Measurement of language laterality using functional transcranial Doppler ultrasound: a comparison of different tasks

Zoe V.J. Woodhead,Holly A. Rutherford,Dorothy V.M. Bishop

doi:10.12688/wellcomeopenres.14720.1

Abstract

Background: Relative blood flow in the two middle cerebral arteries can be measured using functional transcranial Doppler sonography (fTCD) to give an index of lateralisation as participants perform a specific task. Language laterality has mostly been studied with fTCD using a word generation task, but it is not clear whether this is optimal. Methods: Using fTCD, we evaluated a sentence generation task that has shown good reliability and strong left lateralisation in fMRI. We interleaved trials of word generation, sentence generation and list generation and assessed agreement of these tasks in 31 participants (29 right-handers). Results: Although word generation and sentence generation both gave robust left-lateralisation, Bland-Altman analysis showed that these two methods were not equivalent. The comparison list generation task was not systematically lateralised, but nevertheless laterality indices (LIs) from this task were significantly correlated with the other two tasks. Subtracting list generation LI from sentence generation LI did not affect the strength of the laterality index. Conclusions: This was a pre-registered methodological study designed to explore novel approaches to optimising measurement of language lateralisation using fTCD. It confirmed that sentence generation gives robust left lateralisation in most people, but is not equivalent to the classic word generation task. Although list generation does not show left-lateralisation at the group level, the LI on this task was correlated with left-lateralised tasks. This suggests that word and sentence generation involve adding a constant directional bias to an underlying continuum of laterality that is reliable in individuals but not biased in either direction. In future research we suggest that consistency of laterality across tasks might have more functional significance than strength or direction of laterality on any one task.

Highlights

Relative blood flow in the two middle cerebral arteries can be measured using functional transcranial Doppler sonography to give an index of lateralisation as participants perform a specific task
We considered two questions: (a) whether the laterality index on a task was related to the number of words produced, and (b) whether bimodality of laterality indices was an artefact of the method of computation, which involved identifying a peak
Shapiro-Wilk tests confirmed that laterality indices (LIs) values from all three conditions were normally distributed (LG: p = .084; sentence generation (SG): p = .577; word generation (WG): p = .752)

Summary

Introduction

Relative blood flow in the two middle cerebral arteries can be measured using functional transcranial Doppler sonography (fTCD) to give an index of lateralisation as participants perform a specific task. Conclusions: This was a pre-registered methodological study designed to explore novel approaches to optimising measurement of language lateralisation using fTCD It confirmed that sentence generation gives robust left lateralisation in most people, but is not equivalent to the classic word generation task. Measuring lateralisation with Functional Transcranial Doppler ultrasonography (fTCD) Functional transcranial Doppler ultrasonography (fTCD) is a minimally invasive technique which is portable and relatively inexpensive, making it possible to investigate hemispheric dominance in relation to typical and impaired language function across a wider population This method measures bilateral blood flow in the middle cerebral arteries (MCA) – which may be regarded as a proxy of cerebral activity in regions supplied by these vessels due to neurometabolic coupling (Lohmann et al, 2006). The principal drawback of fTCD is that, while it is reliable at detecting increased blood flow in one hemisphere during an activation task with good temporal resolution, it cannot be used to localise function within a hemisphere

Objectives

Methods

Results

Discussion

Conclusion