You are viewing the SNL 2017 Archive Website. For the latest information, see the Current Website.

Poster C34, Thursday, November 9, 10:00 – 11:15 am, Harborview and Loch Raven Ballrooms

Mapping Both Lesion and Behaviour Structures in Stroke Aphasia

Ying Zhao1, Ajay Halai1, Matthew Lambon Ralph1;1Neuroscience and Aphasia Research Unit, School of Biological Sciences, University of Manchester

Acquired language and general cognitive deficits are common symptoms after stroke to the left hemisphere. Research has focussed on mapping these impairments onto neural substrates; classically this work relied on post mortem studies (in the late 19th century) and subsequently on soldiers returning with head injuries after the World Wars in the 20th century. Advances in medical technologies (e.g., CT and then MRI) have allowed for detailed in vivo examination of brain injuries, which have radically improved lesion-symptom mapping methods. Whilst spatial specificity has improved over history, from large areas of damage to millimetre precision, there is an underlying issue that is rarely addressed in lesion-symptom mapping research, which relates to the fact that damage to a given area of the brain is not random but constrained by the brain vasculature. Standard lesion-symptom mapping does not take this inherent statistical structure of patients’ lesions into account. The aim of this study was to uncover this lesion statistical structure and to relate it to the language and cognitive impairments in a group of seventy left hemisphere, post-stroke, chronic aphasic cases. We applied a data reduction method, varimax rotated principal component analysis, to the patients’ brain lesion maps. The underlying structure in the lesion maps revealed 20 components of which 17 were interpretable, with most of them reflecting the distribution of middle cerebral artery (MCA) sub-branches. In addition, we extracted the underlying statistical structure from a neuropsychological test battery that consisted of 21 language and cognitive assessments, revealing a four factor behavioural solution, reflecting: phonological ability, semantic ability, executive-demand and speech fluency. We used stepwise regression in order to predict behavioural factors from the principal lesion components and found significant models for all four core abilities (all p’s < 0.001). For each model, we projected the beta weights into the brain space. Phonological ability was predicted by two components, which were located in the inferior longitudinal fasciculus, posterior segment of the arcuate fasciculus and inferior frontal gyrus. Three components significantly predicted semantic ability and were located in the anterior temporal lobe extending to the medial temporal lobe, supramarginal gyrus, and angular gyrus. Executive-demand was predicted by two components covering dorsal edges of the MCA, while speech fluency was predicted by two components that were located in the middle frontal gyrus, precentral gyrus, and subcortical regions (putamen and thalamus). The identified lesion territories map very closely with detailed angiography studies and, for the first time, we have linked these to core behavioural deficits.

Topic Area: Language Disorders

Back to Poster Schedule