Modeling responses of macaque and human retinal ganglion cells to natural images using a convolutional neural network.

Linear-nonlinear (LN) cascade models provide a simple way to capture retinal ganglion cell (RGC) responses to artificial stimuli such as white noise, but their ability to model responses to natural images is limited. Recently, convolutional neural network (CNN) models have been shown to produce light response predictions that were substantially more accurate than those of a LN model. However, this modeling approach has not yet been applied to responses of macaque or human RGCs to natural images. Here, we train and test a CNN model on responses to natural images of the four numerically dominant RGC types in the macaque and human retina - ON parasol, OFF parasol, ON midget and OFF midget cells. Compared with the LN model, the CNN model provided substantially more accurate response predictions. Linear reconstructions of the visual stimulus were more accurate for CNN compared to LN model-generated responses, relative to reconstructions obtained from the recorded data. These findings demonstrate the effectiveness of a CNN model in capturing light responses of major RGC types in the macaque and human retinas in natural conditions.

Decomposition of retinal ganglion cell electrical images for cell type and functional inference.

Identifying neuronal cell types and their biophysical properties based on their extracellular electrical features is a major challenge for experimental neuroscience and the development of high-resolution brain-machine interfaces. One example is identification of retinal ganglion cell (RGC) types and their visual response properties, which is fundamental for developing future electronic implants that can restore vision. The electrical image (EI) of a RGC, or the mean spatio-temporal voltage footprint of its recorded spikes on a high-density electrode array, contains substantial information about its anatomical, morphological, and functional properties. However, the analysis of these properties is complex because of the high-dimensional nature of the EI. We present a novel optimization-based algorithm to decompose electrical image into a low-dimensional, biophysically-based representation: the temporally-shifted superposition of three learned basis waveforms corresponding to spike waveforms produced in the somatic, dendritic and axonal cellular compartments. Large-scale multi-electrode recordings from the macaque retina were used to test the effectiveness of the decomposition. The decomposition accurately localized the somatic and dendritic compartments of the cell. The imputed dendritic fields of RGCs correctly predicted the location and shape of their visual receptive fields. The inferred waveform amplitudes and shapes accurately identified the four major primate RGC types (ON and OFF midget and parasol cells), a substantial advance. Together, these findings may contribute to more accurate inference of RGC types and their original light responses in the degenerated retina, with possible implications for other electrical imaging applications.

Unified classification of mouse retinal ganglion cells using function, morphology, and gene expression.

Classification and characterization of neuronal types are critical for understanding their function and dysfunction. Neuronal classification schemes typically rely on measurements of electrophysiological, morphological, and molecular features, but aligning such datasets has been challenging. Here, we present a unified classification of mouse retinal ganglion cells (RGCs), the sole retinal output neurons. We use visually evoked responses to classify 1,859 mouse RGCs into 42 types. We also obtain morphological or transcriptomic data from subsets and use these measurements to align the functional classification to publicly available morphological and transcriptomic datasets. We create an online database that allows users to browse or download the data and to classify RGCs from their light responses using a machine learning algorithm. This work provides a resource for studies of RGCs, their upstream circuits in the retina, and their projections in the brain, and establishes a framework for future efforts in neuronal classification and open data distribution.

An offset ON-OFF receptive field is created by gap junctions between distinct types of retinal ganglion cells.

In the vertebrate retina, the location of a neuron's receptive field in visual space closely corresponds to the physical location of synaptic input onto its dendrites, a relationship called the retinotopic map. We report the discovery of a systematic spatial offset between the ON and OFF receptive subfields in F-mini-ON retinal ganglion cells (RGCs). Surprisingly, this property does not come from spatially offset ON and OFF layer dendrites, but instead arises from a network of electrical synapses via gap junctions to RGCs of a different type, the F-mini-OFF. We show that the asymmetric morphology and connectivity of these RGCs can explain their receptive field offset, and we use a multicell model to explore the effects of receptive field offset on the precision of edge-location representation in a population. This RGC network forms a new electrical channel combining the ON and OFF feedforward pathways within the output layer of the retina.

Premotor and Motor Cortices Encode Reward.

Rewards associated with actions are critical for motivation and learning about the consequences of one's actions on the world. The motor cortices are involved in planning and executing movements, but it is unclear whether they encode reward over and above limb kinematics and dynamics. Here, we report a categorical reward signal in dorsal premotor (PMd) and primary motor (M1) neurons that corresponds to an increase in firing rates when a trial was not rewarded regardless of whether or not a reward was expected. We show that this signal is unrelated to error magnitude, reward prediction error, or other task confounds such as reward consumption, return reach plan, or kinematic differences across rewarded and unrewarded trials. The availability of reward information in motor cortex is crucial for theories of reward-based learning and motivational influences on actions.