Language | HXI - Human-centered eXtended Intelligence

ConversHIVe: AI-Based Communication Training for HIV Prevention and Care

Mon, 14 Sep 2026 00:00:00 +0000

Overview

Every interaction between an HIV care team and a person with HIV can build trust or trigger disengagement. Communication quality is not peripheral in HIV care: it is a mechanism through which equity is enacted or undermined in real time. Implicit bias, often linked to factors like sexual orientation and race, can adversely affect access to care, especially for people who have already faced stigma and discrimination.

Addressing implicit bias requires self-awareness and practice in difficult communication situations. Those opportunities are scarce. HIV clinics and community-based organizations operate under limited time, staffing shortages, turnover, and restricted access to sustainable training infrastructure.

ConversHIVe is an AI-driven platform that makes that practice available on demand. It builds on our UnBIASED work, which established social behaviors such as dominance, interactivity, engagement, and warmth as measurable indicators of implicit bias. ConversHIVe uses Large Language Models to provide rapid simulation and feedback: care teams practice with simulated patients grounded in the lived experiences of people with HIV, and receive interpretable feedback derived from multimodal analysis of how the interaction actually went.

A design principle distinguishes these simulated patients from goal-directed conversational agents: they hold information back. Trust has to be earned over the course of the conversation before a patient’s underlying needs surface, which is what makes the rehearsal realistic.

ConversHIVe-Live, the project’s second phase, moves from validation to real-world implementation, embedding the training in routine HIV service delivery. It reframes communication not as a one-time competency but as a continuously improvable part of the care infrastructure.

Funding and External Collaborations

ConversHIVe is funded by the California HIV/AIDS Research Program (CHRP) under its Low Barrier Technology Interventions for HIV Prevention and Care program, with a second phase supporting real-world implementation.

The work is a collaboration between the HXI Lab and the Owen Clinic at UC San Diego, the AntiViral Research Center (AVRC) Community Advisory Board, and San Diego community-based organizations including Christie’s Place, San Ysidro Health Center, and Father Joe’s Villages.

Publications

AffAdapt: AFFect-driven ADAPTive AI Personas for Seamless Conversations

Nishanth Chidambaram, Kaustubh Paliwal, Kayla Hom, Shaoze Zhou, Chen Chen, Manas Bedmutha, Nadir Weibel

UNDERSTAND: Uplifting the New generation through DBT Education and Resilience for Social Triggers, Anxiety, Negativity, and Depression

Fri, 11 Jul 2025 00:00:00 +0000

Overview

Many college students experience comorbid depression and anxiety, exacerbating symptom severity and functional impairment. While evidence-based treatments like CBT and DBT demonstrate efficacy, they lack real-time, context-aware delivery at moments of heightened distress.

We are developing a mobile health (mHealth) system that leverages ubiquitous computing, wearable sensing, and machine learning to detect physiological and socio-behavioral indicators of distress in situ. Upon detection, the system delivers Just-In-Time Adaptive Interventions (JITAI) informed by DBT principles to provide personalized, contextually relevant support.

By targeting transdiagnostic mechanisms such as emotional dysregulation, interpersonal challenges, and cognitive distortions, our approach aims to improve scalability and relevance across diverse populations with comorbid mood and anxiety disorders.

Funding and External Collaborations

UNDERSTAND is funded by the National Institute of Mental Health (NIMH) at NIH, with the funded project starting 1 August 2026. It is a collaboration between the HXI Lab and the UC San Diego Department of Psychiatry and the Herbert Wertheim School of Public Health, bringing together expertise in DBT, clinical psychology, digital mental health, and ubiquitous sensing.

UnBIASED: Understanding Biased patient-provider Interactions And Supporting Enhanced Discourse

Thu, 23 Sep 2021 00:00:00 +0000

Overview

Healthcare bias – based on patients’ race, gender, sexual orientation – and other factors lead to health disparities, such as lack of appropriate treatment and inadequate pain support. Such biases are often unintentional and “hidden” in communication between patients and doctors.

Existing approaches to address hidden bias are limited because they are removed from actual patient-doctor interactions in which bias hides. Technology offers an opportunity to design new approaches that can make den bias more visible and thus addressable.

We are investigating a new approach to address hidden healthcare bias by improving patient-doctor communication in primary care. This approach monitors body language for signs of bias and provides feedback to raise awareness of patients and doctors for opportunities to adjust their communication style.

We are partnering closely with patients and doctors to ensure this approach is guided by their experiences and needs. Through this collaborative effort, we expect to gain a deep understanding of how hidden bias is experienced and how we can address it better in the future.

More Info here: http://unbiased.health

Funding and External Collaborations

UnBIASED ia a 5-year project, funded by the National Library of Medicine (NLMR01LM013301), and it is a collaboration between the University of Washington and the HXI Lab at UC San Diego. Our ultimate goal is to create tools to support patients and the next generation of doctors to have bias-free interactions that promote healthcare access, quality, and equity.

SocialLM

Assessing communication at scale is the bottleneck. SocialLM asks whether large language models can track social behaviors directly from clinical transcripts without fine-tuning, and finds that they can, but unevenly: performance varies by patient race and by segment of the visit. Because that variability is itself an equity problem, the work introduces an agreement-weighted ensemble that improves both accuracy and stability, giving a practical route to social signal tracking at scale.

VOLI: Voice Assistant for Quality of Life and Healthcare Improvement in Aging Populations

Thu, 23 Sep 2021 00:00:00 +0000

Overview

According to the latest US Census Bureau predictions, by 2035 older people are projected to outnumber children for the first time in US history. This brings significant societal challenges based on their unique living and health-related conditions stemming from reduced sensory, motor, and cognitive capabilities, as well as multiple chronic conditions. Technology can play a pivotal role in meeting the needs of older adults in ways that preserve their independence. Voice represents a natural choice for interaction between an aging individual and their caregivers, social networks, and healthcare providers, and it becomes key for those with visual or mobility impairment.

We are working on a personalized and context-aware voice-based digital assistant to improve the quality of life and the healthcare of older adults, and consequently, to reduce caregiving burden and optimize the interactions with healthcare and service providers.

We strive for innovations in natural language understanding, deep learning, and human-computer interfaces that leverage information from EHRs, clinical ontologies, and novel patient-level terminologies to support among others the clinical use case of detecting symptom changes and medication side effects.

More Info here: http://voli.ucsd.edu

Funding and External Collaborations

VOLI is a NIH/NSF Smart and Connected Health (SCH) funded by the National Institute of Aging (NIA) at NIH. It is a collaboration between the HXI Lab and a number of experts at UC San Diego’s Qualcomm Instititute, Computer Science and Engineering, and School of Medicine in healthcare, expert systems in clinical care, EHR integration, the aging population, patient monitoring, patient self-report, machine learning for natural language processing and understanding, experimental prototyping, field studies, and software engineering of large-scale systems.