h-Life: an intelligent personal health assistant

Health Informatics Journal http://jhi.sagepub.com

h-Life: an intelligent personal health assistant D.G. Nanou, A.M. Poulis, A. Likas and D.I. Fotiadis HEALTH INFORMATICS J 2002; 8; 20 The online version of this article can be found at: http://jhi.sagepub.com/cgi/content/abstract/8/1/20

Published by: http://www.sagepublications.com

Additional services and information for Health Informatics Journal can be found at: Email Alerts: http://jhi.sagepub.com/cgi/alerts Subscriptions: http://jhi.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav

Downloaded from http://jhi.sagepub.com at PENNSYLVANIA STATE UNIV on February 7, 2008 © 2002 SAGE Publications. All rights reserved. Not for commercial use or unauthorized distribution.

04 Q-HIJ8.1_Nanou.qxd

26/4/02

11:08 am

Page 20

h-Life: an intelligent personal health assistant D.G. Nanou, A.M. Poulis, A. Likas and D.I. Fotiadis

In this paper the architecture and subsystems of the h-Life platform are described. The system is currently under implementation and therefore all system components and modules are described according to their design specifications. h-Life aims at developing a user-friendly, personalized and ubiquitously accessible knowledge-based system that will provide citizens with an intelligent tool for monitoring the state of health and lifestyle. The h-Life system is based on the integration of several components: ●

●

●

●

●

●

D.G. Nanou BSc Medical Technology and Software Development Unit Email: [email protected] A.M. Poulis BSc MSc Student Email: [email protected] A. Likas PhD MSc BSc Lecturer Email: [email protected] D.I. Fotiadis PhD MSc BSc Assistant Professor Email: fotiadis@ medcs.cs.uoi.gr Department of Computer Science University of Ioannina and Ioannina Biomedical Research Institute 45110 Ioannina, Greece

a knowledge repository related to health, nutrition, fitness and lifestyle a user repository where personal health records and user profiles are stored a plan repository for the storage of medical plans an intelligent adviser and plan selection mechanism, which provides advice and plans based on the user profile an alerting and reminding module that issues alerts (to ensure the timely execution of a selected plan) and notifications concerning events of potential interest a customer-facing module, which ensures personalized interaction with the system and customizes the user-system interaction according to the user profile.

Based on the latest technology in intelligent and communication systems, the h-Life environment aims at providing the user with personalized, ‘certified’ lifestyle information and consultancy, tailored to individual needs and available anywhere and at anytime. Keywords Intelligent systems, knowledge-based systems, personal health records, personalization, mobile access

Increasingly-capable telehealth systems and the Internet are not only moving the point of care closer to the patient, but the patient can now assume a more active role in his or her own care. Coupled with the migration of the healthcare industry to electronic patient records and the emergence of a growing number of enabling healthcare technologies (e.g. novel biosensors, wearable devices and software agents), these technologies demonstrate unprecedented potential for delivering ubiquitous, highly automated and intelligent healthcare services [1]. Current trends and continuously increasing developments in the area of health monitoring and decision support systems promote the use of Internet technologies and Web-based systems for customized advice and consultancy, tailored to individual needs. We already observe several communities of patients with conditions such as cardiovascular disease, diabetes, renal failure, asthma or obesity, as well as healthy individuals (with low or high risk factors), turning to the Internet to seek out the latest available information, advice and support to monitor their lifestyle and level of health [2] [3]. The increasing role of IT in the delivery of patient-centred healthcare has led to the development of a wide range of applications and platforms, targeting the individual needs for health and lifestyle information and monitoring [4] [5] [6] [7]. Advances in telecommunication technologies [8] [9], telemedicine platforms and services [10] [11] [12], human–computer interaction [13] [14], knowledge discovery [15] [16] [17], clinical decision support [2] [3] [18] and health monitoring systems [2] [3] [19], motivated our research and led to the design and implementation of the h-Life system. Building an advanced health telematic system always involves consideration of wellestablished standards that specify: ● ● ●

●

●

INTRODUCTION ●

Worldwide, the healthcare industry is experiencing a substantial paradigm shift with regard to patient-centred homecare, due to the convergence of several areas of technology.

Health Informatics Journal (2002) 8, 20-28

●

medical aspects, such as diagnostic and intervention coding (e.g. UMLS) medical informatics aspects on architecture and security (CEN/TC251) technical aspects such as communication protocols (e.g. TCP-IP, WAP, future UMTS) safety aspects, usually determined by agencies such as the International Standards Organization or the US Food and Drug Administration (FDA) communication aspects, such as standards for information modelling and exchange (e.g. XML) security and privacy aspects, usually set up by national legislation digital signatures and database aspects (e.g. International Agency for Cancer Research and WHO specifications).

Downloaded from http://jhi.sagepub.com at PENNSYLVANIA STATE UNIV on February 7, 2008

2002 SAGE Publications. rightsLondon reserved. Not 7NX for commercial use or unauthorized © The Continuum Publishing Group Ltd 2002, The©Tower Building, 11 YorkAll Road, SE1 and 370 Lexington Avenue, distribution. New York, NY 10017, USA.


26/4/02

11:08 am

Page 21

h-Life: an intelligent personal health assistant

In the lifestyle information domain, most of the existing products are books with advice on nutrition and exercise or Web-based systems, offering consultation and in some cases appropriate schedules upon customer request. Only a limited number of these systems provide extra functionalities such as user profiling, information categorization and intelligent search, support and updating of users’ health records, alert and reminding mechanisms, customization features and easy ubiquitous access through various terminals and network. Nevertheless, no system exists to integrate all the above-mentioned features and even those that provide a subset of them address either specific aspects of healthy lifestyle or refer to a limited audience. Employing state of the art technologies, standards, existing platforms and implementations, we designed and are currently developing a personal health assistant system. h-Life is an intelligent health environment that provides patients and citizens in general with a tool that can be used for the monitoring and improvement of their health and lifestyle. By providing a technologically innovative platform for monitoring and interaction with the customer, the h-Life system integrates all actors (citizens, patients, medical doctors, nutrition companies, gym chains and hospitals) involved in the healthy lifestyle industry. The platform combines a large scale of functionalities including quality assurance and certification of provided information, user profiling (medical, cultural, lifestyle and personal assets), customized plan selection, plan execution monitoring, customized issuing of alerts and reminders, evaluation of plans, user-friendly interaction and easy access through various terminal devices and network interfaces. In this paper we describe the overall system architecture, the system modules, the related databases and its functionality. Details and methods used to achieve the above functionality are described elsewhere [20] [21].

Current status of the system Data from operational trials are not currently available as the whole system is still under development. The system has first been engaged in an identification and descriptive analysis of main and specific user needs and of existing constraints. The primary user groups targeted by the project include: ● ● ●

© The Continuum Publishing Group Ltd 2002.

healthy individuals individuals with minor health problems (high/low risk factors) patients

21

Among the three generic categories the following specific cases were selected: ● ● ● ● ● ●

individuals/persons with no known health problems patients suffering from cardiovascular diseases diabetic patients (and insulin-dependent diabetics) patients with renal failure patients with asthma overweight individuals

All cases (except the second, that is patients with cardiovascular diseases) will be faced both at the ‘prevention level’ (providing information on prevention and specific clinical tests required to diagnose the disease) and at the ‘monitoring level’ (providing therapeutic/supporting medical plans to improve the health condition). Patients suffering from cardiovascular diseases were considered a critical and extremely complex case in terms of monitoring and interdependencies with other diseases. It was therefore decided that this case would be faced only at a ‘prevention’ level. A functional description and analysis of current and future needs has been carried out (concerning both functional and non-functional requirements) as well as a detailed study of the state-of-the-art related technologies, products and standards. Currently, we are at the phase of implementing the system knowledge repository. All potential legal and ethical implications or constraints related to medical procedures have been addressed, so as to ensure that the system will not be considered as an automatic diagnosis and medical advice system. Therefore the system aims to be mainly provided and promoted through organized service providers (e.g. diet clubs, insurance companies, gym chains, hospitals) as an added value supporting service. All users will always supervised by responsible medical professionals (even users that are not customers of these service providers but ordinary Internet users) that observe their progress and take appropriate actions if any need arises.

SYSTEM ARCHITECTURE The h-Life system exploits all technological advances in the areas of intelligent systems, knowledge engineering and personalized information delivery. The main functionality of the system is achieved through the interaction and co-operation of several intelligent components. In addition, the delivery of critical health and lifestyle information independently of the loca-



22

26/4/02

11:08 am

Page 22

Health Informatics Journal

tion of the user is based on an intelligent multimodal mechanism for information filtering and delivery. The h-Life system presents a flexible and open architecture that can easily be extended to accommodate future needs. It consists of two knowledge repositories: the health information repository and the user repository. These repositories are equipped with knowledge processing mechanisms and their functionality is supported by other intelligent components such as the user-profiling module, the advising module, the information filtering module and the alerting and reminding mechanism. The system functionality is shown in Figure 1. On the right hand side, the impact of the h-Life system to various actors and aspects of healthy lifestyle is depicted and related to the overall system functionality. The health information repository contains information relating to a healthy lifestyle: diet and exercise plans, medical advice and guidance, clinical protocol instructions and rules. This information is user-independent, provided by the knowledge suppliers of the system (hospitals, health centres, the nutrition industry, etc.) and is appropriately

certified, classified and stored. The user repository stores user-related knowledge. This knowledge mainly concerns the personal medical record of each user, as well as the plans and medical instructions that he or she follows. The intelligent user profiling module classifies the user according to his or her personal characteristics, interests and medical history. It constitutes a key component of the system, since it provides the other modules with criteria for information filtering, in order to define whether some specific information item in the knowledge repository may be of interest to a user. Another important component that is responsible for the intelligent functionality of the system is the consultancy and advice subsystem. This subsystem implements several mechanisms: ●

●

The Plan Selection Mechanism, which accepts as input the profile and medical record of a user and suggests several plans that match the user characteristics. The Consultancy and Advising Mechanism, that provides advice to the user either upon request or when specific updates

Fig. 1 h-Life system impact and functionality © The Continuum Publishing Group Ltd 2002.



26/4/02

11:08 am

Page 23


●

have been performed on either the user’s health record or the Medical Knowledge Repository. The Monitoring, Alerting and Reminder Module, that is responsible for the regular execution of the medical plans followed by the user.

Another important module responsible for the customization of the information presented to the user is the ‘Communicator’ module. Communicator is the actual ‘customer-facing’ mechanism of the h-Life platform according to the type of network connection and the functional capabilities of the employed device.

information is hierarchically organized into specific domains where each domain corresponds to each of the above areas of interest. The knowledge of each domain is stored in the form of relations, rules and plans. The repository also contains unstructured information in the form of free text, which is accessed using appropriate keywords. The information of the health knowledge repository may be accessed in a variety of ways: ●

● ●

OVERALL ARCHITECTURAL APPROACH In this section we present h-Life’s architectural approach and the basic functionalities and technical features of each system component, shown in Figure 2.

23

●

Users may simply retrieve information on a specific subject (i.e. diabetes, symptoms, medication etc). Users may ask the system for advice on a specific health problem they face. Users may ask the system to propose a plan in order to improve their health and lifestyle (diet plan, exercise plan, etc.). The system may asynchronously decide (without any previous user inquiry) to warn users, in cases where the monitoring of their health indicates that some indices exceed the normal values (i.e. weight gain instead of loss), or to inform the user on new products and services.

Health knowledge repository

User repository

The health knowledge repository constitutes the core of the system. It stores user-independent knowledge related to diseases, risk factors, nutrition, exercise, healthcare providers, healthy lifestyle services and products, etc. To facilitate knowledge engineering, the stored

The user repository stores personal information concerning the users of the system. More specifically, for each user a separate record is created where the medical history, the followed health plan and the profile of the user are stored. The database also stores temporal information (e.g. measurements) with which the user (or the doctor) periodically provides the system. The user repository is a key component of the system, since it stores critical personal information that is exploited by other modules of the system. Any update of the user record is considered as an event that may trigger some rules in the knowledge repository or in the plan-monitoring module. Moreover, this information is used to extract statistical data related to the success of specific plans or the effectiveness of various treatments. This statistical information is accessible by medical, pharmaceutical and lifestyle experts and content providers supporting and improving, in this way, their decision-making procedures.

Intelligent user profiling The profiling module is used for the classification of a user based on medical condition and personal characteristics. This module provides as output a user profile, which is stored in the user database and is subsequently employed in any information-processing Fig. 2 System overview © The Continuum Publishing Group Ltd 2002.



24

26/4/02

11:08 am

Page 24


operation concerning this user. The module is activated when a user accesses the system for the first time and prompts the user to fill in a profile form, with information regarding general health condition, special needs, individual preferences, particular interests, financial status and other relevant characteristics. Using this information, the intelligent mechanism of the module extracts user-specific features: a set of user characteristics (keyword attributes) that are subsequently inserted as keywords in all user-related operations. The user profile also contains information related to the cultural habits of the users (e.g. eating habits, local environmental conditions, work characteristics, and so on). This information is either directly entered to the system by the end-user or retrieved from databases with relevant information according to the user’s location.

Plan evaluation module

Consultancy and advice module

Plan monitoring, alerting and reminder module

This module provides the user with advice either upon request (synchronous communication) or in the form of unexpected messages (asynchronous user notification). It is the module that actually combines the two large repositories of the system. In the case of asynchronous notification, it monitors any update events in the user database and forwards these events to the knowledge repository along with user profile information. Then it identifies the rules with conditions matching the profile and takes appropriate action, for example it may issue a warning to the user or notify them about events or new products that may be of interest. In the case of synchronous communication, the module operates in a similar way, the only difference being that the input to the knowledge repository comes from an explicit user query. The module identifies all answers (advices) that match the query and then performs information filtering using the profile characteristics (keyword attributes) so that the user finally receives explicitly customized information.

This is responsible for the evaluation and assessment of the predefined services and health plans in the knowledge repository. Indexes and measurable targets are identified in collaboration with medical and lifestyle experts. Their values are computed and compared with health indices (such as blood pressure, BMI, or other disease-specific indices) stored in the database to indicate the success of a healthy lifestyle plan. If the index value satisfies specific health criteria (rules extracted from medical expertise) then the plan is acceptable. The process evaluates the followed plan according to personal user profile and modifies the classification of the specific plan, if that need arises. The whole system has been developed on the basis of classification and matching rules.

The alerting system is activated during the monitoring phase of the regular execution of plan that a user has committed to follow. Through a temporal knowledge-based system, the system checks whether the plan instructions were followed or not. The alerting module issues notifications to the user concerning the actions to be taken in the near future, and where the user does not follow the plan, it sends alerts and warnings both to the user and the corresponding doctor, using the communicator module described in the next section. Furthermore, alerts are activated in cases where updates take place to the medical database (e.g. insertion of new products), as well as in cases where a user profile has been modified (such as during insertion of new update measurements). A user may select a preferred way of receiving alerts (e.g. receiving SMS form alerts through a mobile phone or by email). An alert is sent in the following instances: ●

Plan selection mechanism The plan selection mechanism is a rule-based module which uses the rules extracted from the user’s profile database and plan repositories. Upon user request, and using information from the user database, the module searches the plans database in the knowledge repository, identifies the most relevant plans (using the rule-based matching mechanism) and displays these plans to the user. If the user agrees to follow a plan, then the module automatically updates the plan record in the user’s database.


●

Medication. The system can remind the hLife user to take a prescribed medication. The alert reminds him which medication plan must be followed every day as well as the specific time schedule. Diet. This applies to those following a diet proposed by the h-Life system. Frequently, this diet changes, depending on the person’s level of health and the supervising doctor’s suggestions. The role of the alert system in this case is to inform users about the date on which the new diet will start. The system continuously monitors the users’ health condition and when necessary it decides that the diet should be



26/4/02

11:08 am

Page 25


●

●

changed earlier (e.g. if the user gains 6 to 8 kilos since they were last weighed instead of losing them). In this case the alert should inform the specific person about the unexpected and unscheduled change. In addition an appropriate message is sent to the supervising doctor or the Content Provider who supports this plan (so as to update the efficiency and rating of that plan). Doctor’s appointments and checkups. This reminds users of their scheduled health tests (such as glucose or blood test) or visits to the doctor. Products. The system informs users on new products or services added in the h-Life system database that may be of particular interest (according to each user’s personal profile). These products can be nutrition supplements, new gymnastic centres, pharmaceutical products or medical equipment/device (e.g. manometer, thermometer, etc.).

The alert and reminding mechanism will utilize a variety of terminal devices and network interfaces. According to the module functional design, personal computers, mobile phones, PDAs and possibly other communication devices (UMTS-enabled) will be used to remind users of online guidance, interactive checklists, record-embedded ‘to do’ lists and workflow commitments.

Communicator: ‘customer-facing’ module The communicator is a ‘customer-facing’ module that enables personalized and cus-

Fig. 3 Communicator architecture


25

tomizable access according to the user’s profile, network connection, interaction device capabilities and position. The communicator creates an abstraction level for documents, user-facing equipment and network type, providing a personalized access services from terminals using various networks. This allows the same application to be accessed on a wide variety of user-facing equipment and provides a common and expandable method of interfacing applications to the next generation UMTS networks and terminals. A brief overview of the communicator’s architecture is given in Figure 3. The communicator comprises the following four basic modules: COM-data. This includes the User Preferences and Capabilities Repository containing information about device capabilities as well as the user preferences. All these features will be defined and handled in a mark-up based format and therefore can be easily expanded and updated. COM-device. This module provides the framework for classifying devices based on characteristics such as the preferred language, sound on/off, images on/off, class of device (PC, mobile, palmtop etc.), screen size, available bandwidth, version of XML/HTML supported by the device. This information will be collected from device manufacturers. The device schema is based on user-agent profiles and consists of description blocks for key components such as: hardware platform, software platform, browser attributes, network characteristics, protocol characteristics and a set of attributes pertaining to WAP capabilities supported by the device. This includes details on the capabilities and characteristics related to the WML Browser, WTA, and so on. Additional components may be added to the schema to describe capabilities pertaining to other user agents such as an email application or hardware extensions. This framework is applied to describe receiver and sender capabilities and preferences, and also document characteristics. To describe a profile, the communicator uses predicate expressions (‘feature predicates’) on collections of media feature values (‘feature collection’) as an acceptable set of media feature combinations (‘feature set’). The communicator uses a vocabulary of common and compatible features in order to be able to match a document profile against a receiver profile and determine compatibility between them. COM-connection. This is used to define several types of information delivery formats based on user global position, network capabilities and



26

26/4/02

11:08 am

Page 26


priority of information, e.g. this could be higher or lower resolution images and small file sizes in case of high priority or vital information. When information is sent to a Web browser, the decision would be higher quality and multi-media output. In the case of using a small screen device (e.g. cell phone or PDAs), the decision would be faster transmission and smaller file size. In case of emergency and submission of high priority vital information the decision could be batch-mode information delivery to all available user-facing equipment. COM-interaction. This includes intelligent internal server-side interactions, as the user and device profile information needs to be matched with the document profile. COMdata profiles (based on XML) express the required functionality for what the system perceives as optimal rendering, and how the user wants them to be rendered. This procedure takes care of the global user position, network/protocol capabilities and status, supporting bi-directional (Web access) and userdirectional interactions (SMS messages).

●

●

●

MEDICAL DATABASES AND KNOWLEDGE MANAGEMENT One of the most significant tasks in the h-Life system is the specification of the health knowledge repository and the design and implementation of the independent medical databases. These databases – currently under implementation – contain essential information items for adapted and personalized health plans and also information in the domain of health and healthy lifestyle. A categorization of the databases that contain user-independent information is given below: ●

●


Healthy products database. Contains information related to healthy nutritional products as well as nutritional supplements and vitamins. The database includes information on each product such as price, places to buy it, nutritional information, caloric value, contraindications (examples where use must be avoided are suggested). Diet plans database. Contains a variety of diet plans classified according to duration, target group, scope and possible results, in order to enable the system to easily select the most appropriate diet according to the individual user profile and the target user group. The particular way in which diets must be stored in order to be effectively retrievable is a critical aspect. Access to the

●

●

registered data is available only to authorized nutritional experts. The content providers participating in the project have authorized access only to their own medical plans and are responsible for updating the information. Physical exercise plans database. This database stores plans concerning independent physical exercise. These plans consist of text with instructions as well image/video previews. They are categorized according the features of the target group (e.g. physiotherapy), goals to be achieved (e.g. strengthening of specific muscles, rehabilitation) and level of difficulty. Service providers database. Contains information related to the healthy lifestyle product and service providers. A classification of service providers is the following: healthy lifestyle product providers, gyms, hospitals, doctors, dieticians, pharmacies. Pharmaceutical products database. This database will be developed at a later stage of the project. The project view was to reuse existing work concerning medication and drugs. The current thinking is to have a licence agreement with other actors that have developed systems on drugs or pharmaceutical products and just implement an interface module so that the h-Life users can retrieve personalized information on these products. Small medical devices database. Designed to provide information on medical devices that can be used for everyday individual measurements and monitoring (e.g. pressure-gauge, instant test strips, spyrometer) as well as other equipment (such as health monitoring or wearable devices). Health conditions database. Provides information on several health conditions, such as diabetes or asthma: symptoms, the clinical tests required for diagnosis of the condition and other valuable information.

Other information provided by the h-Life system concerns news, articles and links to scientific and medical magazines, prevention advice, first aid instructions and statistical information concerning plans’ effectiveness and users’ satisfaction. The latter is only accessible by the medical professionals and the content providers. A significant component in the h-Life knowledge management mechanism is the decision-support system. h-Life’s decisionsupport mechanisms fall into two categories: ●

Mechanisms that assist healthcare and lifestyle experts to determine the current situation: ‘what is true about a patient’



26/4/02

11:08 am

Page 27


●

(user categorization, correct diagnosis and plan monitoring) Mechanisms that assist in making decisions about the actions that should be performed: ‘what to do for the patient’ – (plan selection, modification of plans, consultancy on medication and alerting).

The first case is based on the ‘fixed data’, available in the databases and can be performed offline. The final results are presented to interested decision makers. On the other hand, determination of the appropriate action for a patient is made ‘online’, combining the user provided information and a set of critical decision rules, extracted from medical and lifestyle provider’s experience.

INFORMATION PERSONALIZATION AND MAINTENANCE The h-Life system acts as a ‘personalized’ health monitoring and disease prevention system (not an automatic diagnosis system) based on user profiles, providing a service tailored to the individual user’s preferences or interests. The h-Life user model is a knowledge source, which contains explicit assumptions on all user-related aspects that may be relevant for the behaviour of the system. The intelligent user-profiling module is an essential part of the system whose function is to incrementally build up a user model, to store, update or delete entries, to maintain the consistency of the model and to supply the h-Life system with assumptions about the user. The preferred manner in using the profile is to store information that enables personalization on an individual basis. A content-based filtering method is applied to lifestyle-related metadata (extracted from the core databases of the system) in order to evaluate the selected document’s relevance by matching the keywords contained in a user profile with the keywords extracted from the data. Citizens, whether ill or not, individuals, families or groups accessing the system will have direct, easy and customized access to information related to their health and lifestyle. They are also able to update their personal medical record with new data from personal monitoring devices, drug prescriptions and medical diagnoses and they will get personalized high quality advice. h-life users (customers of health clubs, health centres and individuals can access the system through the WWW and get relative information over second and third generation mobile platforms such as UMTS. © The Continuum Publishing Group Ltd 2002.

27

On the h-Life implementation, the most efficient location for storing this type of information is the server so the user profiles do not have to be transmitted through the network. Social or collaborative filtering is another effective way of taking advantage of user profiles. This method collects the user profiles of a group of people and generates recommendations based on the similarities of the profiles. In order to implement collaborative filtering, hLife establishes an intelligent matching mechanism that compares all user profiles (using as start-up a predefined set of user features) and extracts grouping information. The same procedure is applied to the communicator module components in order to classify the ‘user facing’ equipment according to a predefined and supported set of device capabilities.

CONCLUSIONS AND FUTURE WORK The h-Life system focuses on a major and novel technological area. It can be described as a platform for personalized health information delivery to the individual. This area is very important from the health point of view, but the approach could also be expanded to other areas such as tourism, business, education or administration. An important technological aspect is the exploitation of the capabilities of third generation mobile protocols for the provision of advanced services through a wide range of communication devices. The alerting and reminding mechanism embedded into the system will drive new options in organizations which are experimenting with user interaction agents, offering users online guidance, interactive checklists, record-embedded ‘to do’ lists, workflow commitments, scheduled meetings and events and case-based reusable artefacts. Users will have the opportunity to improve and monitor their lifestyle in a secure and safe way. Healthcare professionals and service providers (such as gym chains or insurance companies) will be provided with a unique tool for the reorganization of customer monitoring methods, providing their customers with an added value service. The health and lifestyle industry will be able to promote services and products and obtain significant feedback so as to improve the quality of their work. h-Life will help healthy lifestyle providers participate in new market opportunities, improve the quality of their products and develop new products that match customer preferences. Thus h-Life enhances links among customers and healthy lifestyle product



28

26/4/02

11:08 am

Page 28


and service providers and opens new business opportunities. The h-Life system will in the near future be integrated and expanded to provide increased functionalities.

Acknowledgements The h-Life project is funded by the European Commission (IST-2000-26353).

REFERENCES [1] Warren S, Craft R L. Designing smart health care technology into the home of the future. Sandia National Laboratories, 1999. www.sandia.gov/cis/6200/telemedicine/docs/futurehome.pdf [2] Shortliffe E H, Perreault L E (eds). Medical informatics: computer applications in health care and biomedicine. Heidelberg: Springer-Verlag, 2001. [3] Van Bummel J H, Musen M A. Handbook of medical informatics. Heidelberg: Springer-Verlag, 1997. [4] BodyMedia, http://www.bodymedia.com [5] WebMDHealth, http://my.webmd.com [6] Community Care Managed Healthcare Plans, http:// www.ccmhp.com [7] Health Net, http://www.healthnet.com [8] The path towards UMTS: technologies for the information society, UMTS Forum, 1998. [9] The future mobile market: global trends and develop-


ments with a focus on Western Europe, No. 8 Report from the UMTS Forum, 1999. [10] TELMED. The impact of telematics on the healthcare sector, http://www.ehto.org/ht_projects/telmed/toc.htm [11] Telemedicine in the 21st Century: Opportunities for citizens, society and industry. An International Space University Workshop, 4–5 November 1999, Strasbourg, France. [12] Dick R S, Steen E B, Detmer D E. The computer-based patient record: an essential technology for health care. Washington: Institute of Medicine, National Academy Press, 1991. [13] Silverman B G. Computer reminders and alerts. IEEE Computer 1997; 30(1): 42-49. [14] Benyon D. Navigation in information spaces: supporting the individual. Proceedings of INTERACT ‘97, Sydney, Australia, 1997. [15] Giles C et al. (eds.). Neural networks and hybrid intelligent models. IEEE Transactions on Neural Networks 1998; 9, Sept. [16] Fayyad U et al. Advances in data mining and knowledge discovery. Cambridge, MA: MIT Press, 1996. [17] Abernethy N F, Wu J J, Hewett M, Attman R B. Sophia: a flexible Web-based knowledge server. IEEE Intelligent Systems and Their Applications 1999; 14(4): 79-85. [18] Schizas C (ed.). Computational intelligent diagnostic systems in medicine. Journal of Intelligent Systems 1998; 8(1-2). [19] Szolocovic P, Doyle J, Long W L, Kohane I, Pauker S G. Guardian angel: patient-centered health information systems. Technical Report MIT/LCS/TR-064, Massachusetts Institute of Technology, Laboratory of Computer Science, 1994. [20] The h-Life reference architecture. Technical report, Ioannina Biomedical Research Institute, 2001. [21] Intelligent user profiling: the h-life approach. Technical report, Ioannina Biomedical Research Institute, 2001.


h-Life: an intelligent personal health assistant

h-Life: an intelligent personal health assistant

Suggest Documents

intelligent personal digital assistant solution - Harman International

intelligent personal digital assistant solution - Harman International

Personal Intelligent Travel Assistant A distributed approach

HealthPal: An Intelligent Personal Medical Assistant for Supporting the ...

Socially-Aware Animated Intelligent Personal Assistant Agent

An Intelligent Assistant for Context-Aware Adaptation of Personal ...

An Intelligent Assistant for Patient Health Care - CiteSeerX

Towards a Personal Health Management Assistant

An Intelligent Multi-Agent Memory Assistant - CiteSeerX

YPA - An Intelligent Directory Enquiry Assistant - CiteSeerX

An Intelligent Assistant for Public Transport Management

An Intelligent Assistant for Environmental Planning ...

1982 - RABBIT: An Intelligent Database Assistant - CiteSeerX

An Intelligent Sales Assistant for Configurable Products

An Argumentation-Based BDI Personal Assistant

An Eye-Tracking based Personal Assistant for

Towards Intelligent Personal Health Record Systems - ScienceDirect

A Personal Email Assistant

Smart Personal Assistant

Smart Personal Assistant

My Personal Assistant - Smashwords

AliMe Assist: An Intelligent Assistant for Creating an Innovative ... - arXiv

Health Care Application Based on Intelligent Assistant Agent

Assistant Personal Robot - Semantic Scholar