The ISIS research team focuses on three themes:

• Assisted living for dependent patients

• New e-health technologies

• The processing and management of medical data

The "Assisted living" Theme:

Faced with the increase in chronic degenerative diseases that requires monitoring and long-term patient management, and the growing desire of patients to be treated in a family environment in order to protect their social ties, and, finally, faced with a need to reduce costs, currently, in France, like other developed countries, we are seeing new strategic orientations in the structuring of the healthcare system's offer. In particular, this offer takes the form of a transfer of a large portion of care activities from the hospital to the patient’s residence. The hospital is therefore turning into a step in the treatment process and “proximity” structures (home hospitalization, home care, substitution of nursing care at home, etc.) are playing a major role in this global process. This issue relates to individuals who are losing their autonomy (aging, ill and disabled persons) who, in order to continue living alone, must be monitored and assisted in their daily environment. Emergency situations must be identified for the rapid attendance of outside support.
It is within this logic and according to the national (ANR, etc.) and European (“Ambient Assisted Living”) research program orientations related to healthcare, where information and communication technologies (ICT) are considered to be a crucial element to improve practices in this field, that this research project is positioned.

Healthcare professionals agree on the economic and human benefits of patients remaining in their homes. Faced with the aging of the population, it is desirable to allow aging individuals who are dependent to remain in their living environment rather than in specialized institutions. Certain treatments that are usually long and require hospitalization could also be applied at home at low cost and with greater patient satisfaction if a suitable infrastructure is put in place. From a technological perspective, this field requires a convergence between home automation, telecommunication, and monitoring and automatic decision-making techniques to identify situations that may be dangerous to the patient and trigger suitable alarms to the patient’s environment (family and friends, healthcare services, etc.). In this field, the ISIS team is studying systems that combine sensors, data transmission items and related information processing elements to continuously monitor the physiological features of an individual. In particular, the team uses distributed artificial intelligence techniques and multi-agent systems, and is developing projects in cooperation with corporations in the sector to test this type of solution in actual situations.
Meta-surveillance: co-construction and sharing of macro-information
We propose to study assisted home living of a dependent person by going beyond surveillance technologies and data processing methods.
Indeed, it appears that current systems dedicated to assisted home living remain focused on one patient or a juxtaposition of patients. None of the current projects aims at the collectivization of the surveillance of a large number of individuals. The heterogeneity of applications, the partitioning and significant distribution of surveillance data, as well as aspects related to privacy make this task arduous. However, through aggregation and generalization processes, this would allow for the emergence of categories of individuals in order to create adaptable supervision or contribute to the forecasting and measuring of demographic evolutions.

Our work on massively multi-agent systems takes on its full meaning here: for the past decade, we have focused on large-scale multi-agent systems, in particular with the abrupt injection of new agents, the dynamic reorganization of a system and data aggregation methods.
The S(MAD)2D system currently being developed is a system that is capable of creating a classification of patterns of individuals to be supervised based on individual surveillance data. The patterns thereby created and managed can be accessed by the MAD teams and play a number of roles, such as:
•    Offering customized assistance to the person according to the relevant pattern; for example, greater surveillance of a specific behaviour or planning for preventive care.
•    Identifying correlations with tools to evaluate dependence and, for a given individual, the changes in the degree of dependence that require a re-evaluation of the support provided.
•    Obtaining global statistical data on individuals cared for at home.

• The “e-Health and New Technologies” Theme:

The ISIS team is also studying the impact of new information and communication technologies on various players involved in healthcare. New technologies appear every day both in the field of telecommunications and new interaction techniques offered to users: SmartPhones, PDA, UMPC, remote localization, RFID... These technological advances may be put to good use to change practices in the medical sector and lead to new applications. One example of this type of convergence is the Columba bracelet, developed by France Telecom’s research center and targeting Alzheimer patients; it includes a GPS and cell phone and may trigger telephone alarms when the patient goes beyond a predefined geographic area. In this field, the ISIS team uses design methods focused on users resulting from the “Man-Machine Interaction” community to offer new ergonomic applications suited to user tasks and to evaluate the effectiveness and usability of these applications.

The “Medical Data” Theme:

Healthcare sector activities generate data of highly varied types: exam results, diagnoses, treatments, vaccination history, epidemiology, imagery … the computerization, storage and exchange of this data with appropriate security, confidentiality and privacy conditions are major challenges, first to improve efficiency and reduce system costs, but mostly to provide better service to patients and minimize therapeutic incidents of any nature. Recent misadventures with the “shared medical file” (DMP) in France perfectly illustrate the complexity of the implementation of these types of systems. However, in light of the potential benefits, their development is crucial in one form or another. The role of research in information systems, in this field, involves removing all the technological obstacles to development. However, this type of research must not be approached from a techno-centric perspective, but must be directed by the needs of players in the healthcare sector, taking into account research in the humanities and social sciences, as well as the study of practices and customs.

From a technological perspective, the ISIS team is approaching this research issue by using recent developments on the medical semantic web: the semantic web means a set of technologies that aims at making the content of the World Wide Web accessible and usable by software programs and agents thanks to a formal metadata system, in particular using the family of languages developed by the World Wide Web Consortium (W3C), the body responsible for the standardization of Internet technologies. It is noteworthy that the first working group the W3C has put in place in this field is specifically dedicated to the healthcare sector (W3C Semantic Web Health Care and Life Sciences Interest Group), with the objective of removing obstacles to the sharing of related data, which slow down clinical research, and using semantic web technologies to facilitate communication between chemists, biologists, clinicians and other researchers.

The ISIS team is also involved in defining ontologies for medical data, which is part of the international standardization issue. This standardization stage opens the way for the determination of medical web services, interoperable software components that will allow players in the healthcare sector to develop information systems that exchange this data, with all the necessary security and confidentiality guarantees. This is referred to as medical middleware, adapted to the needs of all players (doctors, pharmacists, hospitals, social services, etc.).