Fig. 4: Visualizing neural activity across brain areas and species.
From: Rastermap: a discovery method for neural population recordings

a, Left: Rastermap sorting of neural activity collected from the CA1 region of rat hippocampus using two multi-shank silicon probes, during which the rat ran back and forth along a 1.6-m linear track44. Neuron identity was defined by spike waveform shape: FS, fast spiking (putative interneuron); RS, regular spiking (putative pyramidal cell). Right: tuning curves of each neuron to the position of the rat in the track. b, Left: Rastermap sorting of neural activity collected from the whole brain of a paralyzed zebrafish using light-sheet imaging at a rate of 2.1 Hz22. This period of time included two visual stimuli: phototactic stimuli (one side of the screen is dark) and optomotor response stimuli (moving gratings). Middle: color bar sectioning Rastermap into 18 groups for visualizing spatial patterning. Right: neuron positions in each group plotted in color; all neurons in the recording are in gray. Positions in each plot were collapsed across the z dimension. The color bar denotes the position along the embedding. c, (i) Rastermap sorting of cortical activity collected by wide-field imaging in mice performing a decision-making task48. The voxels in the brain image are colored by Rastermap position. Stimulus events in the task are indicated by colored shaded regions, and reward times are indicated by green lines. (ii) Linear prediction of activity from task and behavior variables shown with the same sorting as in (i). (iii) Same as (ii) with linear prediction from behavior variables only. (iv) Difference between prediction in (ii) and (iii). Eye pos., eye position; aud, auditory; vis, visual; var. exp., variance explained; diff. in var. exp., difference in variance explained.