Ebook: pHealth 2020

pHealth 2020 is the 17th Conference in a series of scientific events bringing together expertise from medical, technological, political, administrative, and social domains, and even from philosophy or linguistics. It opens a new chapter in the success story of the series of international conferences on wearable or implantable micro and nano technologies for personalized medicine.

Starting in 2003 as a Dissemination Activity in the framework of a European Project on Wearable Micro and Nano Technologies for Personalized Health with personal health management systems, pHealth conferences have evolved to become truly interdisciplinary and global events. Meanwhile, pHealth is comprehensively represented in the conference series, which also covers technological and biomedical facilities, legal, ethical, social, and organizational requirements and impacts, as well as the basic research necessary for enabling the future-proof care paradigms. It thereby combines medical services with public health, prevention, social and elderly care, and wellness and personal fitness to establish participatory, predictive, personalized, preventive, and effective care settings. As a result, it attracts scientists, developers, and practitioners from various technologies, medical and health disciplines, legal affairs, politics, and administration from all over the world. The conference brings together not only health service vendor and provider institutions, payer organizations, governmental departments, academic institutions, and professional bodies, but also patients and citizen representatives.

Smart mobile systems such as microsystems, smart textiles, smart implants, sensor-controlled medical devices, and innovative sensor and actuator principles and techniques, as well as related body, local and wide area networks up to cloud services have become important enablers for telemedicine and ubiquitous pervasive health as the next-generation health services. Social media and gamification have added even more knowledge to pHealth as an eco-system.

OECD has defined four basic areas to be managed in the new care model: meeting the challenges of big data; fostering meaningful innovations; understanding and addressing potential new risks; and supporting the concerted effort to un-silo communities for a virtual care future. The multilateral benefits of pHealth technologies for all stakeholder communities, including patients, citizens, health professionals, politicians, healthcare establishments, and companies from the biomedical technology, pharmaceutical, and telecommunications domains offers enormous potential, not only for the improvement of medical quality and industrial competitiveness, but also for managing health care cost.

The pHealth 2020 conference has benefited from the experience and lessons learned by the organizing committees of previous pHealth events, particularly 2009 in Oslo, 2010 in Berlin, 2011 in Lyon, 2012 in Porto, 2013 in Tallinn, 2014 in Vienna, 2015 in Västerås, 2016 in Heraklion, 2017 in Eindhoven, 2018 in Gjøvik, and 2019 in Genoa. The 2009 conference brought up the interesting idea of having special sessions, focusing on a particular topic, and being organized by a mentor/moderator. The Berlin event in 2010 initiated workshops on particular topics prior to the official start of the conference. Lyon in 2011 initiated the launch of so-called dynamic demonstrations allowing participants to dynamically show software and hardware solutions on the fly without the need for a booth. Implementing pre-conference events, the pHealth 2012 in Porto gave attendees a platform to present and discuss recent developments and provocative ideas which helped to animate the sessions. Highlight of pHealth 2013 in Tallinn was the special session on European projects’ success stories, but also presentations on the newest paradigm changes and upcoming challenges of Big Data, Analytics, Translational and Nano Medicine, etc. Vienna, in 2014, focused on lessons learned from international and national R&D activities and practical solutions, and especially from Horizon 2020, the new EU Framework Program for Research and Innovation. Beside reports about technology-transfer support and building ecosystems and value chains to ensure better time-to-market and higher impact of knowledge-based technologies, the acceptability of solutions with particular consideration of security and privacy aspects were presented and discussed in depth. pHealth 2015 in Västerås addressed mobile technologies, knowledge-driven applications and computer-assisted decision support, but also apps designed to support elderly and chronic patients in their daily and possibly independent living. Furthermore, fundamental scientific and methodological challenges of adaptive, autonomous, and intelligent pHealth approaches, the new role of patients as consumers and active parties with growing autonomy and related responsibilities, as well as requirements and solutions for mHealth in low- and medium-income countries were considered. The pHealth 2016 conference aimed at the integration of biology and medical data, the deployment mobile technologies through the development of micro-nano-bio smart systems, the emphasis on personalized health, virtual care, precision medicine, big bio-data management and analytics. The pHealth 2017 event in Eindhoven provided an inventory of the former conferences by summarizing requirements and solutions for pHealth systems, highlighting the importance of trust, and renewed the focus on behavioral aspects in the design and use of pHealth systems. A specific aspect addressed was the need for flexible, adaptive and knowledge-based systems as well as decision intelligence. pHealth 2018 established national and European satellite workshops, so completing the more theoretical consideration of the majority of the papers by organizational and practical experiences. Borrowing from good experiences at former events, pHealth 2018 responded to the national and regional needs for advancing the healthcare systems and their services to citizens and health professionals. pHealth 2019 put a special focus on artificial intelligence (AI) and machine learning (ML) and their deployment for decision support. In that context, ethical challenges and related international manifests were discussed in depth. In view of the advancement of pHealth to P5 medicine and smart systems approaches, the 2020 event additionally covers AI and robots in healthcare, bio-data management and analytics for personalized health, security, privacy and safety challenges, integrated care, and also the intelligent management of specific diseases including the Covid-19 pandemic. In that context, communication and cooperation with national and regional health authorities and the challenges facing health systems in developing countries were also addressed.

pHealth 2020 has been strongly supported by the Czech Institute of Informatics, Robotics and Cybernetics of the Czech Technical University in Prague, the Department of Natural Sciences of the Faculty of Biomedical Engineering of the Czech Technical University in Prague, and the Czech Society for Biomedical Engineering and Medical Informatics.

Following a long-term tradition, the Working Groups “Electronic Health Records (EHR)”, “Personal Portable Devices (PPD)” and “Security, Safety and Ethics (SSE)” of the European Federation for Medical Informatics (EFMI) have been actively involved in the preparation and realization of the pHealth 2020 Conference.

This proceedings volume includes 1 keynote, 5 invited talks, 25 oral presentations, and 8 short poster presentations from 99 authors, representing 16 countries from all around the world. All submissions have been carefully and critically reviewed by at least two independent experts from a country other than the authors’ home, and additionally by at least one member of the Scientific Program Committee. The performed highly selective review process resulted in a full papers rejection rate of 36%, despite of the specific dedication of the addressed community as compared to multi-topic conferences. This process guarantees a high scientific level of papers accepted and ultimately published. The editors are indebted to all authors, as well as to the internationally acknowledged and highly experienced reviewers, for their essential contribution to the quality of the conference and the book at hand.

Neither the pHealth 2020 Conference nor the publication of the pHealth 2020 Proceedings at IOS Press would not have been possible without the aforementioned pecuniary and spiritual support and sponsorship. This includes the Czech Institute of Informatics, Robotics and Cybernetics of the Czech Technical University in Prague, the Faculty of Biomedical Engineering of the Czech Technical University in Prague, and the Czech Society for Biomedical Engineering and Medical Informatics. Other supporters were the European Federation for Medical Informatics (EFMI) and standards developing organizations such as HL7 International, ISO/TC215 or CEN/TC251.

The editors are also grateful to the Members of the International Scientific Program Committee, but particularly for the dedicated efforts members of the Local Organizing Committee and their supporters for carefully and smooth preparation and operation of the conference.

Bernd Blobel, Lenka Lhotska, Peter Pharow, Filipe Sousa

(Editors)

Multidisciplinary and highly dynamic pHealth ecosystems according to the 5P Medicine paradigm require careful consideration of systems integration and interoperability within the domains knowledge space. The paper addresses the different aspects or levels of knowledge representation (KR) and management (KM) from cognitive theories (theories of knowledge) and modeling processes through notation up to processing, tooling and implementation. Thereby, it discusses language and grammar challenges and constraints, but also development process aspects and solutions, so demonstrating the limitation of data level considerations. Finally, it presents the ISO 23903 Interoperability and Integration Reference Architecture to solve the addressed problems and to correctly deploy existing standards and work products at any representational level including data models as well as data model integration and interoperability.

The paper describes the concept of the Industry 4.0 and its reflection in health care. Industry 4.0 connects intelligent production concepts with external factors, including those linked with the production and those linked more with human, as for example intelligent homes or social web systems. Communication, data and information play an important role in the whole system. After explaining basic characteristics of the Industry 4.0 concept and its main parts, we show how they can be utilized in the health care sector and what their advantages are. Key technologies and techniques include Internet of Things, big data, artificial intelligence, data integration, robotization, virtual reality, and 3D printing. Finally, we identify the main challenges and research directions. Among the most important ones are interoperability, standardization, reliability, security and privacy, ethical and legal issues.

Medical data can be represented in various forms. The most common is visualization, but recent work started to also add sonic representation – sonification. In this study we start with a theoretical background, then focus on medical applications. The discussion synthesizes the authors view about the present state of the domain and tries to foresee future potential developments in medicine. In conclusion we present a set of original recommendations for developing new applications with potential use in medicine and healthcare.

The International Patient Summary Standard (EN 17269) normalizes the dataset within the European Guideline on cross-border exchange of a patient summary. This dataset has been widely appreciated and been taken as the basis for projects in both Europe and wider afield, e.g. U.S.A, Canada and more. The dataset is a relatively mature dataset and it is currently in its third iteration (i.e., 2013, 2016, 2020). Even so, to move from a policy-driven guideline to a formal standard was not straight forward. The paper describes how the ‘minimal and non-exhaustive’ dataset could be the basis for a reference standard; one that was intended to facilitate both an ‘implementable’ and ‘sustainable’ solution. In particular, the requirement of ‘extensibility’ for the standard dataset had to be addressed.

Today’s digital information systems and applications collect every day a huge amount of personal health information (PHI) from sensor and surveillance systems, and every time we use personal computers or mobile phones. Collected data is processed in clouds, platforms and ecosystems by digital algorithms and machine learning. Pervasive technology, insufficient and ineffective privacy legislation, strong ICT industry and low political will to protect data subject’s privacy have together made it almost impossible for a user to know what PHI is collected, how it is used and to whom it is disclosed. Service providers’ and organizations’ privacy policy documents are cumbersome and they do not guarantee that PHI is not misused. Instead, service users are expected to blindly trust in privacy promises made. In spite of that, majority of individuals are concerned of their privacy, and governments’ assurance that they meet the responsibility to protect citizens in real life privacy is actually dead. Because PHI is probably the most sensitive data we have, and the authors claim it cannot be a commodity or public good, they have studied novel privacy approaches to find a way out from the current unsatisfactory situation. Based on findings got, the authors have developed a promising solution for privacy-enabled use of PHI. It is a combination of the concept of information fiduciary duty, Privacy as Trust approach, and privacy by smart contract. This approach shifts the onus of privacy protection onto data collectors and service providers. A specific information fiduciary duty law is needed to harmonize privacy requirements and force the acceptance of proposed solutions. Furthermore, the authors have studied strengths and weaknesses of existing or emerging solutions.

The value of data models in general and information models in specific has been evaluated by many scientific papers. UML as one modelling notation has documented its value as a foundation for precise specifications. Analyzing implementation guides for data exchange, they rarely include or are based on information models but simple data sets, if at all, as simple technical representation thereof. This paper wants to argue in favor of information models as a basis for creating interoperability specifications using a quite simple example and to include – or at least reference – them when providing implementation guides. The reader is invited to transfer this example to even more complex scenarios.

In healthcare settings, questionnaires are used to collect information from a patient. A standard method for this are paper-based questionnaires, but they are often complex to understand or long and frustrating to fill. To increase motivation, we developed a chatbot-based system Ana that asks questions that are normally asked using paper forms or in face-to-face encounters. Ana has been developed for the specific use case of collecting the music biography in the context of music therapy. In this paper, we compare user motivation, relevance of answers and time needed to answer the questions depending on the data entry method (i.e. app Ana versus paper-based questionnaire). A randomised trial was performed with 26 students of music therapy. The results show that the chatbot is more motivating and answers are given faster than on paper. No differences in answer relevance could be determined between the two means. We conclude that a chatbot could become an additional data entry method for collecting personal health information.

A lower-extremity exoskeleton can facilitate the lower limbs’ rehabilitation by providing additional structural support and strength. This article discusses the design and implementation of a functional prototype of lower extremity brace actuation and its wireless communication control system. The design provides supportive torque and increases the range of motion after complications reducing muscular strength. The control system prototype facilitates elevating a leg, gradually followed by standing and slow walking. The main control modalities are based on an Artificial Neural Network (ANN). The prototype’s functionality was tested by time-angle graphs. The final prototype demonstrates the potential application of the ANN in the control system of exoskeletons for joint impairment therapy.

Technological advancements in smart assistive technology enable navigating and manipulating various types of computer-aided devices in the operating room through a contactless gesture interface. Understanding surgeon actions is crucial to natural human-robot interaction in operating room since it means a sort of prediction a human behavior so that the robot can foresee the surgeon’s intention, early choose appropriate action and reduce waiting time. In this paper, we present a new deep network based on Convolution Long Short-Term Memory (ConvLSTM) for gesture prediction configured to provide natural interaction between the surgeon and assistive robot and improve operating-room efficiency. The experimental results prove the capability of reliably recognizing unfinished gestures on videos. We quantitatively demonstrate the latter ability and the fact that GestureConvLSTM improves the baseline system performance on LSA64 dataset.

Prediction of a labor due date is important especially for the pregnancies with high risk of complications where a special treatment is needed. This is especially valid in the countries with multilevel health care institutions like Russia. In Russia medical organizations are distributed into national, regional and municipal levels. Organizations of each level can provide treatment of different types and quality. For example, pregnancies with low risk of complications are routed to the municipal hospitals, moderate risk pregnancies are routed to the reginal and high risk of complications are routed to the hospitals of the national level. In the situation of resource deficiency especially on the national level it is necessary to plan admission date and a treatment team in advance to provide the best possible care. When pregnancy data is not standardized and semantically interoperable, data driven models. We have retrospectively analyzed electronic health records from the perinatal Center of the Almazov perinatal medical center in Saint-Petersburg, Russia. The dataset was exported from the medical information system. It consisted of structured and semi structured data with the total of 73115 lines for 12989 female patients. The proposed due date prediction data-driven model allows a high accuracy prediction to allow proper resource planning. The models are based on the real-world evidence and can be applied with limited amount of predictors.

Timely identification of risk factors in the early stages of pregnancy, risk management and mitigation, prevention, adherence management can reduce the number of adverse perinatal outcomes and complications for both mother and a child. We have retrospectively analyzed electronic health records from the perinatal Center of the Almazov specialized medical center in Saint-Petersburg, Russia. Correlation analysis was performed using Pearson correlation coefficient to select the most relevant predictors. We used APGAR score as a metrics for the childbirth outcomes. Score of 5 and less was considered as a negative outcome. To analyze the influence of the unstructured anamnesis data on the prediction accuracy we have run two prediction experiments for every classification task: 1. Without unstructured data and 2. With unstructured data. This study presents implementation of predictive models for adverse childbirth events that provides higher precision than state of the art models. This is due to the use of unstructured medical data in addition to the structured dataset that allowed to reach 0.92 precision. Identification of main risk factors using the results of the features importance analysis can support clinicians in early identification of possible complications and planning and execution preventive measures.

Introduction:

The electronically submitted data from midwives and hospitals to the Netherlands perinatal registry vary significantly in their data definitions, and electronic message versions. The purpose of this article is to describe the semantic cross-mapping tool and execution procedure to prepare the data for statistical analysis.

Methods:

requirements analysis, design, development and testing.

Results:

The tool for governance of versions of datasets, CIMs, data, and value sets is designed, developed, and tested. The test is based on the data-mart of version PRN 1.3 based data from 2019. Data are semantically cross mapped to current version perinatology data 2.2.

Conclusion:

The cross-mapping of PRN 1.3 data to perinatology 2.2 data are defined in the tool, testing revealed this mapping is successful.

Type 2 diabetes is one of the most common chronic diseases in the world. World Diabetes Federation experts predict that the diabetes patients’ number by 2035 will increase by 205 million to reach 592 million. For health care, this diabetes type is one of the highest priority problems. This disease is associated with many concomitant diseases leading to early disability and high cardiovascular risk. A severity disease indicator is the degree of carbohydrate metabolism compensation. Decompensated and subcompensated carbohydrate metabolism patients have increased cardiovascular risks. Therefore, it is important to be able to select the right therapy to control carbohydrate metabolism. In this study, we propose a new method for selecting the optimal therapy automatically. The method includes creating personal optimal therapies. This kind of therapy has the highest probability of compensating carbohydrate metabolism for a patient within a six-month. The method includes models for predicting the results of different therapies. It is based on data from the previous medical history and current medical indicators of patients. This method provides high-quality predictions and medical recommendations. Therefore, medical professionals can use this method as part of the Support and Decision-Making Systems for working with T2DM patients.

In this paper, we describe a strategy for the development of a genetic analysis comprehensive representation. The primary intention is to ensure the available utilization of genetic analysis results in clinical practice. The system is called Personnel Genetic Card (PGC), and it is developed in cooperation of CIIRC CTU in Prague and the Mediware company. Nowadays, genetic information is more and more part of medicine and life quality services (e.g. nutritional consulting). Therefore, there is necessary to bind genetic information with the clinical phenotype, such as drug metabolism or intolerance to various substances. We proposed a structured form of the record, where we utilize the LOINC^® standard to identify genetic test parameters, and several terminology databases for representing specific genetic information (e.g. HGNC, NCBI RefSeq, NCBI dbNSP, HGVS). Further, there are also several knowledge databases (PharmGKB, SNPedia, ClinVar) that collect interpretation for genetic analysis results. In the results of this paper, we describe our idea in the structure and process perspective. The structural perspective includes the representation of the analysis record and its binding with the interpretations. The process perspective describes roles and activities within the PGC system use.

Specific predictive models for diabetes polyneuropathy based on screening methods, for example Nerve conduction studies (NCS, can reach up to AUC 65.8 – 84.7 % for the conditional diagnosis of DPN in primary care. Prediction methods that utilize data from personal health records deal with large non-specific datasets with different prediction methods. Li et al. utilized 30 independent variables, which allowed to implement a model with AUC = 0.8863 for a Multilayer perceptron (MLP). Linear regression (LR) based methods produced up to AUC = 0.8 %. This way, modern data mining and computational methods can be effectively adopted in clinical medicine to derive models that use patient-specific information to predict the development of diabetic polyneuropathy, however, there still is a space to improve the efficiency of the predictive models. The goal of this study is the implementation of machine learning methods for early risk identification of diabetes polyneuropathy based on structured electronic medical records. It was demonstrated that the machine learning methods allow to achieve up to 0.7982 precision, 0.8152 recall, 0.8064 f1-score, 0.8261 accuracy, and 0.8988 AUC using the neural network classifier.

In this paper, we follow up on research dealing with body tracking and motor rehabilitation. We describe the current situation in telerehabilitation in the home environment. Existing solutions do not allow wide adoption due to hardware requirements and complicated setup. We come with the possibility of telerehabilitation using only laptop or mobile web camera. Together with physiotherapists, we have compiled a set of complex motor exercises to show that the system can be practically used.

The paper compares two approaches to multi-step ahead glycaemia forecasting. While the direct approach uses a different model for each number of steps ahead, the iterative approach applies one one-step ahead model iteratively. Although it is well known that the iterative approach suffers from the error accumulation problem, there are no clear outcomes supporting a proper choice between those two methods. This paper provides such comparison for different ARX models and shows that the iterative approach outperformed the direct method for one-hour ahead (12-steps ahead) forecasting. Moreover, the classical linear ARX model outperformed more complex non-linear versions for training data covering one-month period.

Human Activity Recognition (HAR) is becoming a significant issue in modern times and directly impact the field of mobile health. Therefore, it is essential the designing of systems which are capable of recognizing properly the activities conducted by the individuals. In this work, we developed a system using the Internet of Things (IoT) and machine learning technologies in order to monitor and assist individuals in their daily life. We compared the data collected using a mobile application and a wearable device with built-in sensors (accelerometer and gyroscope) with the data of a publicly available dataset. By this way, we were able to validate our results and also investigate the functionality and applicability of the wearable device that we choose for the Human Activity Recognition problem. The classification results for the different types of activities presented using our dataset (99%) outperforms the results from the publicly database (97%).

The demographic change is no longer a prognosis, but a reality seen in everyday life situations and requires mechanisms to make the public and private space elderly-adequate. These required mechanisms need to consider the varying aging process for each individual as well as adapt to the dynamic daily life of individuals characterized by spatial, temporal and activity variance. Developing assistance systems that are user-adaptive within dynamic environments is a challenging task. AI-based cyber-physical assistance systems enable such adaptive, flexible and individual assistance by processing acquired data from the physical environment using cyber resources and delivering intelligent assistance as well as interfaces to further medical services. This contribution discusses a flexible, reusable, and user-specific concept for AI-based assistance systems. Relying on distributed and heterogeneous data, the user’s context is continuously modeled and reasoned over to infer actionable knowledge within a middleware between the data layer and the application layer. To demonstrate the applicability of the concept, the use case of intelligently supporting patients’ medication adherence is shown.