Katie Matton

Graduate Trainee, Massachusetts Institute of Technology

1 active project

Personalized Models of Mental Health

Our goal is to develop new machine learning methods for detecting and forecasting mental health trajectories. We plan to study how different methods for model personalization and data-driven sub-type discovery can be used to (1) improve the accuracy of models…

Scientific Questions Being Studied

Our goal is to develop new machine learning methods for detecting and forecasting mental health trajectories. We plan to study how different methods for model personalization and data-driven sub-type discovery can be used to (1) improve the accuracy of models that forecast changes in mental health and (2) better understand and characterize sub-types of mental illness. This work is important because the methods developed can be used to support timely interventions and enhanced clinical decision making, ultimately improving outcomes for those with mental illness.

Project Purpose(s)

  • Methods Development

Scientific Approaches

We will use the All of Us data as one of several datasets to validate our machine learning methods. We will also use internal datasets we've collected, including a study of behavior and physiology in adults with Major Depression and a study of mental well-being, physiology, and behavior in college students. In addition, we will work with other datasets that include mental health and behavioral measures that are available on the NIMH Data Archive. In our study, we will develop hypotheses about which methods work well in what contexts and we will test them by comparing the accuracy, robustness, and interpretability of each method across these different datasets. A focus of our work is also on understanding which methods work well for which individuals and ensuring that models achieve equitable performance for different sub-populations.

Anticipated Findings

We anticipate that our study will yield new methods for detecting, forecasting, and characterizing patterns in mental health for various sub-populations. From a machine learning perspective, our research will provide new insights regarding how to effectively model heterogenous patient populations. From a psychology perspective, the methods produced by this study can be used to improve our understanding of the symptoms and sub-types of mental illness.

Demographic Categories of Interest

This study will not center on underrepresented populations.

Research Team

Owner:

  • Katie Matton - Graduate Trainee, Massachusetts Institute of Technology
1 - 1 of 1
<
>
Request a Review of this Research Project

You can request that the All of Us Resource Access Board (RAB) review a research purpose description if you have concerns that this research project may stigmatize All of Us participants or violate the Data User Code of Conduct in some other way. To request a review, you must fill in a form, which you can access by selecting ‘request a review’ below.