Biology subjects

Burg, M. F.

Publications and source records attributed to Burg, M. F..

4 recordsLinked to original sources

bioRxiv2025

Interpreting convolutional neural networks to study wide-field amacrine cell inhibition in the retina

Wide-field amacrine cells (ACs) play a unique role in retinal processing by integrating visual information across a large spatial area. Their inhibitory influence has been implicated in multiple retinal functions such as differential motion detection and the suppression of retinal activity during eye movements. However, a coherent understanding of their general function is lacking due to difficulties in directly recording from these cells and identifying effective visual stimuli to activate them. In this study, we used convolutional neural networks (CNNs) to investigate widefield inhibition mediated by wide-field ACs in the marmoset retina. We trained CNNs to mimic the function of the retina by predicting retinal ganglion cell (RGC) population responses to naturalistic movie stimuli and optimising the most exciting inputs (MEIs) to visualise RGCs receptive field (RF) structures. We then optimized suppressive surrounds beyond classical RGC RF boundaries, intended to capture the inhibitory effect of wide-field ACs on RGC activity. These optimized surrounds reduced MEI-elicited activity by 10% to 30%, demonstrating that CNNs not only mimic retinal responses but can also reveal hidden computational aspects of wide-field inhibition. However, suppression strength and generalization varied across architectures and datasets, indicating potential model-specific effects, high-lighting the importance of cautious interpretation. Overall, our approach illustrates how interpretability methods applied to artificial neural networks can offer new hypotheses regarding biological retinal computation, paving the way for targeted experimental validation. The code is available here.

neuroscience↗

bioRxiv2025

Spatial Adaptation of Primate Retinal Ganglion Cells between Artificial and Natural Stimuli

The retina encodes a broad range of stimuli, adapting its computations to features like brightness, contrast, or motion. However, it is unclear to what extent it also adapts to spatial frequency content - as theories of efficient coding would predict - for instance, when switching between natural scenes and white noise. To address this, we analyzed neural activity of marmoset retinal ganglion cells (RGCs) in response to white noise and naturalistic movie stimuli. We trained linear-nonlinear models on both stimuli, evaluated their performance and compared their receptive fields (RFs) across the stimulus domains. We found that the models with spatial filters trained on either one of the stimulus ensembles were not able to predict the neural activity on the other as accurately as the models trained on the target stimulus. This suggests that spatial processing adapts to stimulus statistics. Different RGC types exhibited distinct changes: the midget OFF cells RFs became enlarged under natural movie statistics, resulting in a lower cutoff frequency. Parasol cells did not change their RF size significantly. Large OFF cells RFs decreased in size. All cell types exhibited stronger surrounds under natural movies, resembling the whitening filters predicted by efficient coding. However, quantifying the effect of the filter adaptation on the stimulus power spectrum showed a significant contribution towards whitening only in ON parasol cells. The whitening effect emerged regardless of the training stimulus. These results suggest that while RGCs adapt to the spatial frequency content of the input, efficient coding can only partially account for this adaptation. Significance statementNatural scenes differ from artificial stimuli like white noise, in spatial frequency structure. How the retina adapts to these differences remains unclear. To explore this, we studied responses of four primate retinal ganglion cell types to artificial and natural stimuli. Our results show that some cell types, like midget cells, enlarge their receptive fields and surrounds under natural stimuli, while others, like parasol cells, only enhance surrounds. These changes align qualitatively with the efficient coding theory, which posits redundancy reduction. However, in three cell types, the enhanced surrounds did not significantly whiten responses to natural stimuli, contrary to efficient coding predictions. These findings challenge how fully efficient coding explains retinal adaptation, suggesting that other principles underlie the processing of visual inputs.

systems biology↗

bioRxiv2024

Convolutional neural network models of the primate retina reveal adaptation to natural stimulus statistics

Understanding the nonlinear encoding mechanisms of retinal ganglion cells (RGCs) in response to various visual stimuli presents a central challenge in neuroscience, driving the development of increasingly complex predictive models. Here, we systematically evaluate linear-nonlinear (LN) models - applying various regularization techniques - and convolutional neural networks (CNNs) of increasing depth, to predict RGC responses to white noise and natural movies. Our analysis includes publicly available datasets from marmoset and salamander retinas. We demonstrate that LN models, when equipped with appropriate inductive biases, can achieve robust predictive performance on neural responses to both white noise and natural movie stimuli. The optimal inductive biases vary substantially across datasets and stimulus types, indicating that the LN models performance is susceptible to these choices. This warrants care when using LN models as baselines: their performance is not fixed, and inappropriate design choices can lead to "unfair" comparisons. However, even in the optimal inductive bias scenario, CNNs consistently outperform LN models across conditions, confirming the advantage derived from their nonlinear representation capacity. Investigating cross-stimulus generalization, we observe that models trained on white noise generalize better to natural movies than vice versa. Notably, LN models exhibit a smaller performance gap between in-domain and cross-domain predictions compared to CNNs, suggesting that the nonlinear processing captured by CNNs is more stimulus-specific. Overall, this study provides valuable benchmarks and methodological insights for neuroscientists designing predictive models of retinal encoding.

neuroscience↗

bioRxiv2022

Digital twin reveals combinatorial code of non-linear computations in the mouse primary visual cortex

More than a dozen excitatory cell types have been identified in the mouse primary visual cortex (V1) based on transcriptomic, morphological and in vitro electrophysiological features. However, the functional landscape of excitatory neurons with respect to their responses to visual stimuli is currently unknown. Here, we combined large-scale two-photon imaging and deep learning neural predictive models to study the functional organization of mouse V1 using digital twins. Digital twins enable exhaustive in silico functional characterization providing a bar code summarizing the input-output function of each neuron. Clustering the bar codes revealed a continuum of function with around 30 modes. Each mode represented a group of neurons that exhibited a specific combination of stimulus selectivity and nonlinear response properties such as cross-orientation inhibition, size-contrast tuning and surround suppression. These non-linear properties were expressed independently spanning all possible combinations across the population. This combinatorial code provides the first large-scale, data-driven characterization of the functional organization of V1. This powerful approach based on digital twins is applicable to other brain areas and to complex non-linear systems beyond the brain.

neuroscience↗