Distributed and integrated representations of tools in human parietal and anterior temporal regions revealed by fMRI-RSA
ISHIBASHI, R.; Humphreys, G.; Halai, A.; Tanabe-Ishibashi, A.; Hagura, N.; Ralph, M. A. L.
Show abstract
Classical models of tool knowledge and use are centred on dorsal and ventral parietal pathways. Theories of semantic cognition implicate a "hub-and-spoke" network, centred on the anterior temporal lobe (ATL), that underpins all concepts including tools. Despite their prominence, the two theoretical frameworks have never been brought together and the large discrepancy in the functional neuroanatomy addressed. We undertook a multiple-regression Representational Similarity Analysis (RSA) of task fMRI data with four (motor action, broad function, mechanical function, object structure) feature-based models. The motor action model correlated with the activation patterns in bilateral superior parietal lobules (SPL), while the models of broad function and mechanical effect aligned with the activation patterns in bilateral ATLs. The object-structure model correlated with activation patterns in bilateral middle occipital gyri. The results also showed that the ventral ATL activation patterns corresponded simultaneously with all RDM models except object structure. Furthermore, a standard univariate analysis using tool-familiarity ratings for parametric modulation revealed that classical tool-network regions (frontal, inferior parietal, and posterior middle temporal cortices) were increasingly active as the tool familiarity reduced. These results demonstrate that parietal and ATL regions are both crucial and motivate a major extension and revision of the neuroanatomical framework for tool use. Significance StatementThis study provides definitive evidence for convergent tool representation in human anterior temporal lobe (ATL), outside the traditionally focused parietal lobe as the critical centre for human tool-use ability. For many years the parietal lobe was considered crucial in recognizing and planning use of familiar objects, while regions in temporal lobe received little attention. Our advanced multi-voxel analysis with artefact-resistive fMRI scanning revealed that both non-motor (tool-function) and motor (kinematics for tool use) information convergently represented in the left ventral ATL, while showing other distributed regions encoding distinct types of tool information in anterior temporal and parietal regions. These findings highlight the ATLs crucial role in tool representation and necessitate a significant expansion of neuroanatomical framework for human tool-use ability.
Matching journals
The top 5 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Mapping the task-general and task-specific neural correlates of speech production: meta-analysis and fMRI direct comparisons of category fluency and picture naming 97%
- Atlas for the Lateralized Visuospatial Attention Networks (ALANs): Insights from fMRI and Network Analyses 97%
- Evidence for a Compensatory Relationship between Left- and Right-Lateralized Brain Networks 97%
Similar papers in this journal
- Seeing through the Static: Reduced Imagery Vividness in Aphantasia is Associated with Impaired Temporal Lobe Signal Complexity 97%
- Frontal language areas do not emerge in the absence of temporal language areas: A case study of an individual born without a left temporal lobe 96%
- Overlapping Neural Correlates Underpin Theory of Mind and Semantic Cognition: Evidence from a Meta-Analysis of 344 Functional Neuroimaging Studies 96%
Similar papers in this journal
- Hand-selective visual regions represent how to grasp 3D tools: brain decoding during real actions 97%
- The dimensionality of neural coding for cognitive control is gradually transformed within the lateral prefrontal cortex 97%
- Both default and multiple-demand regions represent semantic goal information 97%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.