Comprehending auditory speech: previous and potential contributions of functional MRI

Abstract

Functional neuroimaging revolutionized the study of human language in the late 20 th Century, allowing researchers to investigate its underlying cognitive processes in the intact brain. Here, we review how functional MRI (fMRI) in particular has contributed to our understanding of speech comprehension, with a focus on studies of intelligibility. We highlight the use of carefully controlled acoustic stimuli to reveal the underlying hierarchical organization of speech processing systems and cortical (a)symmetries, and discuss the contributions of novel design and analysis techniques to the contextualization of perisylvian regions within wider speech processing networks. Within this, we outline the methodological challenges of fMRI as a technique for investigating speech and describe the innovations that have overcome or mitigated these difficulties. Focussing on multivariate approaches to fMRI, we highlight how these techniques have allowed both local neural representations and broader scale brain systems to be described. 3 General IntroductionThe emergence of functional neuroimaging in the 1990s offered the unprecedented opportunity to examine the processing of spoken language in the intact, functioning human brain with a spatial resolution on the order of millimetres. Building upon the highly influential neuropsychological literature of the late 19 th and 20 th Centuries, a key aim of empirical study was uncovering the neural substrates for the comprehension of speech. Here, we offer a review and synthesis of the contributions of functional imaging -in particular, functional magnetic resonance imaging (fMRI) -to our understanding of the functional neuroanatomy of speech comprehension. We also offer an evaluation of the method and ask whether fMRI will continue to contribute substantially to our knowledge, or whether other modalities (or combinations of techniques) hold more promise for the future of the field.Other authors have offered excellent overviews of the state of the art in the functional imaging of language processes more generally (e.g. Price, 2012), and in the specific methodological challenges of examining auditory processing with fMRI (Peelle 2014). For this review, we felt that it was timely to limit our discussion to the use of fMRI in the investigation of speech comprehension. The study of speech comprehension has been at the centre of significant theoretical debate in the last decade, for example in addressing the role of hemispheric asymmetries and motor contributions to speech perception. Advances in experimental design and data analysis have played an important role in refining our understanding of these issues.Here, we highlight in particular the contributions and future opportunities afforded by multivariate analyses of the BOLD response in examining the nature and content of neural representations of the speech signal. Approaches such as Representational Similarity Analysis (Kriegeskorte et al., 2008), provide promise for integrating multiple imaging modalities, allowing examinatio...

Evidence for shared conceptual representations for sign and speech

“…This was shown to be driven by the encoding of individual spoken words. A role for the anterior superior temporal cortex in representing the identity of spoken words is consistent with studies in which the intelligibility of speech has been parametrically varied or contrasted with non-speech sounds 45,46 and the suggestion that spoken word representations are detected in the more superior portion of the ATL 2 . This region was additionally sensitive to speaker identity, suggesting that spoken word forms and speaker characteristics are jointly encoded.…”

How to study spoken language understanding: a survey of neuroscientific methods

“…Probably the best known of these haemodynamic methods is functional magnetic resonance imaging (fMRI: Evans & McGettigan, 2017) in which whole brain images of blood oxygenation can be acquired with high spatial resolution (voxel dimensions of 3 mm or less are common), but with a relatively low temporal sampling rate (typically one image every 2 s). However, an alternative method -functional near infrared spectroscopy (fNIRS: Peelle, 2017), provides a different trade-off with a superior temporal resolution (tens of measurements per second), but correspondingly lower spatial resolution (∼10 mm, depending on the number of emitters/sensors used).…”

“…In contrast, with neuroscientific measures it is relatively common to acquire a continuous measure of the brain's response (e.g. fMRI: Evans & McGettigan, 2017;MEG/EEG: Wöstmann, Fiedler, & Obleser, 2017). However, the temporal nature of speech adds considerable complexity to such experiments.…”

“…Increases to the spatial resolution of brain imaging measures would be welcome, particularly for studies in which it is the fine-grained pattern of neural activity (rather than the overall magnitude or spatial location) that is used as a dependent measure (i.e. multivariate pattern analysis methods, see Evans & McGettigan, 2017 for discussion). We therefore look with interest towards developments in ultra-high field MRI (e.g.…”

A systematic review of neuroimaging approaches to single-subject studies of language processing

“…Blood Oxygenation Level Dependent (BOLD) signal became recognized as a proxy for brain 45 activity. In language research, fMRI has gained popularity due to its noninvasiveness, task 46 flexibility, and high spatial resolution compared to other modalities (Evans & McGettigan 2017). 47…”