fbpx New research aims to search for digital behavioural signals associated with mood disorders in young people | NDARC - National Drug and Alcohol Research Centre

New research aims to search for digital behavioural signals associated with mood disorders in young people

Dr Kit Huckvale was awarded $40,000 in the 2017 round of UNSW Medicine Neuroscience, Mental Health and Addiction Theme and Clinical Academic Group (CAG) Collaborative Research Seed Funding for the project, Towards a hub for mental health predictive analytics research: exploring the feasibility of a youth data registry.

image - New research aims to search for digital behavioural signals associated with mood disorders in young people

Name: Dr Kit Huckvale

Position/s: Postdoctoral Research Fellow, Black Dog Institute, UNSW Sydney

How has the Neuroscience, Mental Health and Addiction Theme and CAG enabled you to develop your research interests?

Support from the Theme and CAG has played a key role in helping me develop “The Living Lab”, a new research theme at the Black Dog Institute that focuses on the potential for personal sensing and momentary assessment (using devices such as smartphones) to identify novel signals associated with mood disorders in young people.

I’m particularly grateful for the availability of seed funding to kickstart feasibility research and technical development activities. Other important – and very practical – contributions from the Theme and CAG have included acting as convenor for multi-stakeholder discussions that helped shape the direction of the research, we well as supporting an international visit from a recognised expert in digital phenotyping for mental health.

Your project, Towards a hub for mental health predictive analytics research: exploring the feasibility of a youth data registry, was successful in the 2017 round of Theme and CAG Collaborative Research Seed Funding. Can you please tell us about the project?       

Our research vision is to be able to search for novel digital behavioural signals associated with the development of mood disorders in young people. Most serious mental illness presents before the age of 25, and early detection raises the prospect of targeted intervention that can reduce the impact and burden of disease. Our hypothesis is that there exist subtle changes in behaviour and cognitive function that are characteristic of trajectories towards future mood disorders. Our focus on digital signals, specifically, reflects the way that increasingly all aspects of our lives and behaviour are captured in data generated by the devices we carry and wear. Rather than monitoring behaviour directly, we aim to mine this ‘digital exhaust’ of sensor logs, usage data and social media posts. We already know that this ‘exhaust’ has the potential to characterise factors with known relationships to mood disorders, such as changes in sleep patterns, physical activity, social contact and speech characteristics, all without requiring active effort from users.

Being able to turn this inherently noisy and high-dimensional data into useful signals, sometimes called ‘digital phenotyping’, demands collection at scale. Seed funding from the Theme and CAG is allowing us to explore, through a structured process of consultation, the feasibility of establishing a digital behavioural registry that would permit large-scale data collection from university students. Anticipating the significant privacy and ethical dimensions to this work, we are using a participatory design approach to map potential barriers and challenge. In parallel, we are developing a new technical platform to collect sensor and other data from users’ smartphones.

What impact do you imagine the project will have?

Whether or not digital sensing can enable individual-level risk prediction in mental health remains an open question, but sensor-based signals hold the potential to also inform the design of new digital interventions that enrich information available for clinical decision making, help individuals gain new insights into their own health and reduce the burden needed for self-monitoring. For example, by capturing automatic information about sleep, physical activity, mood or social contact. From a clinical point of view, this means opening an objective window into patient experience that may be currently hidden, for example, during the trial of new therapy, post-discharge or during remission. For individuals, this means interventions that can be better tailored and personalised while demanding less effort.

How will the project support new collaborations?

The project has already enabled new collaborations. 

Within Australia, being able to articulate our technical vision for data collection has led to a new ARC Hub partnership with Deakin University to develop a fully-fledged self-report, sensing and momentary assessment data collection platform. Black Dog Institute has also been able to convene a multi-stakeholder interest group, drawing researchers from across the Theme – as well as health and third sector providers – to explore the potential for this open platform to accelerate clinical research and improvement. 

Internationally, our developing focus has led to a new collaboration with the Center for Behavioral Intervention Technologies (CBITs) at Northwestern University, led by Professor David Mohr. CBITs are recognized experts in the use of personal sensing for prediction and adaptive intervention design. Akin to genomics, where the surface area for research is larger than that which any single research group can address, our shared goal is to accelerate research progress and early clinical translation by sharing learning and, where appropriate, by coordinating our efforts.

image - New research aims to search for digital behavioural signals associated with mood disorders in young people