applications, particularly within the realm of autism classification.
Download
Paper Nr: 213
Title:

Characterization of Telemedicine Patients to Discover Patient Journeys Using Process Mining
Authors:

Matías Cornejo, Sebastián Valderrama and Eric Rojas
Abstract: Process mining has established itself as a highly valuable tool in healthcare and demonstrated its effectiveness in process discovery, compliance verification and workflow optimization across a variety of clinical settings. However, its application in the analysis of telemedicine medical care has not been explored in depth. The present paper introduces the first stage of the research “Improving the patient journey in telemedicine using process mining” which aims to optimize the care process in telemedicine. In this initial stage, the characterization of patients who utilize this model of care in a hospital network in Chile between 2020 and 2023 is conducted. Accordingly, statistical information from the Red de Salud UC-Christus healthcare network is used to determine the most frequent characteristics of patients in socio-demographic, healthinsurance and clinical terms. Profiles of typical patients who have received treatment via telemedicine will then be constructed. The preliminary results presented herein will serve as a basis for selecting the type(s) of patients who are of particular interest to institutional authorities. In the latter stages of the project, information from the electronic clinical records of the selected patient profiles will be used to build event logs and thereby construct patient journeys through process mining.
Download
Paper Nr: 227
Title:

Investigations on Anger Experience with Other Basic Emotions Using Affective Ising Model
Authors:

Gina Rose N. Tongco-Rosario, Christie P. Sio and Jaymar Soriano
Abstract: Understanding individual differences in anger experiences is pivotal for tailored interventions. This study explores the variability in individual anger experiences, focusing on fear, happiness, and sadness as intertwined emotions. A computational approach leveraging the Affective Ising Model (AIM) was performed to analyze discrete emotion pairs to unravel the complex dynamics of how individuals experience anger. By applying the AIM to individual-level data collected through Experience Sampling Methodology (ESM), the study aims to derive parameter estimates that capture the nuanced emotional landscapes of participants. The investigation seeks to elucidate not only how individuals experience anger but also how it interacts with co-occurring emotions, shedding light on the uniqueness of emotional responses. This nuanced understanding can pave the way for personalized interventions. The parameter estimates derived from the AIM will serve as a basis for tailoring interventions, offering targeted strategies aligned with an individual’s emotional dynamics. Ultimately, this approach holds promise for shaping more effective and personalized interventions to support emotional well-being.
Download
Paper Nr: 229
Title:

Towards Inclusive Digital Health: An Architecture to Extract Health Information from Patients with Low-Resource Language
Authors:

Prajat Paul, Mohamed Mehfoud Bouh and Ashir Ahmed
Abstract: Collection of health information from the underserved community has been a challenge. Their health records are not digitized. The major population of the underserved community is text-illiterate but is not voice-illiterate. This article proposes a speech-based healthcare information collection system as an additional module to the traditional EHR system. Bangla is a language spoken widely across Bangladesh and Western parts of India by 210 million people, but it is still one of the LRLs when it comes to ASR resources. The existing research outcomes indicate the necessity of application-specific language resources for better performance. In addition, a system architecture for collecting speech data from doctor-patient conversations and an automated information retrieval system in the local language are put forward. The system also extends to extracting information that can provide assistance in operations like prescription prediction and creating new health records in digital medical history management systems.
Download
Paper Nr: 231
Title:

Creative Coding for Dance Movement Therapy in Children with Autism
Authors:

Nicolás Araya, Javier Gomez and Germán Montoro
Abstract: People diagnosed with an Autism Spectrum Disorder (ASD) have deficits in social interaction, communication and cognitive development. Children with ASD may also present motor difficulties growing up, which motivates interventions of Dance Movement Therapy (DMT) that helps them to develop social skills and integrate in society. Current technological advances have integrated into DMT interventions, enriched with virtual scenarios, projections, sensors and robot partners. These works have positive outcomes in social skills development and motor skills refinement, even though, due to confinement for COVID-19, online DMT has yet to be further explored. We propose a research methodology for the development of a tool that aims to develop self expression for ASD youth, with the creation of an artistic image based on dance and body movements. Our initial study case is Movarte, a web based tool that creates graphic pieces based on body movement and proxemic areas. 15 users evaluated the application, showing positive outcomes in terms of engagement and novelty, though it was not considered so clear and limited in terms of parameter control. Future research will provide more insight to adapt an interface for DMT in self expression for people with ASD.
Download
Paper Nr: 256
Title:

Are End-Users Participating in the Life Cycle of Healthcare Application Development? An Analysis of the Opportunities and Challenges of the Use of HCI Techniques in the Healthcare Sector
Authors:

José Silva, André Araújo, Fabio Coutinho and Alenilton Silva
Abstract: Health information systems (HIS) play a fundamental role in society, providing a solid technological basis for collecting, storing, processing, and making decisions in the healthcare sector. Due to the significant increase in software solutions and the limitations arising from the difficulties associated with mastering healthcare, it is appropriate to consider including end-users in the application development life cycle. This paper presents a systematic review of the literature to identify the presence of the end-user throughout the software life cycle and assess whether the techniques used in requirements elicitation and application evaluation are aligned with usability standards or best practices. The study resulted in the analysis of twenty-seven studies, indicating that many works do not incorporate end-users at all stages of development. Although many studies involve users in software evaluation, the methods used need more support in usability standards or guidelines. In addition, little evidence of the involvement of healthcare professionals in the development life cycle of healthcare applications was identified in the studies, which indicates that the knowledge of the domain specialist needs to be taken into account. Thus, it will be possible to investigate whether understanding the domain expert specified in openEHR archetypes can improve the life cycle of healthcare applications.
Download