Effective neural connectivity deficits during speech auditory feedback processing in post-stroke aphasia: An fMRI study

Abstract

Post-stroke aphasia often results in significant impairments in speech production and auditory processing, necessitating a deeper understanding of the underlying neural connectivity within sensorimotor networks. In this study, we investigated neural connectivity within sensorimotor networks in individuals with post-stroke aphasia (n = 16) compared to neurotypical controls (n = 16). Functional magnetic resonance imaging (fMRI) data were acquired during speaking and listening tasks, accompanied by normal or randomly pitch-shifted (±100 cents) altered auditory feedback across trials. Dynamic causal modeling and parametric empirical bayes methods were used to analyze effective connectivity. The analysis revealed significant alterations in intrinsic and extrinsic connectivity patterns, indicating primarily decreased connectivity in the aphasia group during listening and mixed modulation during speaking. Notably, increased self-inhibition was observed in the left precentral gyrus (PrCG), suggesting impaired feedforward mechanisms essential for speech production. Additionally, functional reorganization was evident, with increased connectivity from the right PrCG to the right superior temporal gyrus, reflecting potentially compensatory processes following left-hemisphere damage. These findings contribute to the growing evidence of an impaired and redistributed bilateral speech networks following left-hemisphere stroke. Our data emphasizes the importance of examining the receptive and productive aspects of speech impairment within sensorimotor networks for aphasia treatment.