Repository logo
 

Increased recruitment of domain general neural networks in language processing following Intensive Language-Action Therapy – fMRI evidence from people with chronic aphasia.

Published version
Peer-reviewed

Change log

Authors

Dryer, Felix 
Doppelbauer, Lea 
Buscher, Verena 
Arndt, Verena 
Stahl, Benjamin 

Abstract

Purpose: The present study aimed to provide novel insights into the neural correlates of language improvement following Intensive Language Action Therapy (ILAT; also known as Constraint Induced Aphasia Therapy, CIAT).

Method: Sixteen people with chronic aphasia underwent clinical aphasia assessment (Aachen Aphasia Test, AAT), as well as functional magnetic resonance imaging (fMRI), both administered before (T1) and after ILAT (T2). The fMRI task included passive reading of single written words, with hashmark strings as visual baseline.

Results: Behavioral results indicated significant improvements of AAT scores across therapy and fMRI results showed T2−T1 blood oxygenation level dependent (BOLD) signal change in the left precuneus to be modulated by the degree of AAT score increase. Subsequent region-of-interest (ROI) analysis of this precuneus cluster confirmed a positive correlation of T2−T1 BOLD signal change and improvement on the clinical aphasia test. Similarly, the entire default mode network (DMN) revealed a positive correlation between T2−T1 BOLD signal change and clinical language improvement.

Conclusion: These results are consistent with a more efficient recruitment of domain general neural networks in language processing, including those involved in attentional control, following aphasia therapy with ILAT.

Description

Keywords

Aphasia, Humans, Language, Language Therapy, Magnetic Resonance Imaging, Neural Networks, Computer, Stroke

Journal Title

American Journal of Speech-Language Pathology

Conference Name

Journal ISSN

1058-0360
1558-9110

Volume Title

30

Publisher

American Speech-Language-Hearing Association
Sponsorship
MRC (unknown)
Medical Research Council (MC_UU_00005/14)
This work was supported by the Deutsche Forschungsgemeinschaft (pu 97/15-1 and 97/15-2 awarded to F. P.), the Deutsche Akademische Austauschdienst (fellowship to G. L.), and the Einstein Center for Neuroscience Berlin (fellowship awarded to L. D.)