A novel and effective language-specific training: The BCI-supported aphasia training
Mariacristina Musso1, David Hübner, Cornelius Weiller1, Michael Tangermann2; 1Department of neurology, Uniklinik Freiburg, Germany, 2Donders Institute, Radboud Univ. Nijmegen, The Netherlands
Introduction: Since now language therapy for aphasia following stroke focuses on patients’ specific language deficits and it is guided and controlled by an external person, the speech therapist or a trained volunteer (Kelly, 2010). In contrast, close-loop approaches, as brain-computer-based interventions (BCI) already successfully used for motor impairments after stroke, are guided and controlled by the patients self. This study aimed to verify if this therapeutic approach can improve language deficits. For the first time, we developed and implemented a BCI-based training protocol for chronic aphasia patients after stroke. Analyzing the ongoing EEG signals while attending word stimuli, patients receive BCI feedback based on the strength of task-relevant EEG signals. Feasibility of the protocol was previously evaluated in an offline study with 20 NACs. Material and Methods: 10 patients with a left A. cerebra media infarct and chronic aphasia underwent about 30 hours (4 days per week) of effective BCI-supported online training. During each training-session, patients, seated in a ring of 6 loudspeakers, became EEG (32 channel passive Ag/AgCl electrodes) and heard a cueing sentence following to a series of 6 bisyllabic words (concrete nouns) from which one correctly finished the sentences. After each trial, they became a feedback based on whether the attended-word predicted by ERP responses matched the target word of one trial. Each of the 6 words was trained since the patients perfectly produced it and showed a stable ERP response by processing it. Than, a novel word/sentence replaced the trained word. Before and after training all patients underwent an aphasia test battery for language assessment (Aachen Aphasie Test (AAT)  and Snodgramm naming test) as well as for executive functions (TAP, Corsi, digit span and word-fluency test). Two EEG-sessions (64 EEG channels) without feedback were conducted prior to the training to calibrate the BCI system and to determine parameters of the stimulation for each patient as well as after training to compare ERP response. Before and after the training, rs-fMRI scans, anatomical images and diffusion-weighted echo-planar imaging image were acquired. Results and Discussion: First, we found that the BCI-training was feasible, despite a high-word presentation speed and unfavourable stroke-induced EEG signal characteristics. Second, the training induced a sustained recovery of aphasia, which generalized to multiple language aspects beyond the trained task. Specifically, all tested language assessments (Aachen Aphasia Test, Snodgrass & Vanderwart, Communicative Activity Log) showed significant medium to large improvements between pre- and post-training, with a standardized mean difference of 0.63 obtained for the Aachen Aphasia Test, and five patients categorized as non-aphasic at post-training assessment. Third, our data show that these language improvements were accompanied neither by significant changes in attention skills nor non-linguistic skills. Investigating possible modes of action of this brain–computer interface-based language training, rs-functional MRI showed an increase of functional connectivity of the language-network and a decreases of functional connectivity of posterior cingulate cortex with other regions of Default-mode network. We discussed the importance of the rebalancing between the language- and default mode networks for recovery from aphasia.
Topic Areas: Language Therapy, Language Production