CTMM ·
Rodent operant cognition · cross-species translation

The Bussey-Saksida touchscreen chamber

A trapezoidal operant box in which a rat or mouse answers cognitive problems by touching images on a screen, then collects a reward at the opposite end. Our human side runs on the Western University touchscreen platform (human touchscreen in our projects), developed by the same Bussey-Saksida group so that people and mice can perform the same tasks with the same stimuli, rules and outcome measures. This page shows how the rodent system works, how each task maps onto its human touchscreen version, and where that comparability holds or breaks.

1 · Apparatus

Inside the chamber

Top-down view of the touchscreen chamber sound-attenuating cubicle · fan · camera black acrylic walls IR touchscreen IR beams house light (ceiling) tone magazine + tray light pump / pellet dispenser

Select any part of the drawing, or use the buttons, to see what it does and why it matters for translation.

2 · Trial structure

Run a trial yourself

You play the animal. Start each trial at the magazine, answer at the screen, then collect the reward. The program logs the same variables the real software records. Timeout and inter-trial intervals are shortened here so the demo stays brisk.

House light
TONE magazine

Session record

    Human mirror

    3 · Shaping

    Getting an animal to the screen

    Animals are mildly food-restricted (typically to about 85% of free-feeding weight) and pass through a fixed pretraining sequence before any test begins. These stages run in order; each has a performance criterion.

    In humans

    The same competence takes a verbal instruction and a few practice trials. That asymmetry (weeks of shaping versus minutes of instruction) is the largest procedural difference between the species and should be kept in mind when interpreting acquisition measures such as trials to criterion.

    4 · Task battery

    Rodent task and its human touchscreen version

    Human versions follow the Western identical-task approach. Which of these tasks are enabled in the human touchscreen build used in our cohorts still has to be confirmed task by task.

    5 · Comparability

    How far does the parallel go?

    Comparability is a property of each dimension, not of the platform as a whole. The table rates each one by how closely the rodent procedure can match the human test.

    DimensionMatchRodentHuman
    Synthesis

    A shared measurement language, not identical cognition

      Sources

      Key references

      Confidence reflects how certain the citation details and the claim attributed to each source are; verify medium-confidence items before citing.

        Horner 2013 · Mar 2013 · Oomen 2013 · Nithianantharajah 2015 · Heath 2019 · Martins 2025