Summary
A bioRxiv preprint reports that the timing of hippocampal neurons’ spikes relative to theta rhythms carried information about locations and viewed images, including in neurons without corresponding firing-rate changes. In human navigation data, spike timing improved location decoding beyond firing rates alone.
A bioRxiv preprint reports that hippocampal neurons carried information about locations and viewed images through the timing of their spikes relative to theta rhythms—even when their firing rates did not change with those locations or images. In human navigation data, this timing information improved location decoding beyond what could be decoded from firing rates alone.
The researchers analysed recordings from humans navigating a virtual city, rats running on a linear track, and humans viewing images of animals, faces and places. The results suggest that measuring how often a neuron fires can miss information present in when it fires.
Timing adds a second view of neural activity
Neuroscientists often identify place cells by an increase in firing rate at a particular location. The preprint examines neurons that do not show that rate increase, asking whether their spike timing still varies systematically with location. The hippocampus produces theta rhythms: recurring patterns of neural activity that can provide a timing reference for spikes within each cycle.
In both navigation datasets, more neurons than expected by chance fired at a consistent theta phase at particular locations. This pattern remained when the researchers excluded neurons whose firing rates depended on location. In the human navigation data, spike phases from the remaining neurons improved location decoding beyond the firing rates of all neurons.
The authors also report that some phase-tuned neurons fired at progressively earlier points in the theta cycle as participants passed their preferred locations. This resembles phase precession, a pattern classically associated with place cells as an animal moves through a location where its firing rate rises. Here, the phase shift was associated with location even though the neurons had no corresponding firing-rate field.
Image responses also appeared in spike phase
In the image-viewing data, more neurons than expected by chance changed the theta phase of their spikes after particular images appeared. The effect remained when images that altered each neuron’s firing rate were removed from the analysis. Spike phases alone also predicted image category above chance.
Together, the navigation and image results point to a form of neural coding that complements firing rate: information can be reflected in a spike’s timing within a rhythmic cycle, not only in the number of spikes. The authors propose that weaker input may shift firing time, while stronger input can also raise firing rate. That explanation is the researchers’ interpretation of the findings, rather than a mechanism established by the abstract’s reported analyses.
The work is a preprint, and its abstract does not provide participant or neuron counts, effect sizes, or detailed recording and decoding methods. Its reported outcomes concern neural activity and information decoding in specific tasks; they do not directly measure memory performance.