Our work is at the intersection of Human-Computer Interaction (HCI), Artificial Intelligence (AI), eXtended Reality (XR) and Human-Centered Design (HCD). We strive to understand how pervasive and emerging technologies can help bridge the gap between the physical and digital worlds, especially in health & healthcare and education, and above all in life- and time-critical settings where the cost of a breakdown is immediate.
HXI’s work has been funded by the National Institutes of Health (NIH), the National Science Foundation (NSF), the Agency for Healthcare Research and Quality (AHRQ), the Department of Veteran Affairs (VA), the Center for AIDS Research (CFAR), the UC San Diego Office of Research Affairs, the Robert Wood Johnson Foundation, the Moxie Foundation, the Swiss National Science Foundation (SNSF), the European Commission (FP6), the California HIV/AIDS Research Program (CHRP), the Naval Medical Center San Diego, the T. Denny Sanford Institute for Empathy and Compassion, Intuitive Surgical, IBM Research, Lytx, Adobe Research and Boeing.
ConversHIVeConversHIVe-LiveEmpathIQ
An AI training platform for clinical communication: LLM-driven simulated patients stage difficult conversations, and social signal processing turns each rehearsal into interpretable feedback.
A mobile health system pairing wearable sensing and machine learning with just-in-time adaptive interventions, targeting emotional dysregulation and other transdiagnostic mechanisms in the moment.
Research on student well-being at campus scale, conducted with WILLO, UC San Diego’s campus-wide platform, toward support that reaches an entire student population rather than only those who seek it.
Preference-based alignment for speech and audio: how people judge generated output, and whether annotation protocols validated on text transfer to what they hear.
A Design Lab platform for collaborative, data-driven decisions on complex societal problems, combining data integration, modeling, simulation, and immersive visualization.
iContourVRContouriConTutor
Interactive training systems for radiation oncology contouring: iContour, VRContour, and iConTutor deliver practice, immersive delineation, and automated feedback to residents.
Mixed reality for stroke care at a distance: holograms of a patient, or of their imaging, are placed in three-dimensional space and transmitted to a remote stroke neurologist.
ConverSenseConverStandSocialLM
Social signal processing for hidden bias in primary care: automated analysis of patient-provider communication, and feedback designed together with patients and doctors.
DriveSimQuest
Human-centered design and evaluation of automated driving, from the explanations an autonomous vehicle gives its passenger to the sensing and coaching of the person behind the wheel.
Surgical telementoring in mixed reality: a remote expert sees the novice’s operating field in three dimensions and guides the procedure from anywhere.
PintARHoloCPRAcuVR
Foundational XR research at HXI: interaction techniques for headsets and spatial displays, and immersive environments used as controlled settings for studying behavior.
XR systems for remote collaboration on physical tasks, using spatial computing and alternate 3D representations to reduce the effort of communicating across distance.
MemoVisPaperToPlaceSweeperBot
Interfaces spanning 2D documents and 3D space: MemoVis, PaperToPlace, and SweeperBot support design feedback, spatialized instructions, and accessible 3D browsing.
Empirical research and practical guidance for the ethics of digital health research, spanning consent, privacy, and the terms under which data is collected.
These projects come from earlier phases of the lab’s work. The methods, systems, and collaborations they established run through the research above, and each page remains available in full.
A context-aware voice assistant for older adults, supporting everyday quality of life and connecting them to their healthcare.
Learning aids for online education at scale: minimally-invasive context about a learner drives timely AI and peer support.
A multimodal sensing approach to acute stroke assessment: body posture, facial droop, pupil tracking, and speech are combined into a per-patient signature of stroke-related deficits.
Sketch-based prototyping for augmented reality: designers draw spatial experiences directly in the space where they will be used.
Multimodal data capture for real-world studies: synchronised sensor streams, with real-time and near-time analysis through an extensible plugin interface.
Real-time resuscitation guidance in mixed reality: visual and spatial cues sit on the patient, so a novice sees what to do where it has to be done.
A programming environment in AR and VR built on embodied tools and spatial metaphors rather than the traditional wall of text.
Two do-it-yourself mask designs built during the COVID-19 shortage, with tested filtration performance and open instructions for making them.