Cerebellar activation associated with model-based estimation of tool-use consequences | Behavioral and Brain Functions | Full Text

Cerebellar activation associated with model-based estimation of tool-use consequences | Behavioral and Brain Functions | Full Text



Behavioral and Brain Functions

Cerebellar activation associated with model-based estimation of tool-use consequences

• Sayako UedaEmail authorView ORCID ID profile,
• Hiroyuki Sakai,
• Kenichi Ueno,
• Kang Cheng^ and
• Takatsune Kumada

^Deceased

Behavioral and Brain Functions201915:8

https://doi.org/10.1186/s12993-019-0158-y

©  The Author(s) 2019

• Received: 18 November 2018
• Accepted: 30 March 2019
• Published: 16 April 2019

Abstract

Background

Dexterous tool use is considered to be underpinned by model-based control relying on acquired internal models of tools. In particular, this is the case in situations where available sensory feedback regarding the consequences of tool use is restricted. In the present study, we conducted an fMRI study to identify cerebellar involvement in model-based estimation of tool-use consequences using tracking tasks with different levels of visual feedback.

Methods

Twenty healthy right-handed adults participated in this study. These participants tracked a moving target on a screen with a cursor controlled by a joystick using their right hand during fMRI scanning. For each trial, the level of visual feedback for cursor position was randomly selected from three task conditions, namely, Precise, Obscure, and No conditions.

Results

A conjunction analysis across all task conditions found extensive activation of the right cerebellum, covering the anterior lobe (lobule V) and inferior posterior lobe (lobule VIII). Also, contrasts among the three task conditions revealed additional significant activation of the left superior posterior lobe (Crus I) in the No compared to the Precise condition. Furthermore, a post hoc psychophysiological interaction analysis revealed conditional modulation of functional coupling between the right, but not the left, cerebellar region and right frontoparietal regions that are involved in self-body perception.

Conclusions

Our data show that the left Crus I is the only region that was more active in a condition where no visual feedback for cursor position was available. This suggests that the left Crus I region plays a role in model-based estimation of tool-use consequences based on an acquired internal model of tools.

Keywords

• Cerebellum
• Tool use
• Internal model
• Estimation
• fMRI