Project Talia – AI for Improved Mental Health

Work on Project Talia has now been retired. We continue to actively explore the healthcare and AI space, with other projects within Microsoft Health Futures (opens in new tab).

One in four of us, at some point in our lives, will be affected by a mental health condition. Good mental health and well-being are fundamental to our general health and quality of life. It enables us to build resilience against everyday stresses, to work productively, to have fulfilling relationships, and to experience life as meaningful. Mental health presents one of the most challenging and under-investigated domains of machine learning research. In Project Talia we are exploring how we can best leverage AI to help improve the effectiveness of important mental health services.

Collaboration with SilverCloud Health (opens in new tab)

In this project, we are collaborating with SilverCloud Health (opens in new tab), the leading digital therapeutics platform for mental and behavioral health. This partnership aims to jointly explore how AI can be used to enhance SilverCloud Health’s digital mental health services that deliver cognitive-behavioral treatment (CBT) programs to a large and growing number of people in need of effective care. Using probabilistic machine learning frameworks, the aim is to identify new routes for personalizing treatments and improving patient engagement and clinical outcomes.

More Effective Digital Mental Healthcare with AI

For improving mental health through AI, our research focuses on the following strategies:

	Stratify Understand patient sub-types which respond best to treatment + interventions
	Personalize Tailor content and delivery to achieve optimal therapy outcomes for individual patients
	Predict Identify which patients are more likely to drop-out for earlier intervention, or different programs
	Intervene Intervene timely to ensure earlier intervention and improved outcomes
	Improve Identify successful patterns in supporter behaviour in relation to patient sub-type to improve therapy effectiveness

Events

• 11/2023 – Plos one paper: Deep learning for the prediction of clinical outcomes in internet-delivered CBT for depression and anxiety | PLOS ONE (opens in new tab)
• 09/2022 – Microsoft Research Blog: AI Models vs. AI Systems: Understanding Units of Performance Assessment – Microsoft Research (opens in new tab)
• 08/2022 – ToCHI paper: Designing Human-Centered AI for Mental Health: Developing Clinically Relevant Applications for Online CBT Treatment – Microsoft Research (opens in new tab)
• 11/2021 – arxiv paper: Understanding the Information Needs and Practices of Human Supporters of an Online Mental Health Intervention to Inform Machine Learning Applications – arxiv.org (opens in new tab)
• 05/2021 – World Psychiatry paper: The promise of machine learning in predicting treatment outcomes in psychiatry (opens in new tab)
• 08/2020 – ToCHI paper: Machine Learning in Mental Health: A Systematic Review of the HCI Literature to Support Effective ML System Design (opens in new tab)
• 07/2020 – JAMA Network Open paper: A Machine Learning Approach to Understanding Patterns of Engagement With Internet-Delivered Mental Health Interventions (opens in new tab)
• 07/2020 – Microsoft Research Blog: A path to personalization: Using ML to subtype patients receiving digital mental health interventions (opens in new tab)
• 04/2020 – CHI 2020 paper: Understanding Client Support Strategies to Improve Clinical Outcomes in an Online Mental Health Intervention (opens in new tab)
• 03/2020 – Microsoft Research Blog: Data-driven insights for more effective, personalized care in online mental health interventions (opens in new tab)
• 10/2019 – Collaboration with SilverCloud Health announced at Microsoft Future Decoded (opens in new tab)
• 10/2019 – Microsoft Research Blog: Microsoft collaborates with SilverCloud Health to develop AI for improved mental health (opens in new tab)
• 09/2019 – ACII 2019 workshop: Machine Learning for Affective Disorders (opens in new tab) (ML4AD)