Improving quality of life and reducing health risks are increasingly important concerns in our society. In different situations, including the presence of chronic diseases, or the desire to adopt healthier daily habits (e.g. concerning diet, physical activity, etc.), these improvements are only possible if an effective behaviourchange is also produced. This change can span from small alterations to daily routine to radical changes in lifestyle. It has been shown that personalized interventions are crucialin order to maximize the efficacy of behaviourchange, and therefore the ultimate goals set for a given program. Custom and tailored programs are nowadays feasible, in part thanks to advances in personal data analytics and personalized digital health. Different models exist to describe behaviourchange strategies and actions (e.g. HAPA model (Schwarzer et al., 2011) and the I-Change Model (De Vries et al., 2005)), and different technological solutions (e-Health, mHealth, serious games, reminders, chatbots,social networks) have been developed in several use-cases (diabetes, smoke cessation, overweight, active-ageing, rehabilitation, re-adaptation, etc.).
However, the effort required in order to adapt these models to the appropriate technologies in a given use-case, remains prohibitive and result in ineffective or partial implementations, with little or incomplete personalization interventions. As a consequence, there is no clear methodology that allows to effectively model the profile of a patient, with the goal of using artificial intelligence (AI) techniques to adapt and personalize treatments, recommendations, and other health-related interventions. Therefore, even if different digital solutions and AItechniques have been shown to provide significant improvement to personalized treatments, it remains challenging to reuse and apply these methods to other use cases, or to establish a well-defined workflow for enabling patient-tailored interventions.
With this project, we envision a methodological approach that establishes in a systematic way thedifferent steps that can guide the implementation of a personalized e-health program, adapting existing AI techniques and reusing software building blocks that provide profiling and targeted interventions. The need for the development of this methodological approach is actually coming from our experience in health monitoring, and the involvement and discussions with several companies, healthcare providers and researchers in the area of eHealth. As a main result, the PROFILES project will allow us to setup new research grant proposals, both at national and international levels.
The main goal of the project is to study and develop a methodology for modelling patient profiles, so that they can be fed to AI techniques in order to offer personalized interventions targeting behavioralchange. We argue that this methodology could be reused in several scenarios, even across different diseases or health problems, and different types of interventions. The methodology should be centeredon a user profile, and on how to define the features and parameters that are essential for feeding AI techniques. It should also help in choosing the appropriate AI techniques (machine learning methods, classification algorithms, etc.) and the type of interventions, depending on the nature of the use case, and the requirements that have been established. Finally, the methodology should also allow to define metrics that will be used in order to evaluate the overall procedure.