Led by researcher François Brémond, Stars is a founding member of the university host team CoBTek (Cognition Behaviour Technology), led by psychiatrist Professor Philippe Robert, who is also the coordinator of the Centre Mémoire de Ressources et de Recherche (CMRR) at Nice University Hospital. To date, around one hundred experiments have been conducted at the CMRR with elderly people who are healthy or suffering from dementia.
As the population ages, the number of dependent people is going to increase considerably in the years to come. Factor in the people suffering from Alzheimer's disease and other neurodegenerative illnesses, and we have a real social problem.
The Stars and CoBTek teams will help us to understand better how Alzheimer's affects the daily life of its sufferers and will enable us to identify, and above all quantify in an objective way, its impact on the behaviour of these people. The researchers have developed a new software platform that not only enables analysis of video data, but will also eventually pair these video data with audio data and recordings made by other types of sensors (physiological, for example), such as actimeters.
The Behaviour Analysis system implemented by the Stars team shows the analysis of the universal movements, the most basic of which are changes in posture (such as standing up or sitting down), walking and falling. During the experiments, the patient is placed in a room that resembles a normal room in a house. He or she is observed while carrying out certain free tasks, such as preparing a hot drink or a meal, or during a ‘rest’ activity, such as reading a book or watching the television. These activities are filmed live and the patients’ behaviour analysed by specially designed software.
‘Among other things, we are studying the loss of certain functionalities in a patient ,’ explains François Brémond. ‘For example, whether or not he or she is capable of organising and executing a specific task, as well as long periods of inactivity (of the kind that may follow a fall, for instance). We are moving past the limits of current activity recognition systems, as we can objectively measure a whole range of movements from daily life .
The activity analysis programs are either written on-site or found in open-source form. Stars-CoBTek also works with a company that is developing specialist algorithms for detecting when someone falls (Link Care Services ) and with robotics experts. We should also note that the teams are taking part in the Az@Game project, selected as part of the French government’s Future Investments programme for the development of the digital economy, which aims to develop interactive games (Serious Games) for the use of Alzheimer’s sufferers, as well as professional and family carers.
Eventually, the platform will improve patient care by providing additional information to doctors and other healthcare professionals. It could also help and reassure family members, who often lack the means to deal with the illness. One of the objectives of the project is to enable dependent people to continue living in their own homes. The reasons for this are socio-economic – it is less expensive for the State and, what is more, patients prefer to stay in their own environment rather than having to use a care home or a similar establishment.
Questions for Professeur Philippe Robert
What made you decide to work with Stars?
CobTek is a university host team that is developing new information and communication technologies in several health-related fields. Our research focuses, among other things, on Alzheimer's disease and on ways of preventing dependence and loss of autonomy. Our partnership with Stars allows us to mix theoretical and clinical research - a first in this field.
What specific contribution does the Stars research make to the treatment of Alzheimer’s and other cognitive disorders?
It assesses behaviour, autonomy and cognitive processes such as memory and language. Assessment is now much more objective, compared to the days when we had to interview the patient’s spouse or friends and family to find out about his or her day-to-day behaviour.