Understanding how the brain converts incoming sensory signals into a deliberate choice has long been a central puzzle in neuroscience — and one with real implications for conditions ranging from ADHD to OCD and addiction, where decision-making circuits go awry. New imaging work in mice offers a precise window into the temporal architecture of prefrontal activity during the moment of choice.
Using in vivo calcium imaging in male mice performing a touchscreen visual cue-location choice task, researchers mapped population-level activity in the medial prefrontal cortex (mPFC). The data revealed that mPFC neural populations maintain a stable, low-dimensional, and continuous internal trajectory across decision trials — a kind of reproducible temporal 'clock' that tracks progression through pre-stimulus, stimulus, and post-stimulus phases. A circular phase-decoding algorithm successfully extracted where in the trial the animal was from population dynamics alone. At the single-neuron level, a substantial fraction of mPFC cells showed reliable, time-locked firing patterns. Hierarchical clustering resolved these diverse response profiles into organized temporal groupings that collectively spanned the full task epoch. Critically, choice-relevant information could be decoded across these mPFC ensembles, while the primary somatosensory cortex produced near-chance decoding performance — isolating prefrontal circuitry as specifically encoding the decision variable.
This finding aligns with an emerging theoretical framework called 'sequential attractor dynamics,' in which prefrontal circuits progress through discrete activity states rather than holding static representations. It extends similar observations from working memory tasks into the domain of perceptual choice, suggesting a unifying computational principle. The study's primary caveat is that it uses male mice only, limiting generalizability given known sex differences in prefrontal circuit organization. The imaging approach also captures a subset of neurons and cannot resolve synaptic-level mechanisms. As basic neuroscience in an animal model, translation to human cognition remains indirect, though the circuit principles identified are broadly conserved. Methodologically, the phase-decoding approach is a notable contribution that could prove useful in parsing temporal structure in other cognitive paradigms.