All Research Projects
PRISM — Patient-facing Real-time Interface for Speech and Medical Communication
Exploring Real-Time AI Captioning to Improve Communication and Accessibility for Patients with Hearing Loss
Research area
Project Leads

Yew Song Cheng, BM BCh
Med devices lead, UCSF Otolaryngology Innovation Center
Share

The Challenge: Communication Barriers at the Point of Care
For patients who are deaf or hard of hearing, the clinical encounter presents a unique and often overlooked challenge. In Otolaryngology and Audiology clinics, hearing loss is frequently the primary reason for the visit — meaning the very barrier a patient is seeking treatment for also limits their ability to participate in the encounter itself. Standard alternatives like phone-based interpretation are unworkable for this population, and many patients cannot access or afford traditional hearing rehabilitation technologies.
A Practical Solution: Real-Time Captioning
Speech-to-text (STT) technology transcribes spoken language into written text in real time, bypassing the auditory pathway entirely. Early evidence suggests that real-time captioning during clinical visits can improve patient recall of medical information, reduce anxiety, and increase confidence in decision-making. Despite this, integration into routine outpatient workflows remains largely unstudied.
What We Are Doing
This prospective pilot study is enrolling approximately 50 adults with hearing loss receiving care in UCSF Otolaryngology and Audiology clinics. During routine visits, participants use AI-based real-time STT captioning displayed on a tablet, glasses, or other visual device. The study evaluates whether captioning can be feasibly integrated into standard workflows, whether it improves patient comprehension, and whether it enhances satisfaction and engagement during clinical visits.
Looking Ahead
This work reflects a broader commitment to accessible, equitable care — ensuring that the clinical encounter itself is not a barrier for the patients who need it most.


