Search

Although we can measure muscle activity and analyze their activation patterns, we understand little about how individual muscles affect the joint torque generated. It is known that they are controlled by circuits in the spinal cord, a system much less well understood than the cortex. Knowing the contribution of the muscles towards a joint torque would improve our understanding of human limb control. We present a novel framework to examine the control of biomechanics using physics simulations informed by electromyography (EMG) data. These signals drive a virtual musculoskeletal model in the Neurorobotics Platform (NRP), which we then use to evaluate resulting joint torques. We use our framework to analyze raw EMG data collected during an isometric knee extension study to identify synergies that drive a musculoskeletal lower limb model. The resulting knee torques are used as a reference for genetic algorithms (GA) to generate new simulated activation patterns. On the platform the GA finds solutions that generate torques matching those observed. Possible solutions include synergies that are similar to those extracted from the human study. In addition, the GA finds activation patterns that are different from the the biological ones while still producing the same knee torque. The NRP forms a highly modular integrated simulation platform allowing these in silico experiments. We argue that our framework allows for research of the neurobiomechanical control of muscles during tasks, which would otherwise not be possible., Comment: 15 pages, 10 figures

[Purpose] The purpose of the current study was to reveal the association between posture control and muscle activity by measuring the trunk and hip joint muscle activities in the upright and slump sitting positions in both the healthy participants and patients with recurrent lower back pain. [Participants and Methods] We recruited eleven patients of recurrent lower back pain and ten healthy participants. During the maintenance of the two types of posture, upright and slump, we collected the surface electromyography data. We assessed the following muscles: rectus abdominis, external oblique, thoracic erector spinae, lumbar erector spinae, internal oblique, lumbar multifidus, iliacus, serratus anterior, rectus femoris, tensor fascia latae, and gluteus maximus. We studied the differences in spinal-pelvic curvature and muscle activity between the upright and slump positions in each group. [Results] In the healthy group, comparison of the muscle activity in upright and slump positions for both the trunk (external oblique, internal oblique, lumbar erector spinae, and lumbar multifidus), and the hip muscles (iliacus and gluteus maximus) showed a significant decrease in activity in the slump position. In the group with recurrent lower back pain, although the external oblique, lumbar erector spinae and lumbar multifidus showed reduced activity in slump position, these values were smaller when compared to those in the healthy group. [Conclusion] This study aimed to clarify the relationship between posture (upright and slump) and the activity of the trunk and hip joint muscles in the healthy participants and the patients with recurrent lower back pain. The results indicated that postural changes caused by recurrent lower back pain significantly affected the activity of the muscles involved in controlling the posture.

There have been many advances in the field of reinforcement learning in continuous control problems. Usually, these approaches use deep learning with artificial neural networks for approximation of policies and value functions. In addition, there have been interesting advances in spiking neural networks, towards a more biologically plausible model of the neurons and the learning mechanisms. We present an approach to learn continuous muscle control of a multi joint arm. We use reinforcement learning for a target reaching task, which can be modeled as partially observable markov decision processes. We extend proximal policy optimization with a liquid state machine (LSM) for state representation to achieve better performance in the target reaching task. The results show that we are able to learn to control the arm after training the readout of the LSM with reinforcement learning. The input current encoding used for encoding the state is enough to have a good projection into a higher dimensional space of the LSM. The results also show that we are able to learn a linear readout, which is equivalent to a one-layer neural network to learn to control the arm. We show that there are clear benefits of training the readouts of a LSM with reinforcement learning. These results can lead to demonstrate the benefits of using a LSM as a drop-in state transformation in general.

EEG-neurofeedback (NFB) became a very popular method aimed at improving cognitive and behavioral performance. However, the EMG frequency spectrum overlies the higher EEG oscillations and the NFB trainings focusing on these frequencies is hindered by the problem of EMG load in the information fed back to the subjects. In such a complex signal, it is highly probable that the most controllable component will form the basis for operant conditioning. This might cause different effects in the case of various training protocols and therefore needs to be carefully assessed before designing training protocols and algorithms. In the current experiment a group of healthy adults (n = 14) was trained by professional trainers to up-regulate their beta1 (15–22 Hz) band for eight sessions. The control group (n = 18) underwent the same training regime but without rewards for increasing beta. In half of the participants trained to up-regulate beta1 band (n = 7) a systematic increase in tonic EMG activity was identified offline, implying that muscle activity became a foundation for reinforcement in the trainings. The remaining participants did not present any specific increase of the trained beta1 band amplitude. The training was perceived effective by both trainers and the trainees in all groups. These results indicate the necessity of proper control of muscle activity as a requirement for the genuine EEG-NFB training, especially in protocols that do not aim at the participants’ relaxation. The specificity of the information fed back to the participants should be of highest interest to all therapists and researchers, as it might irreversibly alter the results of the training.

The purpose of this work is to better understand how animals control locomotion. This knowledge can then be applied to neuromechanical design to produce more capable and adaptable robot locomotion. To test hypotheses about animal motor control, we model animals and their nervous systems with dynamical simulations, which we call synthetic nervous systems (SNS). However, one major challenge is picking parameter values that produce the intended dynamics. This paper presents a design process that solves this problem without the need for global optimization. We test this method by selecting parameter values for SimRoach2, a dynamical model of a cockroach. Each leg joint is actuated by an antagonistic pair of Hill muscles. A distributed SNS was designed based on pathways known to exist in insects, as well as hypothetical pathways that produced insect-like motion. Each joint’s controller was designed to function as a proportional-integral (PI) feedback loop and tuned with numerical optimization. Once tuned, SimRoach2 walks through a simulated environment, with several cockroach-like features. A model with such reliable low-level performance is necessary to investigate more sophisticated locomotion patterns in the future.