CLEAR: aphasia rehabilitation with EEG-based feedback - A randomised controlled trial

INTERVENTION: Group 1: Patients with left hemispheric stroke and aphasia will complete a 30 hours of a BCI‐based language training (Breitenstein et al. (2017) four days per week (Teasell et al. 2018; Brady et al. 2016). The training consists of many short runs of 30 seconds to one minute each. Each run begins with a set of cues played through speakers or headphones. Only one word can complete the cue sentence in a meaningful way. This word is therefore defined as the target word. The patient is first asked to infer the target word for the given clue sentence. Then a quick sequence of words is presented, within which the target word and various non‐target words are played several times. During the presentation of the word sequence, the patient's task is also to concentrate on the target word and ignore the non‐target words. Using machine learning and signal processing methods, the BCI system analyses the collected EEG responses. The results of this analysis are used to provide feedback to patients on how well they were able to distinguish target and non‐target words. This feedback is intended to help patients train linguistic networks in a targeted way. Group 2: Patients with left hemispheric stroke and aphasia are given 30 hours of training based on the same above mentioned task. However, unlike the first group, they have the additional task of pressing a button each time they hear the target word. The accuracy and reaction time of the button presses is used to provide feedback to the patients on how well they could distinguish target and non‐target words. CONDITION: ; I69.3 I69.3 Sequelae of cerebral infarction PRIMARY OUTCOME: The primary outcome is the language perfromance on the Aachen Aphasia Test (AAT), the most widely used standardised aphasia test in Germany. This test is administered by a speech therapist before, after and 3 months after the training. Individual changes in language functions are evaluated on the basis of the AAT manual. The T‐transformed scores of the AAT subtest results (a standardisation in which the scores are transferred to a scale with a mean of 50 and a standard deviation of 10) will be compared by repeated‐measures analyses of variance between the groups, across the different measurement time points (pre, post and follow‐up). If necessary, the Greenhouse Geisser (G‐G) correction is applied. SECONDARY OUTCOME: Secondary outcomes are: the quantification and qualification of a) functional communication through the Communication Activity Log (CAL) and spontaneous speech, b) naming abilities through the Snodgrass & Vanderwart Naming Test c) cognitive abilities based on a test battery consisting of the Digit Span Test, the Block Tapping Test, the TAP and the Trail Making Test. In addition, the following will be measured in the pre‐ and post‐training : a) maximum amplitudes, latency and onset of different event‐related potentials (especially P300 and N200) b) functional connectivity and c) structural connectivity . INCLUSION CRITERIA: aphasia (confirmed using the Aachen Aphasia Test [AAT]) aged between 18‐80 the presence of aphasia in AAT right‐handed first‐ever left‐hemispheric ischemic stroke; time onset of stroke at least 6 months ago German as mother language sufficient (cognitive and language) functions to comply with study task and requirements