Presentation

Laura Schiffl¹, Simon Jacob¹; ¹Department of Neurosurgery, Klinikum Rechts der Isar, Technical University of Munich, Munich, Germany

We will present an early-stage project investigating cortical (re-) organization and potential behavioral rehabilitation of language in a post-stroke aphasic patient. Language training will be complemented by chronic multi-site intracortical recordings of electrophysiological neural signals (single unit activity and local field potentials) using the Utah multielectrode array. In particular, we will be probing the interactive two step model of speech production [1] by investigating neural correlates of phonological and phonetic encoding with regard to the precise time course of lexical selection, phoneme integration and control processes. We are planning to implant a total of four electrode arrays in distributed cortical regions of an individual patient’s right frontoparietal language homologues to gain insights into the role of single neurons, their local circuits and long-range connections during language processing. The neuronal recordings will be carried out in multiple sessions per week while the patient is undergoing computerized and automated language training that includes productive and receptive tasks. This will allow us to correlate neurophysiological signals and behavioral performance with high spatiotemporal resolution. Following implantation, we plan to collect data for at least one year. The study will be conducted as a single-case, multiple baseline study. The recruited participant is a 51-year-old, right-handed, native German speaker who has suffered infarction of large parts of the left frontal, temporal and parietal lobe following a middle cerebral artery stroke in 2016. While the stroke acutely resulted in a global aphasia and severe apraxia of speech, the patient is now presenting with main symptoms of word-retrieval and -production deficits together with severe agrammatism. The patient’s language comprehension abilities remain largely intact. Results of the study may inform the implementation of language processes on the cellular and microcircuit level. Beyond potential rehabilitative effects of the high-intensity language training that might be mirrored by a change in the response profiles and functional recruitment of different cell ensembles throughout the study, we are interested in the underlying neuronal mechanisms of successful and unsuccessful word production as well as word and sentence comprehension. Recent advances in intracranial multi-electrode technology for recording large-scale brain activity have mostly been applied in the fields of sensory-motor rehabilitation or substitution [2]– including speech [3] – while language or other cognitive functions remain under-researched. New innovative approaches are needed in the field of the neurobiology of language that allow for an investigation of language functions with higher temporal and spatial resolution than can be achieved by conventional, non-invasive methods such as EEG or fMRI. The described study is highly interdisciplinary and combines contributions from fields as diverse as systems and computational neuroscience, artificial intelligence, computational linguistics, ethics, and social sciences. [1] Foygel, D., & Dell, G. S., Journal of Memory and Language, 2000, 43(2), 182–216. [2] Hochberg, L. R. et al., Nature, 2006, 442(7099), 164–171. [3] Moses, D. A. et al., The New England Journal of Medicine, 2021, 385(3), 217–227.

Topic Areas: Methods, Language Therapy