Sensorimotor impairment of speech auditory feedback processing in aphasia

Abstract

We investigated the brain network involved in speech sensorimotor processing by studying patients with post-stroke aphasia using an altered auditory feedback (AAF) paradigm. We combined lesion-symptom-mapping analysis and behavioral testing to examine the pervasiveness of speech sensorimotor deficits and their relationship with cortical damage. Sixteen participants with aphasia and sixteen neurologically intact individuals completed a speech task under AAF. The task involved producing speech vowel sounds under the real-time pitch-shifted auditory feedback alteration. This task provided an objective measure for each individual's ability to compensate for mismatch (error) in speech auditory feedback. Results indicated that compensatory speech responses to AAF were significantly diminished in participants with aphasia compared with control. We observed that within the aphasic group, subjects with lower scores on the speech repetition task exhibited greater degree of diminished responses. Lesion-symptom-mapping analysis revealed that the onset phase (50-150 ms) of diminished AAF responses were predicted by damage to auditory cortical regions within the superior and middle temporal gyrus, whereas the rising phase (150-250 ms) and the peak (250-350 ms) of diminished AAF responses were predicted with damage to the inferior frontal gyrus and supramarginal gyrus areas, respectively. These findings suggest that damage to the auditory, motor, and auditory-motor integration networks are associated with impaired sensorimotor function for speech error processing. We suggest that a sensorimotor integration network, as revealed by brain regions related to temporal specific components of AAF responses, is related to speech processing and specific aspects of speech impairment, notably repetition deficits, in individuals with aphasia.

Keywords: Aphasia; Auditory feedback; Lesion-symptom-mapping; Sensorimotor integration; Speech motor control.

Figure 1

The auditory-motor integration model of speech. In this model, the auditory-motor interface transforms speech motor plans into forward prediction of auditory feedback. The auditory system compares forward predictions with actual speech feedback to detect prediction errors in response to altered auditory feedback (AAF). The auditory system also detects sensory prediction errors in response to AAF by comparing the intended auditory target with actual feedback from speech. The generated sensorimotor errors are translated into corrective signals by the auditory-motor interface to adjust the speech motor parameters to control speech output in response to AAF.

Figure 2

The experimental paradigm for altered auditory feedback (AAF). The task involves steady production of the speech vowel sound /a/ for approximately 2–3 seconds while a pitch-shift stimulus (PSS) at ±100 cents alters the online auditory feedback of speech. The PSS stimuli are delivered at a randomized time between 750–1250 ms after the onset of speech.

Figure 3

Lesion overlap maps in aphasic speakers (n = 16). The maps show lesion distribution on coronal (top) slices in MNI space for the sample, with warmer colors representing more lesion overlap across aphasic speakers (dark red areas represent lesion overlap across at least N = 8 individuals).

Figure 4

The overlaid profiles of grand-average speech compensation responses to altered auditory feedback (AAF) for A) upward (+100 cents) and B) downward (−100 cents) pitch-shift stimuli in 16 speakers with aphasia and 16 control individuals. The temporal profile of responses are divided into onset (50–150 ms), rise (150–250 ms), peak (250–350 ms), and rebound (350–450 ms) time periods relative to the onset of pitch-shift stimuli at 0 ms.

Figure 5

Anatomical representation of lesion on the superior and middle temporal gyrus associated with diminished responses to AAF at the onset phase of speech compensation (50–150 ms) overlaid on A) reconstructed brain surface and B) coronal slices in MNI space. C) Demonstrates correlation between speech compensation ratio in aphasia normalized to the mean of control group and lesion volume on the superior and middle temporal gyrus. D) Demonstrates correlation between speech compensation ratio and patients’ scores on the Western Aphasia Battery (WAB).

Figure 6

Anatomical representation of lesion on the inferior frontal gyrus associated with diminished responses to AAF at the rising phase of speech compensation (150–250 ms) overlaid on A) reconstructed brain surface and B) coronal slices in MNI space. C) Demonstrates correlation between speech compensation ratio in aphasia normalized to the mean of control group and lesion volume on the inferior frontal gyrus. D) Demonstrates correlation between speech compensation ratio and patients’ scores on the Western Aphasia Battery (WAB).

Figure 7

Anatomical representation of lesion on the supramarginal gyrus associated with diminished responses to AAF at the peak of speech compensation (250–350 ms) overlaid on A) reconstructed brain surface and B) coronal slices in MNI space. C) Demonstrates correlation between speech compensation ratio in aphasia normalized to the mean of control group and lesion volume on the supramarginal gyrus. D) Demonstrates correlation between speech compensation ratio and patients’ scores on the Western Aphasia Battery (WAB).