Fully agree! Visual, vestibular and also proprioceptive sensory systems all have prominent velocity signals.

🧠📈 #neuroskyence

Paper was recently featured in J Neuroscience; author copy through OSF.
www.jneurosci.org/content/45/2...
osf.io/emskv/

With sky-less Matthias Albrecht, Markus Gruber, and Robert Peterka. Supported by AFF @uni-konstanz.de

Our data is from sway behavior during standing balance. We believe the mechanism is generally relevant for the internal reconstruction of body orientation and #selfmotion for #motorcontrol, #perception, and #navigation.

In a linear system, visual scene motion evokes body sway at just those frequencies contained in the scene motion. We predicted a distortion using simulations, where body sway is evoked at frequencies NOT contained in the scene motion. This non-intuitive prediction was confirmed in experiments.

This reflects a reduced reliance on vision, quantitatively reproduced by the RFM mechanism.
Furthermore, the RFM model - once tuned - also predicts sway in new sequences (Prediction).

A typical behavior is that subjects sway increases less than proportional when increasing the movement amplitude of the visual reference: the 'gain' (amplitude ratio between body sway and stimulus) is reduced with increasing stimulus amplitude (Model fit left-to-right).

We implemented this mechanism in a closed-loop balance control model and compared its behavior with the sway responses of 24 subjects viewing a moving visual screen in a virtual reality head-mounted display.

The magic comes in, when assuming that the RFM reconstruction is blocked at small velocities and corrects (= is added to) the visual self-motion signal when exceeding a certain velocity threshold.
Why magic? It explains human behavior related to context dependent changes in sensory contributions.

In our new paper, we present evidence that the CNS continuously estimates the movement of the visual reference by comparing visual to other sensory cues. This reconstruction of Reference Frame Motion (RFM) happens in velocity. Why?

The underlying question is a fundamental problem in neuroscience: how does the nervous system construct body orientation from multiple, potentially conflicting sensory inputs?
We might have a solution.

Interesting paper, thread and discussion!

Germany has this DEAL consortium, which currently is focused on promoting Open Access. Would be cool to push this towards non-profit journals only.

deal-konsortium.de/en/