Presentation
Search Abstracts | Symposia | Slide Sessions | Poster Sessions | Lightning Talks
Optimising 7T-fMRI for imaging the anterior temporal lobe
There is a Poster PDF for this presentation, but you must be a current member or registered to attend SNL 2023 to view it. Please go to your Account Home page to register.
Poster A116 in Poster Session A, Tuesday, October 24, 10:15 am - 12:00 pm CEST, Espace Vieux-Port
Saskia L. Frisby1, Marta M. Correia1, Minghao Zhang1, Christopher T. Rodgers1, Timothy T. Rogers2, Matthew A. Lambon Ralph1, Ajay D. Halai1; 1University of Cambridge, 2University of Wisconsin-Madison
Progress in our understanding of semantic cognition - the ability to comprehend and produce language, recognise and classify objects, and understand everyday events – depends on the collection of good-quality data that enable testing of, and adjudication between, different hypotheses (Frisby et al., 2023). The hub-and-spoke model (Lambon Ralph et al., 2017) posits that the anterior temporal lobe (ATL) functions as a semantic “hub”, binding modality-specific semantic information into generalisable transmodal and transtemporal representations. This theory is supported by converging evidence from neuropsychology, PET, MEG, and TMS; however, fMRI evidence for the importance of the ATL is relatively sparse. One reason for this is that the ATL is located close to the air-filled sinuses, making it susceptible to signal dropout and distortions (Visser et al., 2010; Devlin et al., 2000) – a problem that is exacerbated in ultra-high-field fMRI (7T-fMRI). This study explored the best method of imaging the ATL with 7T-fMRI while preserving image quality elsewhere. We compared the capabilities of five whole-brain acquisition sequences for revealing activation in the ATL and in other regions implicated in semantic cognition (Humphreys et al., 2015). A parallel transmit (pTx) sequence has recently been developed to reduce signal loss by using multiple transmit elements, controlled independently, to homogenise excitation over the brain tissue (Ding et al., 2022). The other four sequences comprised a 2x2 factorial design. One factor varied was the number of echoes (1 or 3) – it is well known that T2* decay differs across the brain and a shorter echo time is better suited to detect activity in the ATL, while medium and longer echo times are optimal for other regions (Halai et al., 2014; Poser & Norris, 2009). The second factor varied was the multiband factor (1 or 2) - acquiring multiple slices in parallel both improves temporal signal-to-noise (tSNR) and counteracts long repetition times associated with multi-echo sequences (Puckett et al., 2018). Healthy volunteers (N=20) performed a semantic association task that is known to generate activity in the ATL and other regions at 3T (Jung et al., 2019). We found that, compared to a standard (single-echo single band) sequence, the parallel transmit sequence did not offer improved contrast in the ATL. Multi-echo, however, offered improved contrast and precision of model fit, both in the ATL and the wider semantic network, while multiband also improved precision of model fit. In addition to these univariate results, exploratory analyses indicate that multi-echo sequences improve multivariate decoding within our regions of interest. To date, fMRI studies using 7T have not focused on imaging the ATLs due to magnetic susceptibility artefacts. Our results suggest that modified protocols such as a multiband, multi-echo sequence can reliably detect activation in susceptible regions. This will be a valuable resource for the language neuroscience community – it will enable empirical evaluation of language models such as the hub-and-spoke model; it will allow fair adjudication between competing theories; and it will provide a foundation for new findings, both within and beyond the field of semantic cognition.
Topic Areas: Methods, Meaning: Lexical Semantics