Dissociating Contributions of Expectation and Repetition Suppression Using High-Field fMRI and MEG.
Presented During:
Saturday, June 28, 2025: 11:30 AM - 12:45 PM
Brisbane Convention & Exhibition Centre
Room:
M1 & M2 (Mezzanine Level)
Poster No:
2030
Submission Type:
Abstract Submission
Authors:
Jorie van Haren1, Federico de Martino1, Sonja Kotz1, Floris Lange2
Institutions:
1Maastricht University, Maastricht, Limburg, 2Donders Institute, Nijmegen, Gelderland
First Author:
Co-Author(s):
Introduction:
To make sense of the soundscape of our surroundings, listeners continuously use contextual information to form prediction about what is likely to occur next whilst suppressing repeated sensations. Although both prediction (Tang et al., 2021) and repetition suppression (Todorovic & de Lange, 2012) aid neural sound processing, differentiating between the two remains challenging. Events that elicit surprise often coincides with changes in low-level attributes, triggering a release form adaptation. Despite their co-occurrence, expectations may modulate neural responses beyond what can be explained by repetition suppression alone.
Here, we explore influences from past, present and inferences towards the future in shaping auditory perception. We present probabilistically sampled sequences of pure tones and employ ultra-high-field (7T) functional magnetic resonance imaging to examine layer-specific effects of low-level tuning, expectations, and repetition suppression. Complimentary, we use magnetoencephalography to decern the temporal dynamics of the repetition suppression – prediction interplay.
Here, we explore influences from past, present and inferences towards the future in shaping auditory perception. We present probabilistically sampled sequences of pure tones and employ ultra-high-field (7T) functional magnetic resonance imaging to examine layer-specific effects of low-level tuning, expectations, and repetition suppression. Complimentary, we use magnetoencephalography to decern the temporal dynamics of the repetition suppression – prediction interplay.
Methods:
We outline a content-specific (tuning dependent) modelling approach, encompassing: 1) Low-level tuned activation – the expected voxel activation patterns based on population receptive field dynamics; 2) Expectations – prior probabilities and post-surprisal quantified using the DREX ideal observer model (Skerritt-Davis & Elhilali, 2021); and 3) Repetition suppression – a double-exponential decay model capturing both rapid adaptation to (near) repetitions and lingering effects across many stimuli (Fritsche et al., 2021).
To disentangle the unique contributions of repetition suppression and expectations, we applied variance partitioning. We then further separated expectations into priors and prediction errors and repeated the analysis to assess their independent contributions. For the MEG data, time-resolved regressions were used to derive temporal response functions for each component.
To disentangle the unique contributions of repetition suppression and expectations, we applied variance partitioning. We then further separated expectations into priors and prediction errors and repeated the analysis to assess their independent contributions. For the MEG data, time-resolved regressions were used to derive temporal response functions for each component.
·Experimental Methods
Results:
In early auditory areas – Heschl's gyrus and planum temporale – the variance explained by the tuning model, repetition suppression, and expectations showed substantial overlap. In later regions – planum polare and superior temporal gyrus – the unique contributions of each model became more distinct.
Expectations explained variance beyond low-level tuning and repetition suppression, with the strongest effects in deep cortical layers and minimal contributions in superficial layers. Disentangling expectations further revealed priors and prediction errors predominantly in deep layers, with prediction errors minimally present in superficial layers.
In the temporal domain, MEG analysis showed repetition suppression emerging first, characterized by a rapid onset and sustained response. This was followed by priors, marked by a peak and dip in time. Finally, a distinct surprisal response resembling Mismatch Negativity signaled the detection of violated expectations.
Expectations explained variance beyond low-level tuning and repetition suppression, with the strongest effects in deep cortical layers and minimal contributions in superficial layers. Disentangling expectations further revealed priors and prediction errors predominantly in deep layers, with prediction errors minimally present in superficial layers.
In the temporal domain, MEG analysis showed repetition suppression emerging first, characterized by a rapid onset and sustained response. This was followed by priors, marked by a peak and dip in time. Finally, a distinct surprisal response resembling Mismatch Negativity signaled the detection of violated expectations.
Conclusions:
Our findings demonstrate the key role of expectations in auditory perception across cortical hierarchies. While low-level processes like tuning and repetition suppression encode predictable input, expectations explain additional variance, reflecting the formation of flexible priors that prepare listeners for change.
Layer-specific results reveal a distinction between feedforward and feedback processes. Feedback signals, including priors and prediction errors, are strongest in deep layers and largely bypass middle layers, likely reflecting prediction integration and model updating. Temporally, repetition suppression enables rapid adaptation to stable input, followed by priors and expectation violations, which signal deviations and drive updates.
Together, these findings illustrate how past (repetition suppression), present (tuning), and future (priors, surprisal) information interact to balance stability and sensitivity to (un)predictability.
Layer-specific results reveal a distinction between feedforward and feedback processes. Feedback signals, including priors and prediction errors, are strongest in deep layers and largely bypass middle layers, likely reflecting prediction integration and model updating. Temporally, repetition suppression enables rapid adaptation to stable input, followed by priors and expectation violations, which signal deviations and drive updates.
Together, these findings illustrate how past (repetition suppression), present (tuning), and future (priors, surprisal) information interact to balance stability and sensitivity to (un)predictability.
Higher Cognitive Functions:
Decision Making 2
Modeling and Analysis Methods:
Activation (eg. BOLD task-fMRI)
EEG/MEG Modeling and Analysis
Perception, Attention and Motor Behavior:
Perception: Auditory/ Vestibular 1
Keywords:
Cortical Layers
Data analysis
FUNCTIONAL MRI
MEG
Perception
Other - Predictive processing
1|2Indicates the priority used for review
Abstract Information
By submitting your proposal, you grant permission for the Organization for Human Brain Mapping (OHBM) to distribute your work in any format, including video, audio print and electronic text through OHBM OnDemand, social media channels, the OHBM website, or other electronic publications and media.
The Open Science Special Interest Group (OSSIG) is introducing a reproducibility challenge for OHBM 2025. This new initiative aims to enhance the reproducibility of scientific results and foster collaborations between labs. Teams will consist of a “source” party and a “reproducing” party, and will be evaluated on the success of their replication, the openness of the source work, and additional deliverables. Click here for more information. Propose your OHBM abstract(s) as source work for future OHBM meetings by selecting one of the following options:
Please indicate below if your study was a "resting state" or "task-activation” study.
Healthy subjects only or patients (note that patient studies may also involve healthy subjects):
Was this research conducted in the United States?
Were any human subjects research approved by the relevant Institutional Review Board or ethics panel? NOTE: Any human subjects studies without IRB approval will be automatically rejected.
Were any animal research approved by the relevant IACUC or other animal research panel? NOTE: Any animal studies without IACUC approval will be automatically rejected.
Please indicate which methods were used in your research:
For human MRI, what field strength scanner do you use?
Which processing packages did you use for your study?
Provide references using APA citation style.
Skerritt-Davis, B., & Elhilali, M. (2021). Computational framework for investigating predictive processing in auditory perception. Journal of neuroscience methods, 360, 109177.
Tang, M. F., Kheradpezhouh, E., Lee, C. C., Dickinson, J. E., Mattingley, J. B., & Arabzadeh, E. (2021). Sensory prediction errors increase coding efficiency in mouse visual cortex through gain amplification. bioRxiv, 2021-10.
Todorovic, A., & de Lange, F. P. (2012). Repetition suppression and expectation suppression are dissociable in time in early auditory evoked fields. Journal of Neuroscience, 32(39), 13389-13395.