Background and Aims. Research just started to combine behavioural, computational and neural accounts to investigate the nature of post-stroke aphasia, but the dynamic aspect of aphasia recovery has rarely been integrated. Moreover, executive functions in aphasia recently gained attention as important contributor to language performance.
Lexical picture/word interference embraces both executive control and language processing through distracted confrontation naming, as revealed by functional Magnetic Resonance Imaging. Thus, we for the first time applied this complex paradigm in diagnostic and therapeutic settings for persons with aphasia (PWA) on the group level. In consideration of behavioural and computational data, we aimed at investigating the neural responses of persons with aphasia (a) to a novel treatment method targeting both word processing and language control by means of interferednaming, and (b) to the interfering stimuli to determine preserved and impaired processing and changes thereof.
Method. 19 PWA at a mean age of 51 years (range 21-74) and 26 months post-onset (4-63) as well as 22 matched healthy controls were included in a diagnostic fMRI studycomprising a pure naming test and the lexical fMRI interference paradigm with five
distractor conditions (phonological, associative-semantic, categorical-semantic,
unrelated word, unrelated noise) in a 3T Philips scanner. 12 participants completed the 4-weeks therapy, 11 the fMRI therapy study. During therapy, both comprehension of the interfering words and distracted but cued picture naming were trained.
Results. The distribution of lexical-semantic and -phonological impairments of PWA in the computational model was well-balanced. The novel linguistic executive treatment significantly improved pure and interfered naming for the PWA group, fostering generalisation to untrained items, reducing repetition of distractors, and ameliorating lexical-semantic processing. The predicted behavioural effects of distractor types were only found for the two semantic conditions. Neuroimaging revealed right-hemisphere compensation before treatment, activation decrease related to enhanced efficiency due to treatment, and both differential and overlapping responses to distractor types. Key brain regions encompassed inferior frontal gyrus and right posterior middle temporal gyrus as well as the basal temporal language area, cerebellum and anterior cingulate cortex.
Discussion. The novel inferfered-naming in aphasia therapy revealed to be effective,
but differential effects of distractor types were less prominent than expected.
Behavioural, computational and neural findings converge on identifying semanticexecutive processing as the core source of improvements. This account of linguisticexecutive processing in aphasia offers a step towards a comprehensive theory of neurorehabilitation.