FABIO OTUBOHere’s an interesting fact to ponder while you’re standing still and waiting in a queue: without the constant activity of nerves and muscles, your body would collapse like a puppet released by its master. Standing still requires more than just the orders transmitted by the nervous system through electrical impulses to the muscles that normally remain rigid. If this were so, the human body’s balance would be identical to that of a broom handle: any disturbance—the slightest breeze or even breathing or a heartbeat—would result in a fall. Remaining upright on two legs would require the skill of a circus tightrope walker, who has to move back and forth to keep a plate on the end of a stick. A part of the central nervous system in the human body automatically commands the coordinated contraction and relaxation of the leg muscles, leaving the brain free to pay attention to its surroundings or daydream about something of interest.
“Although we are not aware of it, standing still is a constant challenge for the nervous system,” says André Fábio Kohn, a biomedical engineer at the University of São Paulo (USP). Kohn and his doctoral students have developed a new model to describe how a portion of the spinal cord—the tissue formed by neurons grouped within a canal running through the bones of the spine—coordinates the contraction and relaxation of muscles located below the knee. It is these muscles, which control ankle rotations, that prevent the body from falling forward or backward when standing still.
The model created by Kohn’s team shows that the spinal cord is powerful enough to receive electrical signals indicating muscle tension, process them and then send back commands to control this tension, with very little help from the brain. “Some people think the spinal cord is like an electric cable that connects to the brain, just a bundle passing through, but this idea is wrong. If the brain is the equivalent of a supercomputer, the spinal cord would be a very good computer.”
The muscles simulated by the Kohn team show the same pattern of electrical activity—a combination of continuous and intermittent signals—that neurophysiologists and biomedical engineers have observed in recent experiments on humans. A fast-acting muscle, the gastrocnemius, which, in addition to maintaining posture, helps in jumping and running, operates intermittently, in a more pulsed manner, activated one to two times per second. Yet a slower muscle, one more resistant to fatigue, like the soleus, tends to be activated almost continuously. “Some muscles respond continuously, while others do so intermittently,” says Dr. Júlia Greve, of the University of São Paulo School of Medicine, Institute of Orthopedics and Traumatology. Greve studies therapies that assist in the recovery of elderly patients or patients with neurodegenerative diseases who have difficulty performing movements and maintaining posture. “Control of the nervous system over the muscle sensitivity modeled by Kohn is an important function for the rehabilitation of these people.”
“When you’re standing still and leaning slightly forward, the calf muscles, the soleus and the gastrocnemius contract, while the front of the leg, the tibialis anterior muscle, relaxes,” says Greve. In contrast, the anterior musculature of the leg contracts and that of the calf relaxes against the tendency to fall back. “This synchronization is modulated in the same segment of the spinal cord; the signal that commands one muscle to contract makes the other relax.”
Greve notes that the control of these muscles is only part of the postural control system. To keep the body in a certain position, each segment of the spinal cord needs a copy of the ankle control circuit for the other muscles of the body. In addition, the spinal cord and the motor cortex, the brain region responsible for conscious movements, need to work together to integrate the information received from the nerves linked to the muscles with the information coming from sight, touch and the vestibular system of the inner ear, which provides the reference for where the head is in relation to the rest of the body. “Without this sense, we would fall, Greve says.
After the body has been standing still for awhile, the body begins to use other strategies to maintain balance. In addition to the ankle rotation, the hip begins to move and support of the weight now shifts between one leg and the other. “The human postural control system is a mechanism of extraordinary complexity,” says Daniel Boari, a biomechanics expert at the USP School of Physical Education and Sports. Boari says about 750 muscles control the more than 200 types of independent movements the human body is able to perform. “Each research group has a slightly different view of the neuromuscular mechanisms that act in these situations,” says Robert Peterka, a biomedical engineer at the Oregon Health and Science University in the United States.
Brazilian engineer Hermano Krebs, a researcher at the Massachusetts Institute of Technology (MIT), builds and uses robots to help physical therapy patients who have lost partial movement as a result of lesions in the nervous system. Robots work like automatic physical therapists, by correcting the patient’s movements. Krebs is working with the Kohn team on a project that, if successful, will allow the new computer model to be used to guide rehabilitation therapies. “To improve robotic rehabilitation, it is important to look at the problem from various points of view, with experiments and simulations,” he says.
“Being good at mathematics and computing is not enough to make these models; we need to study physiology and learn from experimental work, in order to improve our intuition about the problem,” says Kohn. He began researching the physiology of the nervous system while still an undergraduate in electrical engineering at the USP Polytechnic School (Poli/USP) in the late 1970s. The origin of his model to control an upright posture dates back to 1994, when he spent a year at a laboratory of the U.S. National Institutes of Health. There, he learned how to measure the electrical activity of nerves and muscles that was recorded by electrodes placed on the skins of volunteers to determine which circuits of neurons in the spinal cord process electrical signals.
These and other experiments revealed that neurons are not simply elements of the electrical circuits that work with regularity like clocks. They fire abrupt and random electrical signals, which are reflected in the movement of the body. Even when a soldier is trained to march with regular steps, there is a small variation in the length of these steps. But, paradoxically, the continuous and smooth movement of a muscle is based on the joint action of hundreds of neurons connected to the muscle fibers, which, firing randomly and in a slightly uncoordinated manner, smooth out each other’s action.
By combining data from their experiments with those of other researchers, Kohn and Rogério Cisi, who at the time was his doctoral student, created a computer model of the spinal cord and the neurons involved in muscle control in 2008. “This is the core of our new model,” says Kohn. In 2013, with two other PhD students, Leonardo Elias and Renato Watanabe, Kohn expanded on Cisi’s model by including detailed descriptions of the muscles responsible for maintaining ankle tonus. The model takes into account, for example, sense organs of the tendons and links between muscle fibers and neurons known as muscle spindles, which act as sensors and inform the nervous system of stretching and the force felt by the muscles.
“We’re aware of the model’s limitations,” says Kohn, who recognizes the simplified way it treats the elements of the motor system. The dendrites, the cell body and the axon of each neuron are represented by electrical circuits that include dynamic aspects of neuronal function, which enables a more realistic reproduction of actual neuronal activity. The complex intertwining of neurons and muscle cells is also reduced. But the most drastic simplification is the human body as a whole, represented by a bar fixed to the floor by a movable joint, which acts as the ankle. In this model, known as an inverted pendulum, the bar remains still by the compensatory action of the soleus, the gastrocnemius and tibialis anterior muscles. “It’s simplified, but not simple,” says Kohn when referring to the model, which includes the representation of thousands of neurons and one million connections (synapses) between them in 5,000 mathematical equations.
The simulations suggest that the information processing performed in the spinal cord can keep a person standing still for at least 30 seconds, with characteristics similar to those of healthy humans. According to the model, the upper portion of the central nervous system, which includes the brain, helps the activity of the spinal cord by sending a special electrical signal. “We roughly imitate how the central nervous system, particularly the spinal cord, tries to process the responses of the senses involved in a certain movement,” says Kohn.
“I believe that Kohn has the best model for representing the circuit between the spinal cord and muscles,” says Krebs, who plans to use this model in reverse. His robots accurately measure variations in the stability of the ankle of a person standing still—that stability changes after a Cerebral Vascular Accident (CVA) because the signals sent to the spinal cord decrease. “With a decreased signal, certain parts of the ankle stop responding, while others respond more actively,” says Krebs. “I want to do the reverse: place measurements of ankle rigidity in the model and use it to find out how the signal is sent from the brain to the spinal cord.”
Is it possible to use the model for robotic therapy or to design a prosthesis that improves the electrical signal from the brain of a patient who has had a CVA? Not yet, says Kohn. The biggest problem is that the model has many variables and, although it acts in a natural way, there is no complete understanding of how each part interacts with another. “Currently, clinical use is impossible,” admits Kohn. Krebs is more optimistic. “Every time I meet up with Kohn, his team is closer to that possibility.”
ELIAS, L.A. et al. Spinal mechanisms may provide a combination of intermittent and continuous control of human posture: predictions from a biologically based neuromusculoskeletal model. PLOS Computational Biology. V. 10, November 2014.