Localization of Free-Field Sound Sources in the Chronic Phase of Mild Ischemic Stroke.

Abstract

Accurate sound localization relies on the transformation of binaural cues into stable spatial representations, yet the neural mechanisms supporting this process remain incompletely understood. Mild stroke provides a unique opportunity to study the vulnerability of auditory spatial processing within distributed neural networks. We investigated azimuthal sound localization in chronic-phase survivors of mild stroke without hearing aid use using broadband, low-frequency, and high-frequency noise and compared it against a headphone-based lateralization task. Most patients exhibited localization accuracy comparable to healthy listeners. However, 4 of 14 patients showed a striking alteration of auditory spatial perception: instead of a continuous mapping of azimuth, responses clustered into bi- or trimodal (left-center-right) categorical patterns. Comparable deficits were also observed in one of the five age-matched control participants. To our knowledge, such localization patterns have not been reported in listeners without neurological disease. Atypical localization was most pronounced with low-frequency stimuli, suggesting a different role of the respective binaural cues in forming a spatial representation of sound. Differences between loudspeaker-based localization and existing headphone-based lateralization data further suggest that these paradigms engage distinct auditory spatial representations. The findings support the view that auditory space is constructed through higher-order, supramodal spatial mechanisms that are particularly vulnerable to right-hemispheric damage. Overall, the data highlight the challenges of quantifying spatial hearing of stroke survivors both on an individual and on a population level. Challenges include previously undocumented stimulus and method dependencies, an unknown mixing of auditory and neurological factors, and the absence of normative data for a nonstroke control cohort.