Neuroanatomical Substrates of Lexical Selection: Contribution and Functional Reorganization of the Left Frontal Aslant Tract

When speaking, we are able to access thousands of words in our mental lexicon and pick the specific word labelling the intended concept in less than a second. Psycholinguists have broken down this intricate process into several more confined processing stages. First, a non-linguistic concept simultaneously activates multiple semantically related lexical representations. This is followed by lexical selection, when a single lexical representation is selected for further processing. Which brain areas carry out lexical selection remains incompletely understood. Clinical and functional neuroimaging evidence suggests that lexical selection heavily relies on the left inferior frontal gyrus (IFG). However, another series of studies implicate a different neuroanatomical structure – the left supplementary motor complex (SMC) – in the same process. Recently, a novel white-matter pathway that connects the IFG and the SMC has been discovered and termed as the frontal aslant tract (FAT). The FAT enables integration between the neural processes carried out by the IFG and the SMC. Therefore, we hypothesized that lexical selection relies on the coordinated network comprising both the IFG and the SMC integrated via the FAT. To test this hypothesis, we investigated whether disruption of the FAT following stroke is associated with a lexical selection deficit. Twenty individuals with subacute or chronic stroke underwent behavioral assessment and magnetic resonance imaging. Behavioral assessment included linguistic tests tapping into lexical selection abilities: sentence completion with varying context constraint and picture-word interference task. Using the magnetic resonance imaging data, we reconstructed the left FAT and extracted its volume, which served as a proxy for tract integrity following stroke. We related the volume of the left FAT and the measures of lexical selection. We replicated the well-established involvement of the IFG in lexical selection. However, we did not observe a robust relation between FAT volume and lexical selection measures. We speculate that disruption of the FAT may have caused a lexical selection deficit early post-stroke, followed by successful recovery prior to the behavioral assessment. This suggests that the network subserving lexical selection may undergo successful functional reorganization over the course of post-stroke recovery. This warrants further studies that track the neuroanatomical correlates of lexical selection throughout the acute to chronic phases after stroke.