The MindTrails Project is a web-based, public research site that offers free interpretation bias training programs to promote healthier thinking patterns for people with anxiety and other emotional difficulties. Research shows that people prone to anxiety tend to think about situations negatively. MindTrails programs give you practice thinking about situations in new ways. This encourages more flexible thinking styles in your everyday life. Learn more about the science.

The Attrition project is for Predicting an early stage of dropout in digital mental health. (e/m)Health is a promising paradigm for the management of one's physical and mental health. The pervasiveness of smartphones has made it possible to deliver in-the-moment treatment for conditions such as anxiety and depression through internet-delivered (e/m)Health interventions. Despite these advances, (e/m)Health continues to suffer from low engagement and high attrition rates. In this work, we examine attrition in the MindTrails program, an online intervention for social anxiety.

Understanding different types of social supports in a network of communication from PositiveLinks. PositiveLinks is a clinic-centered engagement in care program that employs a tailored smartphone app with a private digital social support community to help people living with a chronic condition reach their care goals.

MEDIRL is a framework for modeling the visual attention allocation of drivers in imminent rear-end collisions. MEDIRL is composed of visual, driving, and attention modules. Given a front-view driving video and corresponding eye fixations from humans, the visual and driving modules extract generic and driving-specific visual features, respectively. Finally, the attention module learns the intrinsic task-sensitive reward functions induced by eye fixation policies recorded from attentive drivers. MEDIRL uses the learned policies to predict visual attention allocation of drivers.

Automated vehicles promise a future where drivers can engage in non-driving tasks without hands on the steering wheels for a prolonged period. Nevertheless, automated vehicles may still need to occasionally hand the control back to drivers due to technology limitations and legal requirements. While some systems determine the need for driver takeover using driver context and road condition to initiate a takeover request, studies show that the driver may not react to it.

We present a framework for constructing rich feature spaces from digital text communications. We then demonstrate the efficacy of our framework by applying it to a dataset of private Facebook messages in a college student population (N=103). Our results reveal key individual differences in temporal and relational behaviors, as well as language usage in relation to validated measures of trait-level anxiety, loneliness, and personality. This work represents a critical step forward in linking features of private social media messages to validated measures of mental health, wellbeing, and personality.

The present study assessed the feasibility, target engagement, and preliminary efficacy, and feasibility as primary outcomes of a free multi-session, online cognitive bias modification of interpretations (CBM-I) intervention for anxiety in a large community sample. High trait anxious participants (N = 807) were randomly assigned to a CBM-I condition: 1) Positive training (90% positive-10% negative); 2) 50% positive-50% negative training; or 3) no- training control. Further, half of each CBM-I condition was randomized to either an anxious imagery prime or a neutral imagery prime. Due to attrition, results from six out of eight sessions were analyzed using structural equation modeling of latent growth curves.

Poor adherence to long-term therapies for chronic diseases, such as cancer, compromises effectiveness of treatment and increases the likelihood of disease progression, making medication adherence a critical issue in population health. While the field has documented many barriers to adherence to medication, it has also come up with few efficacious solutions to medication adherence, indicating that new and innovative approaches are needed.

We present Peacock, a bot for social media platforms (i.e., Twitter) that is capable of learning from and imitating influential users. Peacock aims to learn its own voice and develop the potential to influence other users. Peacock's decision problem is modeled as a Markov Decision Process (MDP) and Temporal Difference (TD) learning is used to solve the Markov Decision Process. We initialize a list of influencers and pull the most recent tweets for Peacock to learn from and interact with.