La population des personnes agees atteintes de la maladie d'Alzheimer augmente depuis quelques annees et augmentera encore pendant plusieurs annees. Ces personnes, porteuses de troubles comportementaux sont de plus en plus agees et dependantes. Elles posent des problemes specifiques de prise en charge pour les professionnels soignants au niveau de la securite, de la responsabilite et de la charge de travail, et necessitent en particulier, de la part de leur entourage familial ou professionnel, une vigilance constante qui pent etre difficile a assurer. Nous presentons un systeme multicapteurs qui surveille, en continu, la personne dependante dans sa chambre d'hopital pendant la nuit. Durant la periode d'evaluation, ce systeme contnue par un reseau de capteurs a infrarouge passif, enregistre tous les mouvements nocturnes de la personne agee. En fin de surveillance, son logiciel d'intelligence artificielle traite les donnees et affiche les resultats soit sous forme d'histogramme en trois dimensions, representant I'activite nocturne du patient, soit sous forme d'une liste de deplacements. A la fin de ('evaluation, les traitements de donnees auront lieu en temps reel et le systeme alertera le personnel soignant en cas de chute ou de fugue du patient.
Over the past few years, the number of elderly people with Alzheimer's disease has progressively increased, and will continue to do so for several years to come. Persons with this disease, which is characterized by behavioral disorders, are becoming increasingly older and more dependent. As a result of this, there are a number of problems connected with their stay in a nursing home or special care unit, such as the additional burden the nursing staff or caregivers, management of the security factor, responsibility and workload. Their condition necessitates constant surveillance on the part of family members or professional staff. In this article, a multisensor system has been described, which provides continuous monitoring during the night of the elderly person in his/her hospital room. Over the period in question, the passive infrared sensor system records every movement of the elderly person. At the end of the surveillance period, the artificial intelligence program processes the stored data and then displays the results either in the form of a three-dimensional histogram, or as a list with all movements (apart from those in bed involving less than three movements) noted together with the time, duration and contextual comment. This system alerts the caregivers by sounding an alarm if a person falls or tries to run away. It is therefore useful at several levels: in a hospital environment, in a nursing home, in a private home, and as a research tool.