AMIA 2024Using ASR and NLP, we developed a pipeline to analyze social signals in audio recordings of 782 primary care visits, achieving 90.1% accuracy and fairness across patient groups. The analysis revealed significant disparities in provider behaviors, with more patient-centered communication observed toward white patients, highlighting the potential of automated tools to uncover biases and promote equitable healthcare.
JAMIA Open 2024
Implicit bias perpetuates health care inequities and manifests in patient–provider interactions, particularly nonverbal social cues like dominance. We investigated the use of artificial intelligence (AI) for automated communication assessment and feedback during primary care visits to raise clinician awareness of bias in patient interactions.
Telemedicine and e-Health 2024Augmented reality enables the wearer to see both their physical environment and virtual objects. Holograms could allow 3D video of providers to be transmitted to distant sites, allowing patients to interact with virtual providers as if they are in the same physical space. Our aim was to determine if Tele-Stroke augmented with Holo-Stroke, compared with Tele-Stroke alone, could improve satisfaction and perception of immersion for the patient.
TOCHI 2024MemoVis is a text editor interface that assists feedback providers in creating reference images with generative AI driven by the feedback comments
CHI 2024
Analyzing patient-provider communication through social signals like dominance, interactivity, engagement, and warmth can help improve care by identifying opportunities for better interactions. We introduce a machine-learning pipeline embedded in ConverSense, a web application that visualizes communication patterns across visits. A user study with clinicians and patients highlights its potential to provide actionable, context-specific feedback for enhancing communication quality and patient outcomes.
CHI 2024
Implicit bias among healthcare providers can negatively impact care quality and patient outcomes, necessitating tools to identify and address these biases. Through design sessions with 24 primary care providers, we found they prefer feedback with transparent metrics, trends across visits, and actionable tips presented in a dashboard. These insights can guide the development of interactive systems to support equitable healthcare, especially for marginalized communities.
CHI 2024Through interviews, workshops, and a global survey, this study uncovered gaps between how contouring is taught and what residents need to learn effectively. While faculty often focus on efficiency, residents seek timely, varied, and cognitively rich feedback. Key challenges include limited support for sharing reasoning and balancing clinical with teaching responsibilities. Sociotechnical solutions are proposed to bridge these gaps, such as using senior learners for peer teaching and capturing cognitive insights through in-situ video feedback.
JAMIA Open 2024
This study evaluates an AI system that provides automated feedback on social communication during primary care visits. By analyzing real clinical conversations, the system detects social signals and offers targeted feedback to clinicians. Results show that the feedback is technically reliable, usable, and has the potential to improve communication skills and patient care.
UbiComp 2024
User preferences for mobile sensing and mental health interventions vary significantly across data types, with some sensors more acceptable than others. A university-wide survey reveals that individuals willing to share one type of data are often open to others, highlighting distinct engagement patterns. These insights support the design of inclusive, scalable mental health apps tailored to student needs.