A mechanistically interpretable model of the retinal neural code for natural scenes with multiscale adaptive dynamics

The visual system processes stimuli over a wide range of spatiotemporal scales, with individual neurons receiving input from tens of thousands of neurons whose dynamics range from milliseconds to tens of seconds. This poses a challenge to create models that both accurately capture visual computations and are mechanistically interpretable. Here we present a model of salamander retinal ganglion cell spiking responses recorded with a multielectrode array that captures natural scene responses and slow adaptive dynamics. The model consists of a three-layer convolutional neural network (CNN) modified to include local recurrent synaptic dynamics taken from a linear-nonlinear-kinetic (LNK) model \cite{ozuysal2012linking}. We presented alternating natural scenes and uniform field white noise stimuli designed to engage slow contrast adaptation. To overcome difficulties fitting slow and fast dynamics together, we first optimized all fast spatiotemporal parameters, then separately optimized recurrent slow synaptic parameters. The resulting full model reproduces a wide range of retinal computations and is mechanistically interpretable, having internal units that correspond to retinal interneurons with biophysically modeled synapses. This model allows us to study the contribution of model units to any retinal computation, and examine how long-term adaptation changes the retinal neural code for natural scenes through selective adaptation of retinal pathways.

Molecular Mechanisms of Neurotransmitter Release Control Distinct Features of Sensory Coding Reliability

A key requirement for the repeated identification of a stimulus is a reliable neural representation each time it is encountered. Neural coding is often considered to rely on two major coding schemes: the firing rate of action potentials, known as rate coding, and the precise timing of action potentials, known as temporal coding. Synaptic transmission is the major mechanism of information transfer between neurons. While theoretical studies have examined the effects of neurotransmitter release probability on neural code reliability, it has not yet been addressed how different components of the release machinery affect coding of physiological stimuli in vivo. Here, we use the first synapse of the Drosophila olfactory system to show that the reliability of the neural code is sensitive to perturbations of specific presynaptic proteins controlling distinct stages of neurotransmitter release. Notably, the presynaptic manipulations decreased coding reliability of postsynaptic neurons only at high odor intensity. We further show that while the reduced temporal code reliability arises from monosynaptic effects, the reduced rate code reliability arises from circuit effects, which include the recruitment of inhibitory local neurons. Finally, we find that reducing neural coding reliability decreases behavioral reliability of olfactory stimulus classification.

Emergence of a compositional neural code for written words: Recycling of a convolutional neural network for reading

The visual word form area (VWFA) is a region of human inferotemporal cortex that emerges at a fixed location in the occipitotemporal cortex during reading acquisition and systematically responds to written words in literate individuals. According to the neuronal recycling hypothesis, this region arises through the repurposing, for letter recognition, of a subpart of the ventral visual pathway initially involved in face and object recognition. Furthermore, according to the biased connectivity hypothesis, its reproducible localization is due to preexisting connections from this subregion to areas involved in spoken-language processing. Here, we evaluate those hypotheses in an explicit computational model. We trained a deep convolutional neural network of the ventral visual pathway, first to categorize pictures and then to recognize written words invariantly for case, font, and size. We show that the model can account for many properties of the VWFA, particularly when a subset of units possesses a biased connectivity to word output units. The network develops a sparse, invariant representation of written words, based on a restricted set of reading-selective units. Their activation mimics several properties of the VWFA, and their lesioning causes a reading-specific deficit. The model predicts that, in literate brains, written words are encoded by a compositional neural code with neurons tuned either to individual letters and their ordinal position relative to word start or word ending or to pairs of letters (bigrams).

Decoding neurobiological spike trains using recurrent neural networks: a case study with electrophysiological auditory cortex recordings

Recent advancements in multielectrode methods and spike-sorting algorithms enable the in vivo recording of the activities of many neurons at a high temporal resolution. These datasets offer new opportunities in the investigation of the biological neural code, including the direct testing of specific coding hypotheses, but they also reveal the limitations of present decoder algorithms. Classical methods rely on a manual feature extraction step, resulting in a feature vector, like the firing rates of an ensemble of neurons. In this paper, we present a recurrent neural-network-based decoder and evaluate its performance on experimental and artificial datasets. The experimental datasets were obtained by recording the auditory cortical responses of rats exposed to sound stimuli, while the artificial datasets represent preset encoding schemes. The task of the decoder was to classify the action potential timeseries according to the corresponding sound stimuli. It is illustrated that, depending on the coding scheme, the performance of the recurrent-network-based decoder can exceed the performance of the classical methods. We also show how randomized copies of the training datasets can be used to reveal the role of candidate spike-train features. We conclude that artificial neural network decoders can be a useful alternative to classical population vector-based techniques in studies of the biological neural code.

Modelling the neural code in large populations of correlated neurons

Neurons respond selectively to stimuli, and thereby define a code that associates stimuli with population response patterns. Certain correlations within population responses (noise correlations) significantly impact the information content of the code, especially in large populations. Understanding the neural code thus necessitates response models that quantify the coding properties of modelled populations, while fitting large-scale neural recordings and capturing noise correlations. In this paper we propose a class of response model based on mixture models and exponential families. We show how to fit our models with expectation-maximization, and that they capture diverse variability and covariability in recordings of macaque primary visual cortex. We also show how they facilitate accurate Bayesian decoding, provide a closed-form expression for the Fisher information, and are compatible with theories of probabilistic population coding. Our framework could allow researchers to quantitatively validate the predictions of neural coding theories against both large-scale neural recordings and cognitive performance.

Distinct population code for movement kinematics and changes of ongoing movements in human subthalamic nucleus

The subthalamic nucleus (STN) is theorized to globally suppress movement through connections with downstream basal ganglia structures. Current theories are supported by increased STN activity when subjects withhold an uninitiated action plan, but a critical test of these theories requires studying STN responses when an ongoing action is replaced with an alternative. We perform this test in subjects with Parkinson's disease using an extended reaching task where the movement trajectory changes mid-action. We show that STN activity decreases during action switches, contrary to prevalent theories. Further, beta oscillations in the STN local field potential, which are associated with movement inhibition, do not show increased power or spiking entrainment during switches. We report an inhomogeneous population neural code in STN, with one sub-population encoding movement kinematics and direction and another encoding unexpected action switches. We suggest an elaborate neural code in STN that contributes to planning actions and changing the plans.