Healthy Minds

The Laboratory for Healthy Minds (LHM) develops digital health and AI-driven tools to detect stress-related health risks early and deliver scalable, personalized digital interventions for mental health. Our work integrates multimodal data—spanning wearable biosensors, smartphone-based daily monitoring, autonomic and endocrine measures, and neuroimaging—with clinical science to ensure our digital tools are biologically grounded in how the human brain and body respond to stress, reward, and addiction.

We build multimodal AI models that combine continuous real-world data from smartwatches, biosensors, and smartphones with laboratory and clinical data to identify individual risk trajectories prior to the onset or relapse of stress-related symptoms.

We develop and evaluate breathing- and emotion-regulation interventions delivered via smartphone apps, telehealth platforms, and wearable technology to enhance accessibility and scalability. Additionally, we leverage AI to assist in generating digital health content to improve treatment outcomes for stress- and addiction-related disorders.

Using functional magnetic resonance imaging (fMRI), we examine brain activity and connectivity patterns to understand the neural mechanisms and to predict specific risk factors associated with stress and addiction.

We examine heart rate (HR) and heart rate variability (HRV) using electrocardiogram (ECG), often concurrently with fMRI. HR spontaneously fluctuates, even during resting conditions. HRV is the time interval between heartbeats and allows detection of these fluctuations in specific frequencies. HR and HRV provides insight into autonomic and cardiovascular function. We are interested in HR/HRV’s interactions with brain and behaviors in stress, addiction and coping mechanisms.

Stress responses trigger HPA axis arousal. Glucocorticoids increases dopamine efflux in the prefrontal cortex (PFC), leading to PFC dysfunction and enhancing the reinforcing effects of substances under dyscontrolled stress states. We examine stress hormone responses (e.g., cortisol, ACTH) to understand HPA axis function and its interactions with brain-behavior relationships.

By bridging AI, digital health, neuroscience, and clinical psychology, LHM aims to make effective mental health care accessible, personalized, and affordable—from early risk identification to intervention.