Audio–visual matching of speech and non-speech oral gestures in patients with aphasia and apraxia of speech

Abstract

Background Audio–visual speech perception mechanisms provide evidence for a supra-modal nature of phonological representations, and a link of these mechanisms to motor representations of speech has been postulated. This leads to the question if aphasic patients and patients with apraxia of speech are able to exploit the visual signal in speech perception and if implicit knowledge of audio–visual relationships is preserved in these patients. Moreover, it is unknown if the audio–visual processing of mouth movements has a specific organisation in the speech as compared to the non-speech domain. Methods A discrimination task with speech and non-speech stimuli was applied in four presentation modes: auditory, visual, bimodal and cross-modal. We investigated 14 healthy persons and 14 patients with aphasia and/or apraxia of speech. Results Patients made substantially more errors than normal subjects on both the speech and the non-speech stimuli, in all presentation modalities. Normal controls made only few errors on the speech stimuli, regardless of the presentation mode, but had a high between-subject variability in the cross-modal matching of non-speech stimuli. The patients’ cross-modal processing of non-speech stimuli was mainly predicted by lower face apraxia scores, while their audio–visual matching of syllables was predicted by word repetition abilities and the presence of apraxia of speech. Conclusions (1) Impaired speech perception in aphasia is located at a supra-modal representational level. (2) Audio–visual processing is different for speech and non-speech oral gestures. (3) Audio–visual matching abilities in patients with left-hemisphere lesions depend on their speech and non-speech motor abilities.