Encoding of movement primitives and body posture through distributed proprioception in walking and climbing insects

Area of Science:

• Computational neuroscience
• Animal locomotion
• Robotics

Background:

• Animal limb coordination relies on movement primitives and body representation.
• Insects, lacking specialized posture organs, may integrate proprioceptive cues for postural information.
• Previous work detailed encoding of joint angles and velocities by insect leg afferents.

Purpose of the Study:

• To model how insects derive high-level movement and posture information from distributed proprioceptors.
• To investigate the role of spiking neural networks in processing low-level sensory input.
• To validate computational models against experimental data of stick insect locomotion.

Main Methods:

• Utilized a multi-layer spiking neural network architecture.
• Modeled second-order interneurons using coincidence detection of local leg inputs.
• Validated model performance against experimental kinematics of unrestrained stick insect locomotion.

Main Results:

• Modeled interneurons successfully signaled step cycle phases and transitions (e.g., leg lift-off).
• Third-order interneurons encoded body pitch relative to the substrate using leg joint data.
• Demonstrated that combinations of 2-3 proprioceptive inputs suffice for high-order movement encoding.

Conclusions:

• Simple combinations of local proprioceptive inputs can encode complex movement information.
• The proposed neural network model effectively decodes movement primitives and body posture.
• Findings suggest a plausible mechanism for insect spatial coordination and locomotor control.