An Electrophysiology-Silent fMRI Study to Inform the Neurofunctional Basis of Startle Habituation

1324 

Abstract Submission 

Laura Naysmith¹, Owen O'Daly¹, Ana Beatriz Solana², Florian Wiesinger², Simon Hill¹, Steven Williams¹, Veena Kumari³

¹King's College London, London, England, ²GE Healthcare, Munich, Germany, ³Brunel University, London, England

Laura Naysmith  
King's College London
London, England

Owen O'Daly  
King's College London
London, England

Ana Beatriz Solana  
GE Healthcare
Munich, Germany

Florian Wiesinger  
GE Healthcare
Munich, Germany

Simon Hill  
King's College London
London, England

Steven Williams  
King's College London
London, England

Veena Kumari  
Brunel University
London, England

The acoustic startle reflex provides an important model system for exploring information processing mechanisms and stem from acoustic-focused rodent models, translating to human neuroimaging findings. The startle reflex is subject to modulation, such as habituation which leads to attenuation of the reflex response, and is thought to reflect attentional filtering of redundant information (Thompson & Spencer, 1966). Indeed, startle habituation impairments have been observed in disorders such as Parkinson's disease, schizophrenia, post-traumatic stress disorder, and Huntington's disease (review, McDiarmid, Bernandos & Rankin, 2017).
To better understand information processing in healthy populations and, subsequently, dysfunction in clinical populations, the current study aimed to explore the neurofunctional basis of startle habituation, which remains poorly understood. In a mixed-sex, healthy young adult sample, the study combined simultaneous electromyography (EMG) to measure the startle reflex as an eyeblink, and fMRI to provide a closer coupling of stimulus, behaviour, and BOLD MR signal and hence, greater inference of the association of regional brain activation with startle behaviour and startle stimulus processing. Secondly, we present a major advancement in this work with the application of a silent fMRI sequence, Looping Star (Wiesinger, Menini, & Solana, 2019), which allowed for the continuous presentation of acoustic stimuli by minimising gradient-related acoustic noise. Neural circuitry underpinning startle habituation was hypothesised in the brainstem (Kuhn et al., 2020) and thalamus (McDowell et al., 2006) and when EMG-assessed measures of startle habituation were modelled as a covariate at the group-level fMRI analysis, neural activity was expected to decrease with more startle habituation.

Forty-two participants (M= 23.71 years) were presented with a passive auditory startle paradigm in the 3T MRI scanner, which consisted of loud auditory clicks to elicit startle. EMG measures of startle were acquired in real-time using AcqKnowledge (BIOPAC Systems Inc.) and required online band-pass filtering and offline filtering, of which an innovative pipeline was developed to suit the frequency spectrum of Looping Star. We used a silent functional MRI multi-echo sequence called Looping Star (TE1/ TE2/ TE3/ TR= 0ms/ 17.9ms/ 35.8ms/ 2.62s). Imaging data were pre-processed and analysed in SPM12. An optimal combination of the gradient echoes (TE= 17.9ms and 35.8ms) was used (Fig 1). At the group-level, a linear regression was conducted. EMG data, analysed in SPSS Statistics, calculated startle habituation as a regression slope per participant and these regression slope values were included as a covariate of interest at group-level. We conducted a regions of interest (ROI) approach (thalamus, McDowell et al., 2006; brainstem, Kuhn et al., 2020). We also tentatively conducted a whole brain analysis, as Hermann et al. (2020) highlighted the Default Mode Network (DMN) during startle habituation.

BOLD fMRI activity in the thalamus, brainstem, and a cluster surmising the right putamen and extending to the right insula was observed (Fig 2). BOLD response in these regions correlated significantly positive with startle habituation slope values, thus indicating a decrease in BOLD MR response with more startle habituation (negative startle habituation slope value).

Simultaneous EMG data enriched fMRI findings of the neural basis of startle habituation by identifying a positive relationship between acoustic startle habituation and sub/cortical regions. The direction of relationship between fMRI BOLD and startle habituation can be used to further examine how the acoustic primary startle circuit is mediated, such as through reticular activating system (McDowell et al., 2006) or cortical networks such as the DMN (Hermann et al., 2020) and encourage further fMRI investigations in clinical populations which show aberrations in startle habituation.

Higher Cognitive Functions Other ²

Activation (eg. BOLD task-fMRI) ¹

Other Methods

BOLD fMRI

ADULTS

ELECTROPHYSIOLOGY

FUNCTIONAL MRI

Motor

NORMAL HUMAN

Other - novel silent fMRI