• Non ci sono risultati.

Using Multi-Agent Systems for Healthcare and Social Services

N/A
N/A
Protected

Academic year: 2022

Condividi "Using Multi-Agent Systems for Healthcare and Social Services"

Copied!
20
0
0

Testo completo

(1)

Using Multi-Agent Systems for Healthcare and Social Services

ABSTRACT

Multi-agent systems have been importantly contributing to the development of the theory and the practice of complex distributed systems and, in particular, have shown the potential to meet critical needs in high-speed, mission-critical content-rich and distributed information applications where mutual interdependencies, dynamic environments, uncertainty, and sophisticated control play a role. Therefore, multi-agent systems can be considered a suitable technology for the realization of applications for providing healthcare and social services where the use of loosely coupled and heterogeneous components, the dynamic and distributed management of data and the remote collaboration among users are often the most relevant requirements.

1. INTRODUCTION

Multi-agent systems are one of the most interesting areas in software research and they have been importantly contributing to the development of the theory and the practice of complex distributed systems (see, e.g., Jennings et al., 1995; Muller, 1998; Bordini et al., 2005) and, in particular, have shown the potential to meet critical needs in high-speed, mission-critical content-rich and distributed information systems where mutual interdependencies, dynamic environments, uncertainty, and sophisticated control play a role (Gasser, 2001). Application for healthcare and for providing social services can take outstanding advantage of the intrinsic characteristics of multi-agent systems because of notable features that most healthcare applications share: (i) they are composed of loosely coupled (complex) systems; (ii) they are realized in terms of heterogeneous components and legacy systems; (iii) they dynamically manage distributed data and resources; and (iv) they are often accessed by remote users in (synchronous) collaboration (Moreno & Nealon, 2003; Annicchiarico et al., 2008).

The goal of this chapter is to describe the main reasons why multi-agent systems are considered one of the most interesting technologies for the development of applications for healthcare and social services. It provides some guidelines intended to help identifying the kinds of applications that can truly take advantage of the features of multi-agent systems, and it presents some of the most important international projects that used multi-agent systems in the healthcare and social services domain.

2. BACKGROUND

While a number of definitions have been proposed for identifying a “software agent” and a

“multi-agent system” (see, e.g., Russell & Norvig, 2003; Wooldridge & Jennings, 1995;

Genesereth & Ketchpel, 1994), there is not a single one which researchers generally agree on.

Nevertheless, there is a common understanding that an agent is essentially an autonomous software entity that should at least be designed to operate continuously in dynamic and uncertain environments, reacting to events while showing an intelligent behaviour to pursue its own objectives. An agent usually provides interoperable interfaces for interacting with other agents, either concurrently or cooperatively, exchanging messages formulated according to some syntax,

(2)

semantics and pragmatics. Since an agent behaves proactively, it requires some degree of trust by its user, and it can receive delegations from either human users or other agents in the form of required actions or desired goals, matched with permissions to access necessary resources.

Additionally, some agents may also be able to perform complex reasoning at run-time and also learn and change their behaviour over time, to improve their performances. Mobile agents are even able to move for one computational node to another, to follow their own users or to exploit some local resource more efficiently. Agent-based systems are often realized by loosely coupling various agents, i.e. autonomous software entities, thus modelling a proper multi-agent system, characterised by a higher level of modularity and a richer descriptive model, if compared with a solitary agent working within its environment – either with the presence of users or not.

As agents are often described in terms of the generic properties that they should exhibit, instead of a precise programming interface or algorithm, they are better thought as abstractions capable of capturing the essence of many software systems at different levels of detail, rather than a single technology supporting the realization of distributed intelligent systems. In this sense, diverse systems in the fields of Artificial Intelligence, Databases, Operating Systems and Computer Networks can be coherently considered as first-class multi-agent systems. In particular, agents and multi-agent systems are often considered the highest system level (Newel, 1982;

Jennings, 2000) that we can access today and they are meant to provide a truly novel level of abstraction in the analysis, design and implementation of complex software systems (Bergenti &

Huhns, 2004). Since the agent-based nature of a system comes from the characteristics of its components and of the interactions between them, multi-agent systems are often developed using technologies that have been recently provided for the main purpose of realizing highly interoperable software systems, e.g., Web services, and that have no built-in notion of agent.

2.1 Sociality

Sociality is one of the main distinguishing characteristic of multi-agent systems. In particular, multi-agent systems allow the delegation of goals and tasks as a mean to realize emerging social behaviours. Multi-agent systems are generally considered an appropriate abstraction for modelling complex, distributed systems, even if such a multiplicity naturally introduces the possibility of having different agents with potentially conflicting goals. Agents may decide to cooperate for mutual benefit, or they may compete to serve their own interests. Agents take advantage of their social ability to exhibit flexible coordination behaviours that make them able to both cooperate in the achievement of shared goals or to compete on the acquisition of resources and tasks. Agents have the ability of coordinating their behaviours into coherent global actions.

This characteristic fit quite well the new environment where health care and social services are being provided. In fact, significant societal changes are the mark of these years, including ageing of population, growing value attributed to personal care, preference for living at home also when needing assistance, interconnection and integration of diverse services.

Care provision is usually based either on the effort of relatives, or on hospitals and care centres.

In recent years, both those approaches are being questioned and the situation is changing. In fact, the support of relatives is becoming less and less viable, as family members are frequently occupied with work. On the other side, the relocation of patients, elders and people needing

(3)

social care, outside of their communities is not desirable for their own wellness; in many cases it is not even necessary, as they preserve enough strength and capabilities to stay at home.

Moreover, rooms in hospitals and care centres are in a limited number and the trend is toward their reduction, in spite of a growing number of elders. As a consequence, the traditional approach to provide cares and social services in hospitals and institutional centres is being paralleled with a growing tendency to provide cares in the community, directly. In this kind of

“Community Care” or “Domiciled Care”, services to assisted people are provided directly in their home or in their habitual environment, allowing them to continue to live as much independently as they can. The range of services varies from the delivery of meals and other goods to domestic help, from generic home monitoring and assistance to nursing and medical care. In some cases, thanks to the diffusion of computer networks and computing devices, those activities allow some degree of automation.

Usually, those services are supplied by a number of organisations and agencies, which act largely in an independent way, with different responsibilities and goals. An important fact which is not always emphasised enough, is the large support provided by family, community members and other informal carers who can be paradoxically excluded from the rest of the community care.

However, all those agencies and informal carers need to coordinate their actions, in order to avoid serious service inefficiencies, and so they need the availability of systems to share information about the assisted, without breaching the official confidentiality requirements. The coordination of these autonomous bodies and of single professionals is a challenging problem, that could result in misunderstandings among the involved professionals, ignorance of important notions about the assisted, fragmentation of care, overlapping and duplication of assessments.

Involving autonomous entities, the communication process aimed at reaching a mutually accepted agreement on the overall care management is necessarily a negotiation process. This kind of process is exactly one of the scenarios where multi-agent systems can be applied more conveniently (Jennings, 2001). According to the different goals of the agents involved, which in general are to be considered autonomous, negotiation can be either competitive or cooperative. In particular, contracting is a common negotiation mechanism used to assign a desired task to the most appropriate agent among those available. In this mechanism, agents can take on two roles, i.e., manager and contractor, in a decentralized market structure; a manager have some task to assign, while a contractors can propose itself as the most appropriate executor (Smith & Davis, 1980). The mechanism is basilar for multi-agent systems, and allow an agent to solve an assigned problem without using local resources and capabilities, because it is not possible, practical or convenient. In this case, it would decompose the problem into sub-problems and it would try to find other agents with the necessary resources and capabilities to solve such sub-problems. The communication protocol for assigning sub-problems consists of three steps: first of all, the manager agent announces the contract; in response to the announcement, contracting agents submit their bids; finally, the manager agent evaluates the submitted bids and assigns the sub- problem to the chosen contractors.

2.2 Planning and organization

The current tendency is to manage the provision of cares in complex systems, based on the interaction of a number of different entities, including neighbourhood and institutional care centres, day centres and social security institutions, each of them involving people with various

(4)

roles e.g. health care professionals, social care assistants, assisted people and their relatives. It is not a tendency mainly driven by technology, but nevertheless technology will play a fundamental role in the realization of the underlying information systems, for the overall care management.

Computer networks and innovative software tools, in fact, are able to support the realization of integrated systems which greatly ease the collaboration among the various involved agencies and persons, driving the evolution of those systems towards so-called “virtual organizations”, where the various involved humans operate as part of a virtual community.

In general, organizational structuring is a coordination technique which allows the definition of the organization that governs the interaction among the agents of a system, i.e., the organization that defines the information, communication, and control relationships among the agents of the system (Horling & Lesser, 2005). In the context of integrated care systems, multi-agent systems represent quite a natural solution, as they intrinsically represent autonomous entities and overall

“virtual organizations”, as those concerning the case of health and social services.

Similar organizational structure have been applied to many other fields, as advanced supply chains, virtual enterprises, virtual research laboratories, communities of tech enthusiasts and professionals. In the health and social care sector, this representation describes the collaboration and smooth interaction among all agencies and persons involved in the process. In all these cases, where different but related organisations coexist and cooperate, multi-agent systems have been widely adopted as the best fitting paradigm.

In multi-agent systems, managing and monitoring the distributed execution of the tasks needed to implement the desired functionalities is often based on organizational structuring techniques, together with contracting and multi-agent planning. While contracting solves the problems of searching and selecting the agents that can perform the set of tasks necessary to achieve a goal, it does not give any help in defining the way, i.e., the new plan, in which they are executed to achieve it. Instead, multi-agent planning techniques enable agents to elaborate plans guiding them towards their common/individual goals preventing any possible interference among the actions of different agents (Durfee, 1999; Tonino et al., 2002). In order to avoid inconsistent or conflicting actions and interactions, agents build a multi-agent plan that details all the future actions and interactions required to achieve their goals, and they interleave execution of such a plan with needed planning and re-planning. Multi-agent planning can be either centralized or distributed. In centralized multi-agent planning, there is a coordinating agent that, upon the receipt of all partial plans from individual agents, analyses them to identify potential inconsistencies and conflicting interactions, e.g., conflicts between agents over limited resources. The coordinating agent may eventually attempt to modify such partial plans and then it combines them into a multi-agent plan where conflicting interactions are eliminated. In distributed multi-agent planning, the idea is to provide each agent with a model of other agents plans. Agents communicate in order to build and update their individual plans and the models of other agents until all conflicts are sorted.

Aside of specific health and social services, relevant important changes are happening also at the production and distribution level, with many supply chains that are changing their structures and dynamics. Some of those changes can help to solve some typical problems of social services, such as delivering meals and products at home, providing assistance at home or in the neighbourhood etc. In fact, other than cost efficiency and on-shelf availability, additional parameters are acquiring importance in the design of supply chains, such as traffic congestion in

(5)

urban areas, energy consumption, carbon emissions and the permanent rise in transportation costs. As a result, there is a trend toward information sharing in collaborative supply chains, collaborative warehousing, especially in the form of consolidation centres at the boundary of cities, collaborative distribution and neighbourhood delivery, in cities as well as non-urban areas, including home delivery and pick-up (Global Commerce Initiative & Capgemini, 2008). Quite obviously, the openness, autonomy and negotiation capabilities of software agents are features which adapt very well to all these application areas, too.

2.3 Communication

One of the basic problems of health and social services is the communication among the all persons involved. Assisted people often have a complex mixture of health and social care needs.

Thus, several different professionals may be involved in their care, including General Practitioners, nurses, and social workers. Above all in the case of community care, these professionals might belong to different organisations. Each team, consisting of different health and social care specialists and professionals, must share its expertise and knowledge, including the information gathered during the various assessments, in order to monitor constantly the health conditions of the assisted person and to plan a more accurate care process.

An important issue that most e-health services address regards the possibility of jointly supporting professionals in their highly specialized work. Computer-Supported Cooperative Work (CSCW) is already common practice in telesurgery and teleassistance and it seems an important ingredient of next-generation e-health services. Notably, the inherent cooperative nature of agents and the very fact that many CSCW technologies are already based on agents is another important contribution of multi-agent systems to e-health.

Agent technology, and in particular mobile agents, provide a means by which effective co- operation (information sharing and communication between autonomous information systems) can take place without compromising the security of the assisted and the agencies involved, particularly in the highly volatile environment of community care. Since each agent has complete autonomy it can respond according to the rules of the organisation it represents, providing an effective and assured guardian that is totally under that organisation’s control. No other architectural framework for Distributed Information Systems gives this capability without serious reliability problems.

The integration of the assessment process requires a common language to be used, to support aggregating and sharing information about the health and social care for the assisted people. A common language is also the basis for planning services and monitoring trends in needs and results. Information systems help to automate some of the administration tasks, such as the organization of visits and other aspects of the management of the work, allowing the professionals to concentrate over the actual cares for the assisted person. Furthermore, information systems simplify the involvement of assisted people in their health and social care. In fact, the integration of the assessment process implies the assisted person and their carer are kept informed about their personal care plan, also considering particular problems and with more detailed assessment if appropriate.

(6)

Since simplified user interaction and communication among users are central to e-health and similar systems, this immediately promotes multi-agent systems as ideal candidates to support next-generation e-health services and social applications. Similarly, the central venue that security and privacy-awareness have in multi-agent systems again stresses their importance of them with respect to e-health. In the agents realm, the issues of privacy-awareness are treated under the umbrella of the more expressive notion of trust. Likewise, social services strongly remark the importance of preserving confidentiality and guaranteeing a high level of security for classified information about patients.

2.4 Mobility and system interoperability

Moreover, technological changes intertwine into this scenario, with widespread adoption of various types of computing devices, ranging from laptop and lighter mobile devices to mobile phones, accompanying users through different locations and different countries. Accordingly, e- health deals naturally with mobile users, e.g., in teleassistance scenarios, and it is common understanding that e-health should transparently accommodate fixed and mobile users. So called m-health services should be accessible to anyone, anywhere, anytime, anyhow. In mobile scenarios, various devices can be used to collect, transmit and process vital patients’ data, e.g., heart rate and blood pressure, in real time (De Mola et al., 2006). Such systems are especially important in applications that remotely monitor patients with chronic ailments or in homecare.

Broadly speaking, such systems are designed to access medical information in a mobile and ubiquitous setting. This access may be either (i) the retrieval of relevant medical information for use of healthcare practitioners, e.g., a hospital doctor on his/her ward round; or (ii) the acquisition of patient-generated medical information, e.g., telemonitoring the patient’s health state outside the hospital. In both cases, it is extremely important to ensure that the person retrieving or generating information could interact with a ubiquitous and pervasive e-health system without any obstruction or adaptation of the normal workflow or style of working. The most notable characteristic that such systems should exhibit are: (i) Context and location awareness are to be smoothly integrated, i.e., the access and the visualization of health-related information always depend on the overall contexts of the patient and of the user (Bricon-Souf & Newman, 2007), (ii) fault-tolerance, reliability, security and privacy-awareness are a must in order to accommodate the strict requirements of all healthcare applications, (iii) effective mobile devices are to be used to provide access to relevant health-related information independently of the current physical location and physical condition of the user, and (iv) unobtrusive sensor technology is needed to enable the gathering of physiological information from the patient without hampering his/her daily life.

All in all, mentioned requirements immediately recall the characterizing features of multi-agent systems and it comes with no surprise that many ubiquitous and pervasive e-health systems are realized using multi-agent abstractions and technologies. In particular, the JADE framework and its lightweight version JADE-LEAP (Bergenti et al., 2001) do take special care of transparently and dynamically allocating users and agents on heterogeneous network of different types of devices.

An important issue in e-health is about supporting the interoperability of (legacy) medical information systems in order to enable the integrated provision of services for accessing

(7)

information from different, remote sources. The dream of a single, universally-accepted middleware supporting the development of new services together with the renewal of legacy services was quickly abandoned. Nowadays recent technologies, that were originally intended to support the (semantic) interoperation between heterogeneous services, are commonly adopted in practice. This, again, emphasizes the role of multi-agent systems for providing important contributions to e-health, because of the inherent semantics-awareness of the interaction between agents, which make them ready to deliver semantic interoperability.

A lot of work has been done in the last decade for spreading the use of multi-agent systems for the realization of software applications and services, with particular attention to interoperability among agent platforms and with other systems. Several technological specifications are the results of such work. Among them, the two main results to date are: (i) FIPA specifications (FIPA, 2000), a set of specifications intended to support the interoperability between heterogeneous agent-based systems; and (ii) an agent development framework, called JADE (Bellifemine et al., 2008; JADE, 2008), that implements FIPA specifications and that supports interoperability between agents using consolidated technologies, e.g., Java and CORBA.

3. MULTI-AGENT SYSTEMS FOR E-HEALTH AND SOCIAL SERVICES

The fact that software systems for health and social services are often based on the multi-agent paradigm is easily explained on the basis of the structural and interaction models describing those systems, which closely resemble multi-agent features. The appropriateness of multi-agent systems for this kind of application can be also demonstrated by adapting the well-known grouping criteria proposed in (Barnes & Uncapher, 2000). In this sense, multi-agent systems can greatly contribute to health and social services from three points of view: (i) facilitating the access to services; (ii) improving the quality of services; and (iii) reducing costs.

The long-waited universality and equality of access to healthcare and social assistance, especially for geographically or socially isolated patients, can be promoted by the quick and contextualized access to healthcare-related information from anywhere, at anytime and in the most appropriate way allowed by modern telecommunication networks and distributed, intelligent software applications. In fact, the inherent decentralization that is always assumed in multi-agent systems is vital to facilitate the provision of services also in everyday scenarios, including cares to elderly, disabled or chronic patients. In many cases, this possibility is still uncommon today, for the current lack of basic infrastructures as universally-accessible communication networks and power supplies, but the widespread adoption of multi-agent systems at homes can help reducing medical visits and related waitlists drastically. Finally, agents are good tools to help patients following preventive strategies and supporting self-care on a day-by-day basis. In fact, the proactive nature of agents assists in creating a trusted link between patients and the assistance system, by having agents constantly pushing valuable information to patients, with no need of explicit demands.

With respect to the overall quality of provided services, the most important contribution of multi- agent systems lies essentially on the possibility of feeding highly specialized healthcare professionals with the right information, at the right time, tailored to the patient. The proactive nature of agents and their semantic interoperability support such a need with the possibility of feeding users with information acquired from diverse sources and tailored to the concrete patient at hand. Thanks to the computerization of health records, that is now common practice in

(8)

Western Countries, the transfer of complex health records globally and quickly increases the accessibility, unifies the information at every stage of complex healthcare processes and improves care continuity. Moreover, the longstanding tradition of expert systems that still lives behind the scenes of multi-agent systems can support healthcare professionals in using the provided information for taking the right decisions at the right time. Finally, the transparent integration of mobile terminals helps collecting data to quickly support contextualized healthcare decisions.

Finally, the reduction of the costs of healthcare processes is receiving a growing attention, because of the resources required for the provision of quality healthcare. Also from this point of view, multi-agent systems may be beneficial. The mentioned possibility of agents to provide the right information, at the right time, tailored for the patient, supports efficiency in the overall management of treatments. Moreover, the semantic interoperability of agents enables instant acquisition of information from its natural source, with minimal (if not null) pass of information along chains of intermediaries. Finally, the trusted and privacy-aware support that agents provide to healthcare processes is a valuable means to speed up and optimize many administrative procedures. All in all, we can summarize the contribution that multi-agent systems provide to healthcare from the point of view of cost reductions with an incisive proposition: earlier assistance and structured prevention of the causes of further care.

As is described above, multi-agent systems can become a key ingredient of the next-generation e- health services and applications, however, in the last ten years they are already used for realizing e-health applications and, in particular, for realizing assistive living, diagnostic, physiological telemonitoring, smart hospital and smart emergency applications. In the following, some of the most interesting applications will be introduced.

3.1 Community care and social assistance

The provision of health and social services directly in the community of the assisted people is nowadays preferred by both professionals and patients. In this sense, the integration of smart- home automation is an essential aspect of assisted living, above all for chronically ill patients, elderly, impaired, psychologically or socially diseased people (Mynatt et al, 2000; Liffick, 2003, Mann, 2005). All projects in this domain are notably challenging because they need to naturally accommodate many, if not all, the challenging aspects of the kinds of applications mentioned above. Elderly people tend to suffer from at least one chronic disease, which requires telemonitoring. Additional age impairments make independent living at home difficult and therefore assistance for daily activities is required. Moreover, the additional context information provided by a smart-home environment enhances a better interpretation of physiological sensor information, e.g., whether the patient is running or sleeping has significant influence on the blood pressure. All in all, multi-agent systems are the common denominator of the kinds of uses that we listed above and this is the reason why we believe that they are a key factor for the coherent and successful development of assistive living applications. Two of the most interesting projects are:

CASIS (Jih et al., 2006) and K4CARE (K4CARE, 2007).

VCAP (Sanz Angulo & de Benito Martín, 2011) is namely a Virtual Child Abuse Prevention platform. The main idea behind the project is that the same paradigm driving business towards so-called “Virtual Enterprises” also stands for some in other non-business oriented contexts,

(9)

where “Dynamic Virtual Organization” could be appropriate. VDOs, in this system, come into existence after a new case of abuse has been identified. In this case, a set of institutions and persons must be selected, for providing the personalized answer to the child abuse situation.

Then, selected partners need to work together and coordinate their actions in the new VDO.

When finally an adequate response to the initial problem has been provided, the DVO is dissolved. The system is implemented over JADE, with the inclusion of the JESS rule-engine, for the implementation of the decision modules of agents.

SUPREMA (Vittikh et al., 2007) is a multi-agent system used for the management of social services in the Russian region of Samara. The primary motivation was the large mass of regional and federal laws governing various social benefits, partly overlapping and even contradicting each other. The system is based on personal passports of users and special smart cards, which allow them to access different governmental databases. Ontologies are used to integrate the different legacy sources of information. Citizens are associated with software agents which work uninterruptedly for supporting their access to social services, including education, work, health, security, culture and sport. Other agents are associated with social officers, social organizations, and even social laws. In fact, a new agent is created after each new law is passed, and then it autonomously searches for citizens interested in the new law.

INCA (Beer, 2003) is a system for the management of integrated community care, for monitoring and assisting elder people. Like other systems, it is aimed at integrating different agencies and professionals into a unified system, for communication and collaboration. For this purpose it envisages the adoption of a single agent-based care monitoring facility, to be used by all

professionals, and of cooperative structures and automated agents for planning and managing the care programs. It has been developed as a prototype using the ZEUS agent framework, in the context of the Agentcities project (Agentcities, 2003).

CASIS (Jih et al., 2006) is an event-driven service-oriented and multi-agent system framework whose goal is to provide context-aware healthcare services to the elderly resident in the intelligent space. CASIS framework allows remote caretakers, such as concerned family members and healthcare providers, to closely monitor and attend to the elder’s physical and mental well-beings anytime, anywhere. The smart environment interacts with the elder through a wide variety of appliances for data gathering and information presentation. The environment tracts the location and specific activities of the elder through sensors, such as pressure-sensitive floors, cameras, bio-sensors, and smart furniture. Meanwhile, the elder receives multimedia messages or content through speakers, TV, as well as personal mobile devices. The caregivers may access the elder’s health and dietary information through any Web enable device like a PC or PDA. Context-aware computing enables the environment to respond at the right time and the right place, to the elder’s needs based on the sensor data collected. The environment is further equipped with integrated control for convenience, comfort and safety. In particular, CASIS is able to infer the status of the elder and performs appropriate actions. For example, upon sensing that the elder has fallen asleep, it turns off the TV, and switches the telephone into voice mail mode. It informs and plays back any incoming messages when the elder wakes up.

K4CARE (K4CARE, 2007) is a research project whose main objective was to combine healthcare and ICT experiences coming from several western and eastern European countries to create, implement, and validate a knowledge-based healthcare model for the professional

(10)

assistance to senior patients at home. The main step of the project was to develop a healthcare model to guide the realization of an integrated system of healthcare services for the care of the elderly, the disabled persons, and the patients with chronic diseases. The interaction between health professionals, computer scientists, technology centres, and SMEs has been crucial to define the model and providing detailed information about the selected prototype services. The K4CARE model provides a paradigm easily adoptable in all European countries, being all the proposed structures filtered according to national laws.

TeleCARE (Camarinha-Matos et al., 2003) is the result of an IST project for the design and development of virtual elderly support in community environments. In particular, it supports supervision tasks, intelligent assistance and agenda management, integration and provision of data extracted from distributed sources. It leverages a number of technologies, including multi- agent systems, federated information management, safe communications, hypermedia interfaces, rich sensorial environments, increased intelligence of home appliances, and collaborative virtual environments. It is based on the Aglets agent framework with an extension to support the FIPA standard, on JESS as a rule engine, on ontologies for the definition of a set of basic entities.

3.2 Telemonitoring and aided diagnosis

Telemonitoring, the continuous monitoring of patients, also during their daily activities, is becoming an extremely important application domain in the context of e-health, mainly because of the progression of chronic ailments in the aging society (Meystre, 2005). First, such applications enable healthcare institutions to manage and monitor the therapies of their patients.

Second, they serve as instrument for performing large-scale medical researches and studies.

Third, they support the timely activation of emergency services in case of severe health conditions. Due to the nature of such applications, that are continuously monitoring physiological signals, unobtrusiveness and mobility of the patient are key requirements. Moreover, these applications can offer additional comfort services, like assistive services, information services and communication services, which leads us naturally to the adoption of multi-agent systems for the realization of physiological telemonitoring applications (Rialle et al., 2003). Five of the most interesting projects that use multi-agent systems are: MobiHealth (MobiHealth, 2004), U-R- SAFE (U-R-SAFE, 2005), AID-N (Gao et al., 2007), MyHeart (Amft & Habetha, 2007) and SAPHIRE (Laleci et al., 2008).

Diagnosis is probably the first area of healthcare where computer systems and artificial intelligence techniques were used (Buchanan & Shortliffe, 1984). Diagnosis support systems often need the integration of different sources of data and the on-line or off-line collaboration of different kinds of users. In fact, they are enhanced by telemonitoring systems and provide telemonitoring systems the reasoning capabilities needed to timely raise alarms when severe conditions are detected. These features make multi-agent systems a reference model and technology for their realization Three of the most interesting projects that used multi agent systems for the realization of diagnosis applications are: IHKA (Hashmi et al., 2002), OHDS (Hadzic et al., 2006) and HealthAgents (Croitoru et al., 2007).

MobiHealth (MobiHealth, 2004) is an innovative software platform for Body Area Networks (BANs) is the heart of the architecture of project MobiHealth (MobiHealth), which was funded by the European Commission under the 5th Framework Programme (FP5). It provides plug and

(11)

play sensor connectivity and it handles related issues such as security, handovers and quality of service. It enables monitoring, storage, and wireless transmission, by means of GPRS and UMTS technologies, of vital signals data coming from the BAN of the patient. Possible hardware platforms for this architecture are PDAs or programmable mobile phones which can serve as Mobile Base Units (MBUs). The investigated application scenario is telemonitoring of patients at home. Vital signals are measured and are transmitted along with audio and video to healthcare service providers. The MobiHealth service and application platform enables monitoring, storage and transmission of vital signs data coming from the patient BAN. The platform supports flexible personalization of services and ensures appropriate intervention in response to certain conditions or combinations detected in the vital signs measurements.

U-R-SAFE (U-R-SAFE, 2005) is a research project with the goal of realizing a telemonitoring environment for elderly people and patients with chronic diseases. The project developed a portable device which continuously monitors physiological signals (heart activity, oxygen saturation, and fall detection) and which is able to send an alarm to a medical center if an abnormality is detected. The technology issues tackled in this project cover sensor devices and wireless communication. The concept is to have the elderly person wearing medical measuring devices, all connected via short range Wireless Personal Area Network (WPAN) to a central, portable electronic unit, the so called Personal Base Station (PBS). The short range wireless connection is done using the Ultra Wide Band technology. It is worth noting that, at home or at the hospital, two wireless networks are interconnected. The first is the WPAN worn by the person and the second is the wireless indoor LAN, installed in the house or at the hospital, which allows the PBS communicating via a dedicated gateway to the fixed-access network using radio links.

AID-N (Gao et al., 2007) is a light-weight wireless medical system for triage. The overall goal of the AID-N electronic triage system is to efficiently gather and distribute information on the vital signs and locations of patients in an extremely fault tolerant manner. Typically, monitoring systems like AID-N consist of (i) a central server that medical doctors use to verify the overall conditions of patients, and (ii) portable clients–one for each patient–that patients use to send information about their condition. Such instruments delegate to the patient the task of providing relevant data obtained from classical sensors. Moreover, monitoring devices are not normally capable of proactively operating to autonomously detect anomalies in the conditions of patients, i.e., they always need the direct intervention of the patient. Finally, such devices are quite general and they are not able to adapt to the very specific needs of each patient. In other words, such systems would really benefit from an agent-oriented approach that would support personalization and proactivity.

MyHeart (Amft & Habetha, 2007, MyHeart, 2008) is a research project whose focus was on preventing cardiovascular diseases by intensively applying e-health applications. The work focused, in particular, on the telemonitoring scenario, where sensors integrated in clothing are used to monitor heart activity and the physical activity of the patient. This project emphasizes the importance of specialized sensor and hardware devices to allow unobtrusive measurements. The approach is therefore to integrate system solutions into functional clothes with integrated textile sensors. The combination of functional clothes and integrated electronics to process them on- body defines what MyHeart consortium calls Intelligent Biomedical Clothes. The processing consists of making diagnoses, detecting trends and react on it. MyHeart also comprises feedback devices that are able to interact with the user as well as with professional services.

(12)

SAPHIRE (Laleci et al., 2008; SAPHIRE, 2008) is a research project whose goal was to develop a multi-agent system for the monitoring of chronic diseases both at home and at hospital using a semantic infrastructure. The system is capable of deploying and executing clinical guidelines in a care environment including sparse care providers having heterogeneous information systems. The SAPHIRE multi-agent system addresses such challenges through an enabling semantic interoperability environment.

IHKA (Hashmi et al., 2002) is a healthcare knowledge procurement system based on the use of multi-agent technologies. This system is based on six different agent types. These types are the user interface agent, an agent to convert the search result into a viable format for passing to the UI agent, a query optimising agent which optimises the query, the knowledge retrieval agent that performs the query, the knowledge adaptation agent to adapt the knowledge to the current circumstances and the knowledge procurement agent which if all else fails searches the web for the knowledge. In particular, IHKA features autonomous knowledge gathering, filtering, adaptation and acquisition from some healthcare enterprise/organizational memories with the goal of providing assistance to non-expert healthcare practitioners.

OHDS (Hadzic et al., 2006) is a system that supports the doctors in the diagnostic, treatment and supervision processes of the evolution of new epidemics, based on the exploration of all data pertinent to each case and on the scientific data contained in various professional databases.

OHDS combines the advantages of the holonic paradigm with multi-agent system technology and ontology design, for the organization of unstructured biomedical research into structured disease information. Ontologies are used as ’brain’ for the holonic diagnostic system to enhance its ability to structure information in a meaningful way and share information fast. A fuzzy mechanism ruled by intelligent agents is used for integrating dispersed heterogeneous knowledge available on the web and so, for automatically structuring the information in the adequate ontology template.

HealthAgents (Croitoru et al., 2007; HealthAgents, 2007) is a research project with the goal of improving the classification of brain tumours through multi-agent decision support over a secure and distributed network of local databases. HealthAgents does not only develop new pattern recognition methods for distributed classification and analysis of in vivo MRS and ex vivo/in vitro HRMAS and DNA data, but it also defines a method to assess the quality and usability of a new candidate local database containing a set of new cases, based on a compatibility score. Using its multi-agent architecture, HealthAgents applies cutting-edge agent technology to the biomedical field and it provides an infrastructure for the so called HealthAgents network, a globally distributed information and knowledge repository for brain tumour diagnosis and prognosis.

3.3 Smart-hospitals, emergency, care centres

When located at caregiver’s site, e-health applications are often known as smart-hospital applications. Such applications try to improve the daily activities of doctors and nurses. This is commonly done by providing tools to access patient records or, more generally, clinical information systems, as well as to schedule and track patients and hospital resources in a wireless, mobile, and context-aware manner. It is worth noting that recent projects introduce the use of RFID technology to further improve these applications, as exemplified in the Jacobi case

(13)

study (Fuhrer & Guinard, 2006). Another case of e-health at the caregiver’s site regards emergency management by means of mobile devices. Emergency physicians are able to access the records of their patients in advance while they are still in the ambulance car approaching the location of the patient. If we also consider the triggering of emergency situations and access of current physiological signals, these applications spans the bridge between the caregiver’s and patient’s site. Three of the most interesting projects that use multi-agent systems are: ERMA (Mabry et al., 2004), Akogrimo (Akogrimo, 2007) and CASCOM (Schumacher & Helin, 2008).

HeCaSe (Moreno et al., 2006) is a multi-agent system developed at University Rovira i Virgili, in Spain, in the context of the Agentcities project (Agentcities, 2003). In its latest incarnation it is named HeCaSe2. The system provides secure access to medical records; support for user queries about the medical centres, medical units, or doctors available in a certain area; online set up of specialist visits. The involved professionals, in turn, automatically receive the appropriate medical records of patients. Agents in the system represent users, medical-centres, departments, medical centre unit, doctors. Other agents are used to manage medical-records; finally a broker agent is used as a sort of façade for the overall system in front of a user agent. Enhancements in the latest versions include specific agents for the management of clinical guidelines.

ERMA (Mabry et al., 2004) has the purpose of providing meaningful diagnoses and intervention suggestions to the healthcare personnel acting on behalf of the patient in the cases of emergency trauma with particular emphasis on types of shock and stabilization of arterial blood gases. This system ERMA is based on a set of agents that act as a collaborative team of specialists to realize the monitoring and diagnostic infrastructure for dynamically collect filter and integrate data and reasoning about them through a hybrid approach of fuzzy logic, causal Bayesian networks, trend analysis and qualitative logic.

Akogrimo (Akogrimo, 2007) is a research project whose main goal is the integration of the next generation Grids (NGG) with the next generation networks. The application scenarios of Akogrimo cover smart hospitals, telemonitoring and emergency assistance. The Akogrimo NGGs are able to deal with an environment with rapidly changing context such as bandwidth, device capabilities, and location. Furthermore the architecture of Akogrimo can be immediately deployed in unlicensed mobile access environments such as hot-spot infrastructures, because it assumes a pure IP-based underlying network infrastructure. Target users of an Akogrimo healthcare information system are (i) people demanding mobile ad-hoc and pervasive healthcare services, e.g. due to emergencies or chronic diseases; and (ii) healthcare service suppliers/institutions, i.e., stationary or mobile professionals, including healthcare advisors, pharmacies, nursing services, hospitals, emergency service devices and emergency stations. The reference scenario of Akogrimo covers objective like: (i) early recognition of heart attacks and fast treatment by combining vital parameters monitoring; (ii) aberration and emergency detection;

and (iii) subsequent rescue management. Agents are spread all over in the architecture of Akogrimo, ranging from the end-user application, i.e., Akogrimo Personal Assistants, to infrastructural agents supporting interactions between personal assistants.

CASCOM (Schumacher & Helin, 2008; CASCOM, 2008) is one of the most recent attempts to bring the notable characteristics of agents to e-health. CASCOM is a technology-driven project that brings together three notable new technologies: multi-agent systems, Semantic Web services and peer-to-peer middleware in the scope of mobile and context-aware environments. The main

(14)

delivery of CASCOM project is a general-purpose, open-source middleware that implements a generic architecture, that provides easy, seamless and contextualized access to Semantic Web services anytime, anywhere and using any device. It supports professionals in taking the best decision on a treatment, also with no background knowledge of the patient’s medical history.

Instead, it provides physicians with contextualized information on the fly, acquiring it as needed directly from its source. In fact CASCOM agents interact directly with the Semantic Web services that organizations provide to access needed information in a secure and privacy-aware manner.

5. FUTURE TRENDS AND CONCLUSIONS

The applications for modern healthcare and social services are distributed and complex by their nature. In this kind of scenarios, multi-agent systems have been proved as one of the most convenient paradigms. In particular, multi-agent systems should be considered as a model which provides a novel level of abstraction, rather than a single technology. The approach based on multi-agent systems is concretized into different technologies depending on the concrete needs of various advanced applications. This allows describing various projects and applications that concretely use diverse technologies in terms of agents and multi-agent systems, and in particular the paradigm fits quite well the e-health and social services scenarios, too. In fact, the overall management of cares in vast open environments requires shared and distributed decision making processes, which are based on the communication and coordination of complex and diverse forms of information between involved organizations and people. Therefore, the coupling of multi- agent systems and social cares is quite natural: multi-agent systems have the suitable features for modelling and realizing current and future applications in the health and social fields; by converse, multi-agent systems are offered the suitable requirements for experimenting at best their applicability and, so, for receiving a great contribution to their own evolution and success.

Nevertheless, to provide valuable contribution to healthcare, ICT solutions have to face important challenges. Among the open issues, some are identifiable in any application domain, such as user expectations and acceptance, and lack of centralised control; others are typical of the healthcare domain, such as legal and ethical issues like privacy, integrity and authentication in the exchange of patient information between agents. In general, the adoption of advanced ICT solutions is not happening at the rapid pace fostered by the early optimism. Problems originate from the inherent difficulties of having ICT accepted in healthcare and social services from the technical, social, ethical, political, legal and economical points of view. This is a well discussed topic in the literature and interested readers can refer to some notable works (Barnes & Uncapher, 2000;

Deloitte 2006; Jadad et al., 2000; Laxminarayan & Stamm, 2002; Ohinmaa et al., 1999; Wilson et al., 2004). The distance between the actual research and the real status of the health and social system poses several problems to a wide adoption of ICT. Another obstacle, for the widespread adoption of advanced ICT solutions in healthcare and social services, is the lack of data and methodology for the economic evaluation of the impact of new technologies (e.g. the reimbursement for provided services is not well defined).

In conclusion, the application of a multi-agent approach to e-health and social services doesn’t imply that all those problems are definitely solved. But most of those problems are intrinsic to the application scenario, and the solid theoretical model of multi-agent systems only provide a better ground to identify and trace them. Real programs built on the multi-agent paradigm are still

(15)

evolving towards a complete maturity, and the variety and complexity of the e-health scenario make it one of most interesting application fields, able of verifying the advantages of their use and of conditioning their evolution.

According to Altman (Altman, 1999), one of the most important challenges, in the field of Artificial Intelligence, regards having medical records “based on semantically clean knowledge representation techniques.” Most model of autonomous software agents provide semantic communication as a basic feature. Moreover, thanks to their intrinsically distributed nature and their autonomy, agents provide a clean way to make such records available anywhere, at any time. Agents will improve healthcare organizations, by supporting doctors and caregivers. But the use of agent technology will mean quite a radical and general change in how healthcare and assistance will be provided, with a large impact on patients, too. Finally, agent technology can be a key component for realizing the envisioned “smart use of computation” (Annicchiarico et al., 2007). In all these senses, multi-agent systems will greatly benefit from the current work on their integration with some emergent technologies like, e.g., Web services (Greenwood, D., & Calisti, 2004; Nguyen, 2005), the Semantic Web (Hendler, 2001; Chen et al., 2004) and workflows (Buhler & Vidal, 2005; Negri et al., 2006), that are already and/or will be fundamental components of healthcare and social services (Bicer et al., 2005; Song et al., 2006).

REFERENCES

Agentcities (2003). Testbed for Worldwide Agent Network: Research and Development.

Retrieved March 15, 2011, from http://www.ist- world.org/ProjectDetails.aspx?ProjectId=fbba2dd384a746de8c72b1ef71158b58.

Akogrimo (2007). Akogrimo project Web site. Retrieved May 15, 2011, from http://www.mobilegrids.org.

Amft, O., & Habetha, J. (2007). The MyHeart Project. In L. Langenhove (Ed.), Smart textiles for medicine and healthcare, (pp. 275-297). Cambridge, UK: Woodhead Publishing.

Sanz Angulo, P., & de Benito Martín, J.J. (2011). Developing a Multi-agent Software to Support the Creation of Dynamic Virtual Organizations aimed at Preventing Child Abuse Cases. In Multi- Agent Systems - Modeling, Interactions, Simulations and Case Studies.

Annicchiarico, R., Cortés, U., & Urdiales, C. (Eds.). (2008). Agent Technology and e-Health.

Whitestein Series in Software Agent Technologies and Autonomic Computing. Babel, Switzerland: Birkhäuser Verlag.

Barnes, G.A., & Uncapher, M. (2000). Getting to e-Health: The Opportunities for Using IT in the Health Care Industry. Arlington, VA: Information Technology Association of America.

Retrieved May 15, 2011, from http://www.itaa.org/isec/ehealth/ehealthfinal.pdf.

Beer, M., Hill, R., Huang, W., & Sixsmith, A. (2003). An agent-based architecture for managing the provision of community care - the INCA (Intelligent Community Alarm) experience. In AI Commun. 16, 3 (August 2003), 179-192.

Bellifemine, F., Caire, G., Poggi, A., & Rimassa, G. (2008). JADE: a Software Framework for Developing Multi-Agent Applications. Lessons Learned. Information and Software Technology Journal, 50,10-21.

(16)

Bergenti, F., Poggi, A., Burg, B., & Caire, G. (2001). Deploying FIPA-Compliant Systems on Handheld Devices. IEEE Internet Computing, 5(4), 20-25.

Bergenti F., & Huhns M. N. (2004). On the use of agents as components of software systems. In F. Bergenti, M.P. Gleizes, & F. Zambonelli (Eds.), Methodologies and Software Engineering for Agent Systems, (pp. 19-31). New York, NY: Kluwer.

Bicer, V., Kilic, O., Dogac, A., & Laleci, G.B. (2005). Archetype-Based Semantic Interoperability of Web Service Messages in the Health Care Domain. International Journal on Semantic Web and Information Systems, 1(4):1-23.

Bordini, R. H., Dastani, M., Dix, J., & El Fallah Seghrouchni, A. (Eds). (2005). Multi-Agent Programming: Languages, Platforms and Applications. Berlin, Germany: Springer Verlag.

Bricon-Souf, N., & Newman, C. (2007). Context awareness in health care: A review.

International Journal of Medical Informatics, 76(1), 2-12.

Buchanan, B.G., & Shortliffe, E.H. (1984). - Rule Based Expert Systems: The Mycin Experiments of the Stanford Heuristic Programming Project. Boston, MA: Addison-Wesley.

Buhler P.A., & Vidal, J.M. (2005). Towards Adaptive Workflow Enactment Using Multiagent Systems. Information Technology and Management, 6(1), 61-87.

Camarinha-Matos, L.M., Castolo, O., Rosas, J. (2003). A multi-agent based platform for virtual communities in elderly care. In Emerging Technologies and Factory Automation, 2003.

Proceedings. ETFA '03. IEEE Conference , vol.2, no., pp. 421- 428 vol.2, 16-19 Sept. 2003.

CASCOM (2008). CASCOM project Web site. Retrieved May 15, 2011, from http://www.ist- cascom.org.

Chen, H., Finin, T., Joshi, A., Perich, F., Chakraborty, D., & Kagal, L. (2004). Intelligent Agents Meet the Semantic Web in Smart Spaces. IEEE Internet Computing, 6(8), 69-79.

Croitoru, M., Hu, B., Dasmahapatra, S., Lewis, P., Dupplaw, D., Gibb, A., Julia-Sape, M., Vicente, J., Saez, C., Garcia-Gomez, J.M., Roset, R., Estanyol, F., Rafael, X., & Mier, M. (2007).

Conceptual Graphs Based Information Retrieval in HealthAgents, Computer-Based Medical Systems, 7(20-22), 618-623.

Deloitte Center for Health Solutions (2006). Promoting Physician Adoption of Advanced Clinical Information Systems: A Deloitte Point of View. Retrieved May 15, 2011, from http://www.deloitte.com/dtt/cda/doc/content/us_chs_cis_adoption_21406.pdf

Durfee, E. H. (1999). Distributed problem solving and planning. In G. Weiss (Ed.), Multiagent Systems: A Modern Approach To Distributed Artificial intelligence, (pp. 121-164). Cambridge, MA: MIT Press.

European Commission (2003). European eHealth Ministerial Declaration. Retrieved May 15, 2011, from http:/ec.europa.eu/information_society/eeurope/ehealth/conference/2003.

FIPA. (2008). FIPA Specifications. Retrieved May15, 2011, from http:www.fipa.org.

Fuhrer, J.P., & Guinard, D. (2006) Building a Smart Hospital Using RFID Technologies. In Proc.

of the 1st European Conference on eHealth (ECEH06), (pp. 131-142), Fribourg, Switzerland.

Gao, T., Massey, T., Selavo, L., Crawford, D., Chen, B.R., Lorincz, K., Shnayder, V., Hauenstein, L., Dabiri, F., Jeng, J., Chanmugam, A., White, D., Sarrafzadeh, M., & Welsh, M.

(17)

(2007). The Advanced Health and Disaster Aid Network: A Light-weight Wireless Medical System for Triage. IEEE Transactions on Biomedical Circuits and Systems, 1(3), 203-216.

Gasser, L. (2001). MAS Infrastructure Definitions, Needs, and Prospects. In Wagner, T., Rana, O. (Eds.), Infrastructure for Agents, Multi-Agent Systems, and Scalable Multi-Agent Systems. (pp.

1-11). Berlin, Germany: Springer Verlag.

Genesereth, M.R., & Ketchpel, S.P. (2004). Software agents. Communications of the ACM, 37(7), 48–53.

Global Commerce Initiative, Capgemini (2008). Future Supply Chain 2016 Report. Serving Consumers in a Sustainable Way. Retrieved May 15, 2011, from http://www.futuresupplychain.com/downloads/.

Greenwood, D., & Calisti, M. (2004). Engineering Web Service-Agent Integration. In Proc. of the IEEE International Conference on Systems, Man and Cybernetics, (pp.1918-1925). Hague, The Netherlands.

Hadzic, M., Chang, E., & Ulieru, M. (2006). Soft computing agents for e-health applied to the research and control of unknown diseases. Information Sciences, 176, 1190-1214.

HealthAgents (2008). HealthAgents project Web site. Retrieved May 15, 2011, from http://www.healthagents.net.

Hendler, J. (2001). Agents and the Semantic Web. IEEE Intelligent Systems, 16(2), 30-37.

Horling, B., & Lesser, V. (2005). A Survey of Multi-Agent Organizational Paradigms.

Knowledge Engineering Review, 19(4), 281-316.

Jadad, A.R., Goel, V., Rizo, C., Hohenadel, J., & Cortinois, A. (2000). The Global e-Health Innovation Network - Building a Vehicle for the Transformation of the Health System in the Information Age. Business Briefing: Next Generation Healthcare, (pp. 48-54).

JADE (2008). Jade software Web site. Retrieved May 15, 2011, from http://jade.tilab.com.

Jennings, N. R. (2000). On Agent-Based Software Engineering. Artificial Intelligence, 117, 277- 296.

Jennings, N. R., Corera, J. M., & Laresgoiti, I. (1995). Developing Industrial Multiagent Systems. In Proc. of the First International Conference on Multiagent Systems, (pp. 423-430).

Menlo Park, CA: AAAI Press.

Jennings, N. R., Faratin, P., Lomuscio, A. R., Parsons, S., Sierra, C., & Wooldridge, M. (2001).

Automated negotiation: prospects, methods and challenges. Group Decision and Negotiation, 10(2), 199 -215.

Jih, W., Hsu, J.Y., & Tsai, T. (2006) Context-aware service integration for elderly care in a smart environment. In D.B. Leake, T.R. Roth-Berghofer, & S. Schulz, (Eds.), In 2006 AAAI Workshop on Modeling and Retrieval of Context Retrieval of Context, (pp. 44-48). Menlo Park, CA: AAAI Press.

K4CARE (2007) K4CARE project Web site. Retrieved May 15, 2011, from http://www.k4care.net.

Laleci, G.B., Dogac, A., Olduz, M., Tasyurt, I., Yuksel, M., & Okcan, A. (2008). SAPHIRE: A Multi-Agent System for Remote Healthcare Monitoring through Computerized Clinical

(18)

Guidelines. In R. Annicchiarico, U. Cortés, C. Urdiales, (Eds.) Agent Technology and e-Health.

Whitestein Series in Software Agent Technologies and Autonomic Computing, (pp. 25-44), Babel, Switzerland: Birkhäuser Verlag.

Laxminarayan, S., & Stamm, B.H. (2002). Technology, Telemedicine and Telehealth. In Business Briefing: Global Healthcare, Vol. 3, (pp. 93-96) London, UK: World Medical Association.

Liffick, B.W. (2003). Assistive technology in computer science. In Proc. of the 1st international Symposium on information and Communication Technologies, (pp. 46-51), Dublin, Ireland.

Mabry, S.L., Hug, C.R., & Roundy, R.C., (2004). Clinical decision support with IM-agents and ERMA multi-agents. In Proc. of the 17th IEEE Symposium on Computer-Based Medical Systems (CBMS 2004), (pp. 242-247). Bethesda, MD.

Mann W.C. (2005). Smart Technology for Aging, Disability, and Independence: The State of the Science, Hoboken, NJ: John Wiley & Sons.

Meystre, S. (2005). The Current State of Telemonitoring: A Comment on the Literature, Telemedicine and E-Health, 11(1), 63-69.

MobiHealth (2004) MobiHealth project Web site. Retrieved May 15, 2011, from http://www.mobihealth.org/.

Moreno, A., & Nealon, J. (Eds.). (2003). Applications of Software Agents Technology in the Health Care Domain. Whitestein Series in Software Agent Technology, Babel, Switzerland:

Birkhäuser Verlag.

Moreno, A., Valls, A., Isern, D., Sanchez, D. (2006). Applying Agent Technology to Healthcare:

The GruSMA Experience. In Intelligent Systems, IEEE , vol.21, no.6, pp.63-67, Nov.-Dec. 2006.

Muller, J.P. (1998). Architectures and applications of intelligent agents: A survey. Knowledge Engineering Review, 13(4), 353-380.

MyHeart (2008). MyHeart project Web site. Retrieved May 15, 2011, from http://www.hitech- projects.com/euprojects/myheart.

Mynatt, E.D., Essa, I., & Rogers, W. (2000). Increasing the opportunities for aging in place. In Proc. of the 2000 Conference on Universal Usability, (pp. 65-71). Arlington, VA.

Negri A., Poggi A., Tomaiuolo M., & Turci P. (2006). Agents for e-Business Applications, In Proc. of the 5th International Joint Conference on Autonomous Agents and Multi Agent Systems, (pp. 907-914). Hakodate, Japan.

Newell, A. (1982). The Knowledge Level. Artificial Intelligence, 18, 87-127.

Nguyen X. T. (2005). Demonstration of WS2JADE. In Proc. of the 4th International Joint Conference on Autonomous Agents and Multi Agent Systems, (pp. 135-136). Utrecht, The Netherlands.

Ohinmaa, A., Hailey, D., & Roine, R. (1999). The Assessment of Telemedicine: General principles and a systematic review. INAHTA Joint Project. Finnish Office for Health Care Technology Assessment and Alberta Heritage Foundation for Medical Research. Retrieved May 15, 2011, from http://www.inahta.org/upload/Joint/Telemedicine_1999.pdf.

(19)

Rialle, V., Lamy, J., Noury, N., & Bajolle, L. (2003). Telemonitoring of patients at home: a software agent approach, Computer Methods and Programs in Biomedicine, 72(3), 257-268.

SAPHIRE (2008). SAPHIRE project Web site. Retrieved May 15, 2011, from http://www.srdc.metu.edu.tr/webpage/projects/saphire/.

Smith, R., & Davis, R. (1980) The contract Net protocol: High level communication and control in a distributed problem solver. IEEE Transactions on Computers, 29(12), 1104-1113.

Song, X., Hwong, B., Matos, G., Rudorfer, A., Nelson, C., Han, M., & Girenkov, A. (2006).

Understanding requirements for computer-aided healthcare workflows: experiences and challenges. In Proc. of the 28th international Conference on Software Engineering, (pp. 930- 934). Shanghai, China: ACM Press.

Tonino, H., Bos, A., de Weerdt, M., & Witteveen, C. (2002). Plan coordination by revision in collective agent based systems. Artificial Intelligence, 142, 121–145.

U-R-SAFE (2005). U-R-SAFE project Web site. Retrieved May 15, 2011, from http://ursafe.tesa.prd.fr.

Vittikh, V., Gritsenko, E., Surnin, O., Skobelev, P., Volhoncev, D., Karavaev, M., Shamashov, M., & Tsarev, A. (2007). Multi-agent system of Samara region social services based on social passports and smart cards of citizens. In Proceedings of the 6th international joint conference on Autonomous agents and multiagent systems (AAMAS '07). ACM, New York, NY, USA, Article 277, 7 pages.

Wilson, P., Leitner, C., & Moussalli, A. (2004). Mapping the Potential of eHealth: Empowering the Citizen through eHealth Tools and Services, In Proc of E-Health Conference 2004, Cork, Ireland. Retrieved May 15, 2011, from http://aei.pitt.edu/6092/01/2004_E_01.pdf.

Wooldridge, M. J., & Jennings, N. R. (1995): Intelligent Agents: Theory and Practice.

Knowledge Engineering Review, 10(2), 115-152.

TERMS AND DEFINITIONS

Coordination: Coordination is a process in which a group of agents engages in order to ensure that each of them acts in a coherent manner.

Contracting: A process where agents can assume the role of manager and contractor and where managers try to assign tasks to the most appropriate contractors.

Multi-agent planning: A process that can involve agents plan for a common goal, agents coordinating the plan of others, or agents refining their own plans while negotiating over tasks or resources.

Multi-agent system: A multi-agent system (MAS) is a loosely coupled network of software agents that interact to solve problems that are beyond the individual capacities or knowledge of each software agent.

Negotiation: A process by which a group of agents come to a mutually acceptable agreement on some matter.

(20)

Organizational structuring: A process for defining the organizational structure of a multi-agent system, i.e., the information, communication, and control relationships among the agents of the system.

Software agent: A software agent is a computer program that is situated in some environment and capable of autonomous action in order to meet its design objectives.

Riferimenti

Documenti correlati

In a related fashion, fast revisions of the decision to acquire information in our model lead to episodic adoption of the signal, volatility bursts and reduced use of

Gladney, Muslim Chinese: Ethnic Nationalism in the People’s Republic, Cambridge, MA e London, Harvard University Asia Center, (nuova ed.), 1996.. Bays, A New History

1. Come spesso accade nei periodi storici contrassegnati da gravi crisi sistemiche, il fuoco del nazionalismo e del populismo ha ripreso a divampare, tanto in

Il paese degli Zengi o la costa orientale dell'Africa nel Medioevo (geografia, costumi, prodotti, animali... Graduate School. Dottorato di ricerca in Lingue e Civiltà dell'Asia

High-Resolution X-ray Imaging Studies of Neutron Stars, Pulsar Wind Nebulae and Supernova

Infatti, alcuni studi hanno indicato che livelli moderati di carico sui tessuti svolgono un ruolo di protezione da lombalgie; di contro livelli estremamente bassi

0 Undertake a capacity assessment to create a community-based workforce based on local need (e.g. volumes of health and social care workers, the availability of an

} S aspetterà finché R non manda un messaggio di risposta; ad esempio un messaggio con content: open(left_door), se questa si trova nella belief di R; se non c’è l’agente