A DSL-Based Approach for Detecting Activities of Daily Living by Means of the AGGIR Variables

Abstract

In this paper, we propose a framework for studying the AGGIR (Autonomie Gérontologique et Groupe Iso Ressources—Autonomy Gerontology Iso-Resources Groups) grid model, with the aim of assessing the level of independence of elderly people in accordance with their capabilities of performing daily activities as well as interacting with their environments. In order to model the Activities of Daily Living (ADL), we extend a previously proposed Domain Specific Language (DSL), by defining new operators to deal with constraints related to time and location of activities and event recognition. The proposed framework aims at providing an analysis tool regarding the performance of elderly/disabled people within a home environment by means of data recovered from sensors using a smart-home simulator environment. We perform an evaluation of our framework in several scenarios, considering five of the AGGIR variables (i.e., feeding, dressing, toileting, elimination, and transfers) as well as health-care devices for tracking the occurrence of elderly activities. The results demonstrate the accuracy of the proposed framework for managing the tracked records correctly and, thus, generate the appropriate event information related to the ADL.

Study protocol for a pragmatic cluster randomized controlled trial to improve dietary diversity and physical fitness among older people who live at home (the “ALAPAGE study”)

“…The inclusion criteria are: age ≥ 60 years, who live at home, have health insurance coverage and are able to read and write in French. The non-inclusion criteria are: receiving allocation for dependence (i.e., belonging to one of the fourth-highest groups of the AGGIR grid, a six-level scale used in France to measure the independency levels of elderly people [ 44 , 45 ]), being under guardianship, and having participated in a previous prevention workshop on diet or physical activity within the last 2 years. These criteria aim to target individuals: who, according to the operational partners’ usual practices within the field of healthy aging, are suitable for the preventive intervention program; who meet the research requirements (e.g., are able to give free and informed consent and to complete self-administered questionnaires) and who meet the regulatory requirements (health insurance).…”

A task-model based approach for detecting ADL-related anomalies

“…For this matter, Body Sensor Networks (BSN) as well as Personal Sensor Network (PSN) are involved aiming to define ADL within the SHE. Modeling and detection of complex events methods are mainly supported by languages to model events relying on a set of operators [14,15]. For example, the Allen's temporal logical operators [16], such as FOLLOWS (also called as SEQ or AFTER), to indicate that two events occur after each other, and OVERLAPS, to describes two events occurring simultaneously, are prevalent in every language to represent complex events.…”

“…A comparison between the original task model introduced in Section III and the events occurred within a SHE following the task model requirements is necessary in order to assess the above-mentioned approach. To this end, a SHE simulation over a one-week period is evaluated by means of the framework introduced in [14,15]; consisting on three modules:…”