Lorca Puls, Diego L.; (2019) Using non-invasive stimulation of the undamaged brain to guide the identification of lesion sites that predict language outcome after stroke. Doctoral thesis (Ph.D), UCL (University College London).

Preview

Text Lorca-Puls_10069779_thesis_revised.pdf Download (6MB) | Preview

Abstract

Disrupting the neural activity in the left anterior supramarginal gyrus (SMG) or opercular part of the left inferior frontal gyrus (pOp) with repetitive transcranial magnetic stimulation (TMS) has been demonstrated to cause a transient slowing of response times during phonologically more than semantically demanding tasks. Likewise, a wealth of functional magnetic resonance imaging (fMRI) studies have shown increased activation in SMG and/or pOp for phonological relative to semantic processing. Here I set out to investigate whether, and how frequently, stroke damage to SMG and/or pOp results in persistent phonological processing impairments in a large sample of 262 right-handed English-speaking adults, who were tested at least 1 year after a left-hemisphere stroke. In Experiment I, I compared the effect of damage to different parts of SMG and pOp that were defined by regions of interest from either TMS or fMRI studies of phonological processing in neurologically-normal individuals. I found that the incidence of phonological processing impairments was predicted significantly better by the presence or absence of damage to SMG and pOp regions defined by TMS studies than SMG and pOp regions defined by fMRI studies. Moreover, the discriminatory power (for segregating patients with and without phonological abilities) of the TMS sites was not improved further when combined with the fMRI sites. In Experiment II, I adapted the borders of the TMS SMG and pOp regions to include the surrounding grey and white matter where the presence or absence of stroke damage was consistently associated with the presence or absence of phonological processing impairments. The presence or absence of damage to these new TMS-guided regions was able to explain the incidence of phonological impairments better than the original TMS regions, even in a new sample of patients that was entirely independent of the region identification process. In Experiment III, I showed that damage to the TMS-guided regions accounted for the incidence of phonological impairments substantially better than damage to an alternative set of regions derived from voxel-based lesion-deficit mapping techniques that search the whole brain for areas that are most frequently damaged in those with phonological impairments. However, the best classification accuracy was observed when the analysis took into account a combination of regions from TMS-guided and voxel-based lesion-deficit mapping approaches. In Experiment IV, I investigated the nature of the functional impairment caused by SMG or pOp lesions and found that damage to either region impaired covert and overt phonological processing abilities more than semantic processing abilities, as predicted by prior TMS and fMRI studies of neurologically-normal subjects. Finally, the behavioural effects of damage were remarkably similar (i.e. no statistically significant differences) for both TMS-guided sites (i.e. pOp and SMG). In conclusion, the fact that damage to the TMS-guided SMG and pOp regions impaired phonological processing abilities years after stroke onset, suggests that these regions are critical for accurate phonological processing (both overt and covert) and that other brain areas are not typically able to fully compensate for the contribution that these regions make to language processing. More broadly, the results illustrate how non-invasive stimulation of the undamaged brain can be used to guide the identification of regions where brain damage is likely to cause persistent behavioural effects. By combining these regions of interest with those derived from other lesion-deficit mapping approaches, I was not only able to explain the presence, but also the absence, of phonological processing impairments in a large cohort of patients.

Download activity - last month

Export as CSVXMLJSON

Download activity - last 12 months

Export as CSVXMLJSON

Downloads by country - last 12 months

Export as CSVXMLJSON

Archive Staff Only

View Item