Triggered: using computer vision to measure sound intolerance 

Authors: Samuel Smith¹; Jenna Sugai²; Nicole Buie¹; Kenneth Hancock¹; Daniel Polley¹

¹Mass Eye and Ear, Harvard Medical School; ²Mass Eye and Ear

Background: Unlike senses such as smell, it has long been thought that “hearing isn’t a strong vector for disgust” (Aurel Kolnai, 1929). Yet, millions of people with misophonia would disagree. These are individuals who are triggered, disgusted and distressed by eating- and mouth-related sounds. A socially debilitating disorder, the experience of misophonia has no established measures beyond self-report.

Method: The clinical burden of misophonia appears closely linked to affective reactivity. We hypothesized this may manifest as elevations in physiological and autonomic measures. A cohort of 61 participants (34 misophonia, 27 controls) listened to a diverse battery of sounds, including common misophonia triggers. Simultaneously, quantitative facial videography measurements were made, alongside pupillometric tracking and heart rate recordings. Statistical analyses included group and individual-level differences, and interactions with sound-type and level.

Results: We observed robust evidence that eating- and mouth-related sounds trigger distinct autonomic and physiological responses in individuals with misophonia. Quantitative videography revealed significant elevations in lower face movements, accompanied by increased pupil dilations and heart rates. Differences between misophonia and control groups were specific to sound-type rather than sound-level. Preliminary modeling suggests these combined metrics provide a viable signature for discriminating between misophonia and control groups.

Conclusion: These results highlight that eating- and mouth-related triggers cause substantial distress in individuals with misophonia, accompanied by measurable elevations in autonomic and physiological responses. Current efforts focus on refining these metrics into a predictive model capable of classifying clinical status and quantifying the degree of distress using contact-free technology.