Loneliness and quality of life

Innovation to assess the mental health of the elderly.

The project Evaluation of loneliness and quality of life of the elderly through the use of information and communication technologies is funded by the Generalitat Valenciana. This project aims to study the mental health of the elderly through new information and communication technologies. One of the objectives is to develop simple, automated and less invasive ways of assessing the mental health of the elderly. Another objective of the project is to observe whether there are patterns in the behaviour and language of older people that can be considered indicators of various aspects of mental health, such as loneliness, well-being or quality of life.

A longitudinal study will be carried out with the voluntary participation of a group of elderly people. The participants will be monitored through the information provided by a smartwatch. The data collected will provide information on the person's position and physical activity, both indoors and outdoors. In addition, participants will be able to interact with a conversational assistant, which will run on their cell phones. Through the assistant, they will answer four standardized mental health questionnaires that provide the levels of loneliness, life satisfaction, quality of life and well-being that the person has at the time they respond. In addition to answering questionnaires, they will be able to express their current mood using their own words and simulate a human conversation with the robot. This will result in a database with open texts where elderly people can talk and express openly how they feel.

After the experimental phase, there will be a database with texts and mental health scores, collected by the conversational assistant, and position and physical activity data collected by the smartwatch. On the one hand, the effectiveness of these technological tools for the collection of health information in elderly people will be evaluated. On the other hand, a large database will be obtained with health information of the people participating in the study. Statistical process control techniques and machine learning algorithms will be applied to this data. This will make it possible to analyze and infer the patterns associated with changes in the levels of loneliness, life satisfaction, well-being or quality of life of the participants.

Based on collected texts, sentiment polarity analysis will be applied. The correlation between the indices provided by the questionnaires and the analysis of the language of the texts will be studied. The relationship between the objective variables of positioning and physical activity with the measures obtained in the mental health questionnaires will also be studied. With all this, it is expected to be able to advance in the detection of those elderly people who suffer from some mental health problem.