Automatic Identification of Tinnitus Malingering Based on Overt and Covert Behavioral Responses During Psychoacoustic Testing

Abstract

Tinnitus, or ringing in the ears, is a prevalent condition that imposes a substantial health and financial burden on the patient and to society. The diagnosis of tinnitus, like pain, relies on patient self-report. Subjective self-report measures can complicate the distinction between actual and fraudulent claims and obscure accurate severity assessments. In this study, we combined tablet-based self-directed hearing assessments with neural network classifiers to objectively determine tinnitus severity, and to differentiate participants with tinnitus (N=24) from a malingering cohort, who were instructed to feign an imagined tinnitus percept (N=28). We identified clear differences between the groups, both in their overt rating of tinnitus severity but also covert differences in their fingertip movement trajectories on the tablet surface as they performed the reporting assay. Using only 10 minutes of data, we achieved 81% accuracy classifying patients vs malingerers (ROC AUC=0.88) with leave-one-participant-out cross validation. Objective measurements of tinnitus will improve estimates of tinnitus prevalence and help to prioritize and direct funds for tinnitus compensation.

The human pupil and face encode sound affect and provide objective signatures of tinnitus and auditory hypersensitivity disorders

“…A major goal for research on sensory disorders is to identify and refine a set of objective measurements -akin to what tumor imaging and biopsy data provide for oncologists or EEG measures of epileptiform activity provide to neurologists -that inform future care providers about the subtype and severity of the sensory disorder, the likelihood to benefit from a particular treatment, or whether they have demonstrated an improvement subsequent to treatment. While some measurements have been developed to identify whether an individual has tinnitus, they are insensitive to severity and lifestyle burden 39,58,59 . Here, we found that pupil dilation and facial reactions both demonstrated a correlation with symptom severity, though it was not clear whether these measures were redundant, nor how accurately they could predict an individual's self-reported severity.…”

“…Instead, assessments rely on subjective self-report questionnaires, which can introduce vulnerabilities for malingering and false disability claims 39 and offer less insight into the underlying causes and potential treatments. Here, we hypothesized that involuntary autonomic and behavioral responses may have untapped potential as non-invasive, objective markers of tinnitus and sound sensitivity severity that are relatively easily to implement in laboratory and clinical settings.…”

“…To this point, several neuroimaging studies in participants with tinnitus and sound sensitivity have identified abnormally strong coupling between auditory cortex and the amygdala, insula, anterior cingulate cortex, and medial prefrontal cortex 33,[35][36][37] . Whereas psychoacoustic and low-level auditory assays were generally not correlated with extra-auditory features of sound aversion and psychological burden [38][39][40][41][42][43] , neuroimaging assays of enhanced coupling with extra-auditory networks have shown stronger correlations with individual differences in tinnitus and sound sensitivity severity 37,44 .…”