Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Mapping the task-general and task-specific neural correlates of speech production: meta-analysis and fMRI direct comparisons of category fluency and picture naming
Poster D21 in Poster Session D, Saturday, October 26, 10:30 am - 12:00 pm, Great Hall 4
Gina Humphreys1, Matt Lambon Ralph1; 1University of Cambridge
Improving our understanding of the neural network engaged by different forms of speech production is a crucial step for both cognitive and clinical neuroscience. Semantically driven speech production involves a network of regions, yet the function of each of the sub-regions of the network remains unclear. Here, we directly investigate this issue using two of the most commonly utilised speech production paradigms in research and the clinic: picture naming and category fluency. Harnessing the similarities and differences between the two tasks offers a powerful methodology to delineate the core systems recruited for speech production, as well as revealing task-specific processes. These tasks were formally compared in 1) a meta-analysis of existing neuroimaging studies; and 2) a tightly controlled fMRI study that directly compared the activation from an overt picture naming task , a paced category fluency task, and a non-semantic speech production control task. Together the results showed that both picture naming and category fluency tasks engaged a shared fronto-temporal speech production network, including executive and motor frontal areas, as well as semantic representational regions in the anterior temporal lobe (ATL). This network presumably reflects the core speech production systems. Task-specific differences in the degree of language-laterality, however, were also revealed. Specifically, category fluency was associated with boosted left lateral frontal activation, whereas naming showed an overall weaker but more bilateral pattern with no significant hemispheric differences. In terms of semantic representation, the ATL showed bilateral engagement for fluency and naming tasks, in contrast no evidence was found for the involvement of the angular gyrus (AG) in either task (in fact, both tasks showed AG deactivation relative to rest and the control tasks) thereby questioning the contribution of the AG to speech production. The results have implications for neurocomputational speech production models, and the choice of task best used for clinical assessment of production abilities in patients.
Topic Areas: Language Production,