Motor Loop Sim

Stimulation

Electrode over M1

Contralateral Ipsilateral Sham

Relative to the reaching (right) paw: contralateral is left M1.

3 × 3 mm surface plate · AP +1.5 · ML ±3.0 mm
biphasic, cathodic first

Amplitude (µA) Pulse width (µs) Frequency (Hz) Train (ms)

Train starts at each reach

Training

Sessions Session (ms) Reaches / session Reach (ms)

Spines are pruned and grown between sessions, from the activity just simulated.

Model

Scale Seed

Scale multiplies every population (1 ≈ 500 cells). Run time grows with it.

Spike raster

Session

Amber ticks: stimulus pulses. Shaded: reaches. Left hemisphere is contralateral to the reaching paw.

Electrode field (coronal plane, cathodic phase)

Recruited by each pulse

Chance that a pulse is followed by a spike within 3 ms, per cell type and hemisphere.

Firing rates

M1 → striatum → GPi/SNr ⊣ VL thalamus → M1 · M1 → STN → GPi/SNr

Membrane potential

Spines per cell across sessions

Formed and eliminated

Dendrite view

Cell Session

Grey: present before this session. Amber: formed in it. Blue rings: pruned in it.

Turnover by input source

Reaches to replay

Animal

From To

Outcomes

Simulated stimulation

tDCS Sham Pulsed

A setting of this simulation only; it says nothing about any animal's group.

Electrode side Polarity Current (µA) V/m per mA

3 × 3 mm skull electrode · AP +1.5 · ML ±3.0
field calibrated to Farahani et al. 2024

Trials and model

Frames before Frames after Scale Seed

Run overnight (maestro waits until you are away)

Replay run

no run

Learning across trials (left = contralateral hemisphere)

Dotted lines: session boundaries. Top: the animal's real success rate per session and the model's expectation V.

Trials

Trial raster

Amber: reach onset. Shaded: drive window. Blue: dopamine event.

Compare runs

tDCS effect on firing (same reaches)

Simulated tDCS Simulated sham

Pick two finished runs.

Reward credit (D1 synapses from success-tuned minus failure-tuned M1 cells)

Real outcomes Shuffled

Pick two finished runs.