TL;DR
This paper explores how undulatory locomotion in elongated fish is controlled by a combination of central pattern generators (CPGs) and multimodal sensory feedback. By developing a neuromechanical model and testing it with simulations and a robot, the researchers demonstrate that stretch and pressure feedback work synergistically to enable robust swimming, facilitate terrestrial crawling over obstacles, and explain the remarkable ability of eels to swim even after spinal cord transection.
What Problem Does This Paper Address?
Vertebrate locomotion, particularly undulatory swimming, is a complex behavior arising from the interplay between neural circuits (CPGs), sensory feedback, and the environment. While CPGs are known to generate basic rhythmic patterns, the precise role and interaction of different sensory feedback mechanisms (proprioception from stretch and exteroception from pressure) in generating stable locomotion, adapting to different environments, and maintaining function after injury remain unclear. Specifically, how these feedback loops contribute to both aquatic and terrestrial movement, and how they enable remarkable resilience to spinal cord damage, are key questions. This paper aims to bridge this gap by investigating the combined effects of stretch and pressure feedback on undulatory locomotion in elongated fish.
Methods
The researchers employed a synthetic approach combining neuromechanical modeling, simulations, and robotic experiments.
- Neuromechanical Model: A planar chain model of an elongated fish body was created, consisting of 10 rigid segments connected by joints. Each segment housed a local circuit with a phase oscillator (representing CPGs) and simulated muscle activation.
- Sensory Feedback: The model incorporated two key sensory feedback mechanisms:
- Stretch Feedback (Proprioception): This feedback loop was inspired by control principles for snake-like locomotion and uses the bending angle of anterior segments to influence the current segment's oscillator phase, promoting coordinated bending.
- Pressure Feedback (Exteroception): This feedback, adapted from previous work, uses local pressure forces on the body's sides to entrain the oscillator phase, with a saturation mechanism to prevent excessive responses.
- Simulations: The model was extensively simulated to test various combinations of control components: central phase coupling (C), pressure feedback (P), and stretch feedback (S).
- Robotic Experiments: A robot named AgnathaX, inspired by lamprey locomotion, was used to validate the simulation findings in both aquatic and terrestrial environments. This provided a crucial bridge between abstract models and real-world physics.
- Spinal Transection Experiments: To investigate robustness, the model and real eels were subjected to simulated and actual spinal cord transections, respectively, to observe the effects on locomotion.
Datasets
Not discussed in the paper. The study relies on simulated data generated by their model and experimental data collected from real eels and the AgnathaX robot.
Key Insights and Ideas
- Synergistic Sensory Feedback: Stretch and pressure feedback, when combined, significantly enhance swimming performance by contributing to rapid pattern generation and stability. This suggests a redundancy and complementarity in sensory systems.
- Stretch Feedback for Terrestrial Locomotion: While pressure feedback alone struggles with obstacle-aided crawling, stretch feedback is crucial for generating effective terrestrial locomotion by allowing the fish to exploit reaction forces from environmental irregularities (like pegs).
- Spontaneous Oscillation and Stretch Feedback as Keys to Robustness: The remarkable ability of eels to swim after spinal transection can be explained by two factors: the inherent ability of local CPGs to spontaneously oscillate even without descending drive, and the coordinating role of stretch feedback in maintaining rhythmic activity below the transection.
- Multimodal Feedback for Resilience: The combination of different sensory feedback mechanisms (C, P, S) provides resilience, allowing the system to adapt to significant changes in internal parameters (like intrinsic oscillator frequencies after transection) without compromising function.
Key Results
- Swimming: Combined stretch and pressure feedback (P+S) led to faster convergence to steady-state swimming and improved stride length compared to individual feedback types. However, higher stretch feedback gain increased the mechanical cost of transport.
- Terrestrial Crawling: Stretch feedback (S) alone enabled effective crawling over pegs, while pressure feedback (P) alone was largely ineffective. The combination (P+S) also showed high success rates.
- Spinal Transection: The model successfully replicated the observed phenomenon of eels swimming after spinal transection. Simulations showed that with spontaneous oscillation and sufficient stretch feedback, frequency-locked swimming could be maintained across the transection site, even with varying intrinsic frequencies in the posterior segments.
Limitations and Future Work
- Model Simplifications: The model uses abstract phase oscillators and simplified muscle models. Real biological systems are far more complex. The paper acknowledges that sim-to-real mismatches can occur, particularly in the success rates for robot experiments.
- Stretch Feedback Topology: While the chosen stretch feedback topology was effective, the paper notes that other topologies (e.g., posterior-to-anterior projection) could also contribute to swimming, suggesting further exploration is needed.
- Nonlocal Connections: The current model primarily uses local connections for phase coupling. The paper suggests that investigating longer-range neuronal connections could be an important future direction.
- Energetic Efficiency: The study found that higher stretch feedback gain increased mechanical cost of transport. Future work could explore integrating alternative stretch feedback mechanisms to improve energetic efficiency while maintaining robustness.
- Species-Specific Differences: The paper highlights that salamanders, unlike eels, require weeks to recover after spinal transection, suggesting a lack of spontaneous oscillation or weaker sensory feedback pathways. Further research could investigate these differences.
Summary updated