The FP7 project Help4Mood aims to develop a platform to support the treatment of patients with depression. Major unipolar depression is a significant cause of disability worldwide. Depression affects mood, cognition and behaviour and its core symptoms are persistent low mood and levels of activity, and loss of interest.

The platform that Help4Mood is developing a personal health system administered by a clinician that supports the treatment of people with depression, in their homes. The platform is structured around patient sessions and focuses on monitoring symptoms and engaging the patient in activities based on cognitive behavioural therapy.

This platform will significantly advance the state-of-the-art in computerised support for people with major depression by monitoring mood, thoughts, physical activity and voice characteristics, prompting adherence and promoting behaviours in response to monitored inputs. These advances are delivered through a Virtual agent which can interact with the patient through a combination of enriched prompts, appropriately contextualised spoken output, body movements and facial expressions. Monitoring combines existing (movement sensor, psychological ratings) and novel (voice analysis) technologies as inputs to a pattern recognition-based decision-support system for treatment management.

Description of the way to implement the initiative

The aim of Help4Mood is to develop a system that will help people with major depression recover in their own homes.   The project is researching an interactive system to support the treatment of people who are significantly affected but who are still able to live at home and may also be working. The Help4Mood system is designed to be used together with other forms of therapy, such as self-help, counselling and/or medication.

The system has three main components:

• a Personal Monitoring System that keeps track of important aspects of behaviour such as sleep or activity levels;
• an interactive Virtual agent which asks patients about their health and well-being, provides a portal to trusted health information, and feeds back information collected through monitoring and questionnaires to a clinician;
• a Decision Support System that tailors each session with the virtual agent to the individual needs of the person with depression and that supports clinicians in interpreting the data collected through the virtual agent and the personal monitoring system.

Subjective, self-reported monitoring is a cornerstone of Help4Mood. Patients are able to track their moods, feelings, and experiences, and the system feeds back relevant trends and developments in the form of a weekly report.  Patients are offered varying lengths of interaction at the start of each daily session, but are encouraged to perform the full range of activities every week.

Through daily interactions with the Virtual agent, users track how they are feeling and work through brief, simple exercises based on principles of cognitive behavioural therapy. The Virtual agent is an animated talking head which was designed in response to user surveys.  Virtual agents, when implemented well, have been shown to increase user engagement and help users form an alliance with the system.

Data is entered using a simple graphical user interface. The Virtual agent is customisable, with distinct appearances (in English both male and female) to choose from. The main role of the patient-side Decision Support System is to facilitate intelligent, adaptive collection of self-reported data. This is crucial for the successful long-term deployment of Help4Mood.

The project assesses the effect of depression on overall activity through a Personal Monitoring System that collects activity and sleep data; the effect of depression is also being assessed through simple speech tasks offered during the session with the Virtual agent, researching the connection between speech patterns and self-reported mood.  Special care is being taken to ensure that the patient user interface is as easy and pleasant to use as possible. The Personal Monitoring System has also been designed to be as non-obtrusive as possible, and takes the form of actigraphs worn as wristwatches, keyrings, or belt devices, and custom-built sleep monitors.

Objective monitoring focus on two main aspects, activity patterns, and sleep. In a systematic review of the literature, it was found that there are clear differences between the activity levels of people with depression pre- and post-treatment, and between people with and without depression. While no such clear differences have been observed for time spent asleep, sleep is clearly an aspect that matters a lot to patients. The Help4Mood project focuses not only on sleep duration, but also on sleep quality.

The Decision Support component of Help4Mood plans sessions, making sure they are varied and interesting, and distils the data collected through the virtual agent and the personal monitoring system into a one-page overview. This overview combines text and graphs to help clinicians assess the patient’s current mental and emotional state and guide further treatment decisions. Users and their clinicians can then use this overview to discuss progress and plan further treatment.

General Architecture of the Help4Mood System1

The core Help4Mood system will be implemented on a laptop to be handed to the patients, together with a few custom sensors for measuring sleep and activity patterns. Clinician communication and clinician-side implementation will be adapted to the IT infrastructure at each Help4Mood site.

Since Help4Mood is designed to be a flexible, configurable system and since it is not linked to a specific treatment package, it can be adapted to fit many different clinical settings and pathways across Europe. The project will develop easy-to-use configuration interfaces for site-specific, treatment-specific, and user-specific adaptation. The documentation will also include clinical guidelines on incorporating Help4Mood into existing approaches to treatment.

1 - The mobile phone option shown in the graphic as an activity monitor was discarded fairly early in the project, mainly due to problems with battery life.  Ideally the actigraphs will run for a full 4-week period; this is not possible with phones without regular recharging, which increases the intrusiveness of the system.

Main results, benefits and impacts

The main aim of the Help4Mood project is to provide a closed loop approach supporting the control, communication and treatment management of patients with Major Depression. This approach will be a distributed system with three main components (the Personal Monitoring system, the Virtual agent component and the DSS for Treatment Management) deployed in both places: at the patient’s home and at the clinician’s location.

Return on investment

Lessons learnt

To be added at a later date.