Healthcare information system, solutions and personal decision support in Estonia
The starting point of the case study is the investigation of different health, behavioral and medical data arising during the life span of a person registered in health systems of Estonia. Feasibility of the implementation of different decision support applications is investigated from the prediction, prevention, diagnosis, treatment and rehabilitation point of view.
1. ENVIRONMENT’S DESCRIPTION BEFORE IMPLEMENTATION
There is plenty of data in the Estonian Health Information System (EHIS), Estonian Health Insurance Fund (EHIF) and Estonian Genome Center of the University of Tartu (EGCUT) databases. Some of the data is collected in the structured way corresponding to existing standards but considerable amount of data is presented either as a narrative text or in non-compliant structures which is a challenge to the development of a nation-wide DDS systems.
We have also established that the data required for Digital Decision Support in Personalised Medicine is similar to the data required by conventional medicine in a vast majority of cases. As a major difference, it includes genome data and covers the whole lifespan of an individual rather than just random incidents. Based on the current knowledge, genome data in the field of CVD and diabetes is not widely used for DDS. The situation with cancer prevention and treatment is somewhat better. However, the research of genome data and calculations of risks for all mentioned conditions is evolving rapidly. Therefore it is our suggestion to consider keeping two clinical fields – cancer and combination of CVD and diabetes – for the further development of the DDS systems.
One of the main problems in today’s medicine is that personal information about patient is not widely used. Now we are moving to the Personalized Medicine era.
2. IMPLEMENTATION’S INSPIRATIONS
The development in information technology, the reduced cost of person genome mapping and various sensors inspires us to move to the era of personalized medicine.
3. PROBLEM’S DESCRIPTION
One of the main problems in today’s medicine is that personal information about patient is not widely used.
4. IMPLEMENTATION’S DESCRIPTION
All information about patient from different sources is collected together and special Decision support system is created.
5. REASON OF THE CHOICE OF IMPLEMENTATIONS FOR THE ANALYSIS
In e-government medicine spending is about 15% of all budget, at the same time we are ready for personalized medicine.
APPLICATION’S IMPLEMENTATION PROCESS
The project is initialized by Estonian government and also funded by cgovernment.
6. IMPLEMENTATION’S INFLUENCE ON ENVIRONMENT
“What was the application’s influence on identified problems: was this certain problem solved?
Did the application solve any other problems? Was the implementation’s solution revealed new previously unknown problems? the project is over, is the application still needed? Does the solution applied for the described implementation can be used in other fields?”
7. APPLICATION FUNCTIONAL DESCRIPTION
“What is this application – what is the core idea of this solution? Does the application achieve previously defined objectives? Did the application change or was modified? If so, in what aspect? (Does the described version is the same as it was at the beginning?)
Whether the set out objectives have been achieved?
Whether the application has changed or has been modified? The described version of the application is the previous version or the version after changes or modifacations?”
8. IMPLEMENTATION’S INFLUENCE ON ENVIRONMENT
“What was the application’s influence on identified problems: was this certain problem solved? Has the application solved any other problems?
Has the implementation’s solution revealed new previously unknown problems?
Whether, despite the completion of the project application is still needed?
Does the solution applied for the described implementation can be used in other fields?”
9. PROJECT’S DURABILITY DESCRIPTION
Is the project still running? Is the application still being used? Did any similar implementation be created? What is the functional/meritorical reliance between described implementation and other similar applications/websites which went online later than described project? Is the source code open and available for others? Are there any other applications implemented basing on shared data? Who is managing the project at the moment?
10 Description of the trends in IT environment
Medicine is starting to adopt new treatments, medications and protocols but is lacking far behind where it goes on reflecting on the model of health( care). We basically deliver healthcare the same way it was done a hundred years ago. Now due to the exponentially growing possibilities technology is bringing to the table, we, for instance, will start bringing back health( care) into the homes of people.
Mesko, Bertalan (2014-08-27). The Guide to the Future of Medicine: Technology AND The Human Touch (Kindle Locations 39-43). Dr. Bertalan Meskó (Webicina Kft.).
Some trends in future medcine connected with IT.
Trend 1. Empowered Patients
Healthcare cannot really advance without physicians letting their patients help themselves and be a full partner in making the decisions that affect them.
Trend 2. Augmented Reality and Virtual Reality
Augmented reality (AR) is a real– time view of a real– world environment that is enhanced by computer– generated sound, video, graphics, GPS data, or inputs we may not have thought about yet. Imagine wearing an AR device while you are walking, and receiving promotional offers from shops you pass. Simultaneously you see the real and online worlds superimposed. A company called Metaio, for instance, provides an AR application for technicians to service and repair the Volkswagen XLI without any prior training. Instructions are projected on top of what they are looking at in the auto shop. Getting information via a Google Glass or digital contact lenses could greatly augment the practice of medicine.
Google Glass, a wearable computer with an optical head– mounted display, was made available to testers and developers in 2013. As of July, 2014, it is not yet commercially available, although Google did sell it publicly for one day in April, 2014. Google Glass has a touchpad on the temple piece, a camera, and an optical display. It works like a smartphone by letting users take photos and videos, browse the web, take notes, and make calls. Wearers access the Internet via natural language voice commands such as “OK, Glass, do a search for diabetes”.
Trend 3. Telemedicine and Remote Care
Hugo Gernsback was a pioneer in both radio and publishing. He designed the first home radio set and published dozens of magazines in the early 1900s. He wrote an article about the future of radio telecommunication in the February, 1925 issue of Science and Invention. The device was the “teledactyl” (tele means far, dactyl means finger in Greek) and was meant to allow doctors to see and touch their patients through a viewscreen with robotic arms that was kilometers away. He predicted that the practice of medicine would look much different by the 1970s. From their offices doctors would diagnose and treat patients in their homes via machines and devices that worked through radio waves.
Trend 4. Body Sensors Inside and Out
What do we do when we need to measure different health parameters? We go to a lab and provide blood sample; or to the hospital where they measure blood pressure, ECG, and perform other diagnostic tests. After that we wait and bring the results to the doctor to discuss the next step. If we need a radiology imaging or a laboratory test, it might take a lot of time due to waiting lists worldwide. When I wanted to get my DNA sequenced I provided saliva sample and sent it to direct– to– consumer genomic companies that gave me online access to the results a few weeks later. They even provided a genetic counselor to interpret the data for me over the phone.
Trend 5. The 3D Printing Revolution
A 14– month– old baby in the US had so many heart defects that it made the upcoming operation difficult. To better prepare in detail, hospital officials at Kosair Children’s Hospital in Louisville, Kentucky contacted the J.B. Speed School of Engineering, where a polymer model of the baby’s heart was created with a 3D printer. This provided vital insight ahead of surgery. Once the cardiothoracic surgeon had a model he knew exactly what he needed to do. The model allowed him to reduce the number of exploratory incisions and the overall operating time. This is just one example of how 3D printers could assist medical professionals.
Trend 6. The End of Human Experimentation
Today, new pharmaceuticals are approved by a process that culminates in human clinical trials. The clinical trial is a rigorous process from development of the active molecule to animal trials before the human ones, costing billions of dollars and requiring many years. Patients participating in the trial are exposed to side effects, not all of which will have been predicted by animal testing. If the drug is successful in trial, it may receive approval, but the time and expense are present regardless of the trial outcome. But what if there were another, safer, faster, and less expensive route to approval? Instead of requiring years of “ex vivo” and animal studies before human testing, what if it were possible to test thousands of new molecules on billions of virtual patients in just a few minutes?
Trend 7. Medical Decisions via Artificial Intelligence
In 2011, people witnessed an interesting and at the same time weird competition on the television quiz show Jeopardy. It featured the two best players in the history of the show, Ken Jennings, who had the longest unbeaten run of 74 winning appearances, and Brad Rutter, who had earned the biggest prize of $ 3.25 million. Their opponent was a huge computer with over 750 servers and a cooling system stored at a location so as not to disturb the players. The room– sized machine was made by IBM and named after the company’s founder, Thomas J. Watson. It did not smile or show emotion, but it kept on giving good answers. At the end, Watson won the game with $ 77,147 leaving Rutter and Jennings with $ 21,600 and $ 24,000 respectively.
Mesko, Bertalan (2014-08-27). The Guide to the Future of Medicine: Technology AND The Human Touch (Kindle Locations 2111-2117). Dr. Bertalan Meskó (Webicina Kft.). Kindle Edition.
Disruptive medical model
How to create flexible solutions
Today’s dynamic world requires changes in our information systems development and maintenance methods. A traditional method, in which the manual programming is the main form of development, does not give the necessary flexibility to respond to changes in environments. Our challenge now is to increase domain specialist participation in solutions development and maintenance. One promising direction in the search for the solution is the use the model driven approach.
The important feature of our approach is the model-driven internet of things development method. The goal model, decision model, process model, data model, user interface model, and integration model define the solution, and the user can strictly begin to test the application. All solutions are usable on mobile devices.
The internet of things  refers to uniquely identifiable objects and their virtual representations in computer networks. The term “internet of things” was proposed by Kevin Ashton in 1999. The concept of the internet of things first became popular through the Radio-frequency identification (RFID). If all objects and people in daily life were equipped with identifiers, they could be managed and inventoried by computers. In addition to using RFID, the tagging of things may be achieved through technologies such as NCF, barcodes, QR codes and digital watermarking.
Model-driven engineering (MDE)  is a software development methodology that focuses on creating and exploiting abstract representations of the knowledge and activities that govern a particular application domain instead of on the computing (or algorithmic) concepts.
A decision theory  is concerned with identifying the values, uncertainties and other relevant issues in a given decision with its rationality and the resulting optimal decision.
Our goal is to create a solution (tool) to create and maintain an internet of things solution.
The aforementioned tool should possess the following qualities:
- The System is created using the Model-Driven approach,
- The System must adapt to the changes in environment, which is a self-organization feature,
- The system must work on smart phones,
- The open use of sensors and actuators, including those in the development stage.
Among many definitions of internet of things , the definition of Stephen Haller can be regarded as the best definition:
“A world where physical objects are seamlessly integrated into the information network and where the physical objects can become active participants in business processes. Services are available for interacting with these ‘smart objects’ over the internet, query and change their state, and any information associated with them, took into account security and privacy issues” .
This definition includes all the important features of this approach:
- Physical objects are integrated to the network over the internet,
- These objects are active participants in business processes,
- In the usage of these objects, the privacy and security issues are highly valued.
First, we define the internet of things (IoT) domain model, which is developed by extending excellent models that are found in the literature  .
The Concepts of the IoT Domain Model. According to the IoT definition, the central notion is Physical Entity (PE) in the real world (see Fig 1), which is seamlessly integrated into information networks. Physical Entity also has a self-referential relationship, which is a notably powerful relationship that enables the entity to reflect on the hierarchies of physical things. For example, in the city, we have houses, which have many grounds, each of which has many flats. The top hierarchy of the object constitutes the context of the lower hierarchy of the object. In the domain model, two other key entities of the IoT are introduced: Human User and Digital Artefact (service, software agent, application) that interacts with the PE.
The Physical Objects are represented in IT by Virtual Entities. There are different types of digital representations of Physical Objects:
- Database entries,
- Object (instance of the class),
- 3D models.
Virtual Entities are Digital Artefacts, which can be of two types:
- Active Digital Artefact (software applications as software agent or Services),
- Passive Digital Artefact such as database entry.
There is a significant difference between a Physical Entity (PE) and a Virtual Entity (VE). A VE belongs to a virtual world, and a PE belongs to the real world. However, in our approach, they should be synchronized. To achieve this synchronization, the ICT Device is used, which provides the interface to interact with or gain information about the Physical Entity. Thus, we use sensors, tags and actuators.
The IoT will allow one to do a lot more than considering the situational context. Sensors, Open Data and Big Data provide a completely new face to our solutions. There are many definitions of contexts. The definition  that best satisfies our requirements is:
“Context is any information that can be used to characterize the situation of an entity. An entity is a person, place, or object that is considered relevant to the interaction between a user and an application, including the user and applications themselves”.
Abowd and Day  also introduced the primary context types:
- Location — location data from the GPS sensor,
- Identity — Identify object based on the RFID tag,
- Time — read time from clock, also daytime,
- Activity — what activities are in progress?
There is also a secondary context type, which is derivative information that we can use based on the primary context. For example, using Identity, we can obtain considerable information about a person in social networks and the internet.
In our case, the context information is the place where we can begin to play. Our system can introduce a new situation and sets new context information and state information. Based on the response of the student reactions, we make changes in the state. Thereafter, we evaluate the adequacy of the student responses.
For our approach, the adaptive and self-organizing properties are exceptionally valuable. Instead of attempting to find the most optimal solution based on the available information, we attempt to use self-organization methods. For example, it is known that city public transportation planning is a highly complex and difficult problem. The situation can change notably quickly, and our perfect plan fails. Instead of careful planning, we attempt to use adaptive techniques. For example, we can obtain information about transport needs from the sensors and quickly send our reserve buses to the necessary place.
The most general self-organizing definition is provided in :
„Self-organization is a process in which pattern at the global level of a system emerges solely from numerous interactions among the lower-level components of the system. Moreover, the rules specifying interactions among the system’s components are executed using only local information, without reference to the global pattern“.
The complexity of systems is accompanied by an increase in the importance of self-organizing systems increases.
1.1 Literature review
This article discusses two domains: the internet of things and model driven engineering [9, 10] for smart solutions.
A huge amount of scientific literature is published on the internet of things, and it does not make sense to address the entire matter because there are corresponding surveys . Our chapter focuses on the information that we obtain from the sensors and sages to the decisions that we make as a result. The internet of things for smart environments is defined as : “Interconnection of sensing and actuating devices providing the ability to share information across platforms through a unified framework, developing a common operating picture for enabling innovative applications. This is achieved by seamless ubiquitous sensing, data analytics and information representation with Cloud computing as the unifying framework“.
To create the smart internet of things solutions, the main focus should be given to data storage and analytics. For real-time processing and storing data, the NoSQL databases are used .
Data analytics includes artificial intelligence algorithms for temporal machine-learning and decision-making processes .
There are extensive studies on internet of things and serious games separately, and these topics are notably relevant today. However, there are few studies that address these topics together.
Nevertheless, notably thorough job has been done in this area in the framework of the EU project ELLIOT . As read in the project’s website: „The ELLIOT (Experiential Living Lab for the internet of things) project aims to develop an internet of things (IOT) experiential platform where users/citizens are directly involved in co-creating, exploring and experimenting new ideas, concepts and technological artifacts related to IOT applications and services.“
The requirement engineering is the process to elicit, analyze, validate and document a formal, complete and agreed requirements specification. The ELLIOT approach is an oriented innovation, where innovative thinking is guided by the Synectics method  and supported with predefined scenarios. Game playing and role playing offer two important directions to acquire requirements with a particular focus on the interactions during group work and to the simulated environments.
The model-driven approach is how the current IT solutions can be realized because the current IT solutions no longer satisfy the modern-time requirements. The creation and modification of IT solutions today are costly, time-consuming and mainly manual work. We need new approaches. A promising possibility is to use a Model-Driven engineering approach. One of the central models in this approach is the business process model.
The process: A coherent set of activities that leads to a result (product or service) is conducted by a group of cooperating actors.
In particular, the result can be the outcome of the interaction between (the things in) the domain and the things outside the domain.
A process has goals, which can fit the system goals. These goals must have measures.
For practical reasons, we may not include all state variables of things in the process state. Note that the definition of a process implies that to be a process, a set of ordered activities must lead to a defined goal.
Finally, we recognize that the process might progress through different sequences of states, which depends on the initial state and the state changes that are caused by external events of the process domain.
The method to see the notion of goal is incorporated into the abstract (state-based) view of the previously described process.
Processes  are executed to achieve some predefined outcomes. These results are consistent with some organizational objectives (system goal). We term the desired result of the process as a process goal. A process goal is formed using state variables and must have exact measures. It is notably important to have on-place target values for process goals, which allows a self-adaption of the system [18, 19]. Different options for sub processes enable one to make directed decisions.
The decision-making model consists rule formulas, which in turn consists of Conditions and Actions parts. A rule formula allows you to perform a wide range of calculations. The rule formula offers the basic mathematical operations (addition, subtraction, multiplication, division), more advanced operations (such as exponentiation or binary AND), comparisons (greater than, less than, etc.), and a wide variety of formula functions for different purposes and data types (date and time calculation, string handling and manipulation, conversion functions, system functions, etc.).
The situation modeling.
Now, we describe the situation modeling. In the series of descriptions of the possible conditions are prepared. Data updating is used to fix the situation on our server. For example, a severe decrease in weather is simulated outside the temperature changes in the context data. Now, the user must make adequate decisions to manage the situation. The user’s actions are captured saved and later analyzed by system.
The context in a smart house includes the home environmental conditions (temperature, humidity and lighting) and information about the residents (their location, mobility tracking and behavior).
The users actions can be simple (switch heating on) or complex (described using business process description language BPMN). The user can create hundreds of these descriptions of business processes. In more complicated situations, pre-prepared process archetypes are prepared [9, 21].
We currently lack the simple tools to create smart solutions. This paper proposes some solutions for the problem. First, the model-driven engineering approach can help solving this problem. Second, the use of patterns is also a notably promising direction. One of the challenges is that the domain where we act is also at the development stage; thus, we must address the domain model development in the field of internet of things.
Our future plan includes expanding the application areas of the internet of things (smart cities and smart governance). In addition, we would like to investigate the area of self-organization further.
|||S. Haller, “Internet of Things: An Integral Part of the Future Internet,” Prague, May 13Th 2009.|
|||M. Brambilla, J. Cabot and M. Wimmer, Model-Driven Software Engineering in Practice, Morgan & Claypool Publishers, 2012, p. 173.|
|||M. Peterson, An Introduction to Decision Theory (Cambridge Introductions to Philosophy), 1 edition ed., Cambridge University Press, 2009, p. 325.|
|||D. Kellmereit and D. Obodovski, The Silent Intelligence – The Internet of Things, 1 edition ed., DND Ventures LLC, 2013, p. 168.|
|||A. Bassi, M. Bauer, M. Fiedler, T. Kramp, R. van Kranenburg, S. Lange and S. Meissner, Enabling Things to Talk Designing IoT solutions with the IoT Architectural Reference Model, Springer, 2013.|
|||A. Serbanati, C. M. Medaglia and U. B. Ceipidor, “Building Blocks of the Internet of Things: State of the Art and Beyond,” in Deploying RFID – Challenges, Solutions, and Open Issues, InTech, 2011.|
|||G. D. Abowd and A. K. Dey, “Towards a better understanding of context and context-awareness,” in HUC ’99: Proceedings of the 1st international symposium on Handheld and Ubiquitous Computing, 1999.|
|||S. Camazine, J.-L. Deneubourg, N. R. Franks, J. Sneyd, G. Theraula and E. Bonabeau, Self-Organization in Biological Systems, Princeton University Press, 2001.|
|||J. Arlow and I. Neustadt, Enterprise Patterns and MDA: Building Better Software with Archetype Patterns and UML, Addison-Wesley Professional, 2004, p. 528.|
|||A. Hessellund, “Domain-Speciﬁc Multimodeling,” 2009.|
|||T. L. Koreshoff, T. Robertson and T. W. Leong, “Internet of things: a review of literature and products,” in Proceedings of the 25th Australian Computer-Human Interaction Conference: Augmentation, Application, Innovation, Collaboration, New York, 2013.|
|||J. Gubbi, R. Buyya, S. Marusic and M. Palaniswami, “Internet of Things (IoT): A vision, architectural elements, and future directions,” Future Generation Computer Systems, vol. 29, no. 7, p. 1645–1660, September 2013.|
|||J. S. Veen, B. Waaij and R. J. Meijer, “Sensor Data Storage Performance: SQL or NoSQL, Physical or Virtual,” in CLOUD 2012: 2012 IEEE 5th International Conference on Cloud Computing, 2012.|
|||C. Perera, A. Zaslavsky, P. Christen and D. Georgakopoulos, “Context Aware Computing for The Internet of Things: A Survey,” pp. 1-41, 2013.|
|||EU project ELLIOT, “Experiential Living Lab for the Internet of Things,” [Online]. Available: http://www.elliot-project.eu/. [Accessed april 2014].|
|||H. Duin, J. B. Hauge and K.-D. Thoben, “An ideation game conception based on the Synectics method,” On the Horizon, vol. 17, no. 4, pp. 286 – 295, 2009.|
|||R. Shapiro and S. Khoshafian, iBPMS Intelligent BPM Systems: Impact and Opportunity, Future Strategies Inc., 2013, p. 220.|
|||M. Dumas, W. van der Aalst and A. H. ter Hofstede, Process-aware information systems: bridging people and software through process technology, New York: John Wiley & Sons, Inc., 2005.|
|||A. H. M. ter Hofstede, W. van der Aalst and M. Adams, Eds., Modern Business Process Automation: YAWL and its Support Environment, Springer, 2010, p. 676.|
|||L. Yoga Systems. [Online]. Available: http://www.yogasystems.com/.|
|||T. Beale , “Archetypes: Constraint-based Domain Models for Future-proof Information Systems,” in Proceedings of the 11th OOPSLA Workshop on Behavioural Semantics., 2002.|
|||L. C. Jain and C. Faucher, Innovations in Intelligent Machines-4: Recent Advances in Knowledge Engineering (Studies in Computational Intelligence), 2014 edition ed., Springer, 2013, p. 375.|
|||H. Fei and H. Qi, Intelligent Sensor Networks: The Integration of Sensor Networks, Signal Processing and Machine Learning, CRC Press, 2013, p. 674.|
|||J.-P. Vasseur and A. Dunkels, Interconnecting Smart Objects with IP: The Next Internet, San Francisco, CA, USA: Morgan Kaufmann Publishers Inc., 2010.|
|||P. Harrington, Machine Learning in Action, Manning Publications, 2012.|