Head/Neck Health

Swallow Detectors for Dysphagia Management

LEAD: Bonnie Martin-Harris, PhD, Speech Pathologist and Alice Gabrielle Twight Professor, School of Communication

Base Technology: Mechano-acoustic sensor, published 2020

Technology status: Manufacturable prototypes, software

Clinical status: Pilot studies and Randomized Clinical Trial (NIH/NCI)

Clinical interface: Otolaryngology Head & Neck Surgery

Possible commercial path: Sibel Health

Clinical Participants: Bonnie Martin-Harris, PhD, Theresa Brancaccio, MMus

Engineering Participants: QSIB and QSI-TEAMS

Clinical Goal: Enroll ~200 study subjects with aphasia or dysphagia

Engineering Goals: (1) Validate technology, (2) Develop ML algorithms

Collaborative Goals: (1) Establish home use, (2) Demonstrate remote care

Outcome: Publication in a top clinical journal; IP protection

Head and neck cancers have reached epidemic levels in the United States, leaving many survivors with chronic swallowing disorders (dysphagia) that carry serious health consequences and few effective treatment options. This project is developing a wearable respiratory-swallow sensor, and the algorithms behind it, into an integrated system for both treating and monitoring dysphagia in these survivors. As a therapy, the sensor supports a novel intervention that trains individuals to initiate swallowing during the expiratory phase of respiration, a biomechanically advantageous phase associated with safer and more efficient swallowing. Worn at the base of the throat, the device delivers real-time visual feedback of respiratory and swallowing movements during remotely delivered telehealth sessions, and a Phase II clinical trial is evaluating whether this training improves swallowing coordination, safety, and efficiency. 

As a monitoring tool, the same sensor is being developed to track swallowing function objectively in patients' everyday environments, letting clinicians follow each patient's improvement or decline between clinical visits, so meaningful change can be caught and acted on sooner. This is especially valuable after head and neck cancer, where impairment often progresses gradually and is frequently underestimated by patient self-report, so decline can go undetected until it is well advanced and the risk of aspiration pneumonia is high. Because the sensor's raw signal is not clinically interpretable on its own, the team is developing algorithms that translate it into validated measures of swallowing physiology, calibrated against the gold-standard imaging used in clinical assessment.