Distributed Coding of Evidence Accumulation across the Mouse Brain Using Microcircuits with a Diversity of Timescales

Area of Science:

• Neuroscience
• Computational Neuroscience
• Decision Neuroscience

Background:

• Perceptual decision-making relies on accumulating noisy evidence.
• Understanding the neural basis of this evidence accumulation is crucial.

Purpose of the Study:

• To investigate neural correlates of evidence accumulation across brain areas in mice.
• To explore the diversity of neural mechanisms and timescales involved in evidence integration.

Main Methods:

• Utilized brain-wide electrophysiological recordings in mice.
• Analyzed neuronal firing rates for drift-diffusion model (DDM)-like activity.
• Investigated single and race accumulator models for neural mechanisms and integration timescales.

Main Results:

• DDM-like neurons, showing evidence-sensitive ramping firing rates, are widespread across the brain.
• Multiple evidence accumulation mechanisms (single and race) operate within and across brain areas.
• Integration timescales vary across brain areas hierarchically and within areas, linked to microcircuit properties like recurrent excitation.

Conclusions:

• Evidence accumulation in the brain is not monolithic but employs diverse integration mechanisms.
• Neural circuits integrate evidence over multiple, hierarchically organized timescales.
• Heterogeneity in integration timescales reflects variations in microcircuit parameters, such as recurrent excitation strength.