Internal Dynamics Interact with Proprioceptive Feedback During Movement Execution in an RNN Model of Motor Cortex

ABSTRACT The motor cortex initiates movement by sending temporal patterns to down-stream neural circuitry. The patterns during movement execution are thought to arise from the internal dynamics within motor cortical network. However, external inputs such as proprioception also shape motor cortical dynamics. To investigate the contribution of internal dynamics and proprioceptive feedback to voluntary movement execution, we constructed several motor cortex models receiving different combinations of proprioceptive feedback from a virtual arm to perform a delayed reach task. Considering the delay, noise and source of sensory feedback, we construct a sensory estimation network. We found that the inhibitory stabilized network receives hand kinematics and muscle force generated patterns most similar to those observed in motor cortex neuronal data. Furthermore, we used a disruption strategy to dissect the contribution of internal dynamics and proprioceptive feedback, and found that internal dynamics dominate, while proprioceptive feedback fine-tunes motor command. Analysis of the ablation experiment reveals that proprioceptive feedback improves the robustness against noisy initial conditions. Our results highlights that both intrinsic architecture and external input are critical to produce brain-like neural activity..

Medienart:

Preprint

Erscheinungsjahr:

2024

Erschienen:

2024

Enthalten in:

bioRxiv.org - (2024) vom: 22. Apr. Zur Gesamtaufnahme - year:2024

Sprache:

Englisch

Beteiligte Personen:

Jiang, Hongru [VerfasserIn]
Bu, Xiangdong [VerfasserIn]
Zheng, Zhiyan [VerfasserIn]
Pan, Xiaochuan [VerfasserIn]
Chen, Yao [VerfasserIn]

Links:

Volltext [kostenfrei]

Themen:

570
Biology

doi:

10.1101/2023.12.05.570033

funding:

Förderinstitution / Projekttitel:

PPN (Katalog-ID):

XBI04179656X