Jesus Sanchez1, Victor Sanchez1, Ioan Salomie2, Adel Taweel3, Jim Charvill4 and Manuel Araujo5
Abstract
Studies have shown that the older adults consider autonomy as an essential part of their quality of life, and to stay at home for as long as possible is one of their main wishes. This calls for the need of providing ICT based tools for helping them to cope with their daily life activities such as self-feeding. DIET4Elders project aims to assist the elderly in their nutrition daily activities and helping them in maintain healthy feeding habits. To achieve this goal the project will use new types of sen-sors, reasoning techniques for assessing the feeding habits and will enable the dynamic selection, based on the prescribed diet, of suitable food service provider and potentially facilitate automated shopping.
1. Introduction
The DIET4Elders (Dynamic nutrItion bEhaviour awareness sysTem FOR the Elders) project6 ad-dresses the problem of older adults malnutrition caused by their unhealthy self-feeding habits.
Studies (Sieber, 2010) have shown that in Europe more than 15% of the older population is af-fected by poor nutrition and malnutrition caused by the problems of ageing such as decrease in sensitivity, poor dental health, lack of transportation, physical difficulty, forgetfulness and other issues. The target groups of our project are represented by: (i) older adults living in care homes for which recent statistics say that are affected by malnutrition in more than 60% of cases and (ii) older adults living at home for which statistics says that more that 15% are affected by malnutri-tion. In this context the DIET4Elders project aims to develop an innovative ICT-based system to provide support services to: (i) assist older adults and their informal carers during daily self-feeding activities and is aimed at detecting and preventing the instauration of malnutrition, (ii) help the
[1] Ingeniería y Soluciones Informáticas S.L., Spain, vsanchez@isoin.es
[2] Technical University of Cluj-Napoca, Computer Science Department, ioan.salomie@cs.utcluj.ro [3] King’s College London, UK, adel.taweel@kcl.ac.uk
[4] Tunstall Healthcare Ltd, UK, jim.charvill@tunstall.com [5] COESCO Deza S.L., Spain, maraujo@cocinaculinaria.com [6] www.diet4elders.eu
SESSION A3 123
nutritionists to establish the degree in which the older adults follow their prescribed diet and to dynamically adjust it and (iii) enable the dynamic selection, based on the prescribed diet, of suit-able food service provider and potentially ensuit-able automated shopping.
More specifically, the DIET4Elders system collects data regarding the older adult’s self-feeding activities and the context in which these activities took place followed by data analysis with the goal of assessing the older adults’ diet and of identifying behavioural patterns that are usually as-sociated with unhealthy eating that may lead to malnutrition. The resulting knowledge, after data analysis, will be made available for end-users by especially developed support services to provide user-custom experience and user-adaptable assisting functionality. This knowledge is also used to select and combine the appropriate food ordering services that will provide the right combination of food that will match both the users’ needs and wishes and their recommended diet. To address the project ambitious goals the consortium brings together partners with specific expertise in the following domains: information and computer technology, tele-care, nutrition and clinical informed diet for older adults.
2. DIET4Elders System Architecture
Figure 1 presents an overall view of the DIET4Elders system architectural layers: (i) Monitoring Layer which aims at collecting raw data about older adults’ daily life activities along with the context in which these activities took place, (ii) Analysis and Assessment Layer which aims at determining information and inferring knowledge about the older adult diet and self-feeding behaviour and (iii) Support Services Layer which aims at using the obtained information and knowledge to assist the older adult, informal carer or nutritionist in improving the older adult self-feeding habits and the quality of life. DIET4Elders system is designed by combining techniques and technologies of con-text aware computing, bio-inspired computing, artificial intelligence and knowledge engineering.
The heart of the system is a clinically and nutritionally-based knowledge (ontology) linked to older adults preferences that continually suggest suitable diet alternatives and plans.
Figure 1DIET4Elders system layered architecture
124 SESSION A3
The Monitoring Layer includes the techniques and tools developed for data monitoring and pro-grammatic representation. All data regarding the older adults’ daily life activities and the context in which these activities took place are collected and stored in a database (Older Adults Daily Life Data) at fixed time stamps using a wireless sensor network based monitoring infrastructure composed of, smart containers, smart fridges, accelerometers, health monitoring devices, position sensors, RFID, smart phone, etc. As it can be noticed the data is collected using several physical or virtual sources and is usually represented in heterogeneous formats. Also, there is an evident need for providing an integrated, uniform and programmatic model for data representation so that it can be processed, evaluated, shared and understood by all the DIET4Elders system’s compo-nents defined at each layer. We address these issues by designing and developing the Older Adults Daily Life Context Model (OADLC model) to uniformly and semantically represent all collected data by means of ontologies. Ontology is chosen because it allows sharing the same semantic interpre-tation or meaning of the data and inferring new information or knowledge by means of reasoning processes. To represent the current older adult’s situation, the OADLC model is instantiated using the data collected at the current time stamp regarding the older adults’ daily life activities and the context in which these activities take place.
The Analysis and Assessment Layer aims at: (i) identifying the daily life activity that the older adult is currently carrying out, (ii) constructing the older adult activity diagram representing his/her self-feeding behaviour and (iii) identifying the unhealthy self-self-feeding behaviours. On-line reasoning based techniques are used with the goals of identifying the older adult’s current daily life activity and eventually identifying critical situations (e.g. food chocking, lack of food in the house, too much drugs are taken all at once, etc.) in which urgent alarming and intervention of informal car-ers is required. The identified older adult’s self-feeding activities are chained in activities diagrams describing the overall older adult’s self-feeding behaviour. Prediction techniques are employed on the constructed self-feeding activities diagram and used to proactively detect and identify un-healthy behavioural patterns (as defined by nutritionists).
The Support Services Layer will provide services to (see Figure 2): (i) assist older adults and their informal carers during daily self-feeding activities aiming at detecting and preventing the instaura-tion of malnutriinstaura-tion, (ii) help the nutriinstaura-tionists to establish the degree in which the older adults follow their prescribed diet and to dynamically adjust it and (iii) enable the dynamic selection, based on the prescribed diet, of suitable food service provider and potentially enable automated shopping.
The Diet Aware Food Ordering Service will aggregate the food delivery services of different provid-ers and will guide the older adults and their informal carprovid-ers to order the right food that complies to the nutritionists recommended diet. For this to work effectively, the system will create a virtual electronic marketplace which will define food services in a diet and nutritionally rich context and where food providers can register their services. All the food services registered and available in the market will comply with a service description specification based on semantic annotation.
The Unhealthy Behaviour Notification Service will send notifications to the older adults and their informal carers anytime an unhealthy behaviour is identified by the Analysis and Assessment Layer. The notification will provide information regarding the identified unhealthy behaviour and assistance for the older adult or informal carer on how the behaviour can be corrected.
SESSION A3 125
The Older Adults Remote Diet Monitoring Service provides access for the nutritionists to all data/
information/knowledge regarding the older adult self-feeding behaviour acquired/inferred/learned by the DIET4Elders system. Using this information the nutritionists will be able to assess the de-gree in which the older adults’ follow their prescribed diet. The diet adjustment service allows the nutritionist to dynamically and seamlessly adjust the older adult’s diet by accessing and updating the content a Dietary Knowledge Base.
Figure 2. Support Services Layer details
3. Conclusions
This paper presents the innovative ICT-based system for providing support services to assist older adults and their informal carers during daily self-feeding activities. The use of the presented sys-tem will improve the older adult’s quality of everyday life and will reduce the malnutrition and even-tually their negative effects such as: exacerbation of chronic and acute diseases, speeding up the development of degenerative diseases, delaying recovering from illness and so forth. The project success will subsequently help reducing homecare, healthcare, and other associated costs, which in Europe are currently estimated around 170 billion euro per year.
References
Sieber, C. C. (2010) Malnutrition and appropriate nutritional care, Nutrition Day Conference, www.european-nutri-tion.org
126 SESSION A3