Curing Paralysis Part 3: Neural Bypass
We explored the potential therapeutic value of epidural stimulation and monoamine administration for paralysis patients in the first two parts of this series. These approaches aim to augment existing neural circuitry in hopes of restoring movement to paralized individuals. They tend to work best with simple movement patterns involving large muscle groups, such as walking or standing. However, if a patient has lost movement in regions requiring fine motor control, such as the hands, these approaches are much less efficacious. The hand has about 20 degrees of freedom, meaning there are many more muscles to stimulate, each of which are smaller than the major muscle groups in the legs. Because of this increased complexity and decreased accessibility, it is unlikely epidural stimulation or monoamine treatment will be able to restore hand movement . In cases like these, researchers have instead turned to a method called neural bypass. In this third segment of our series, we will take a closer look at neural bypass in the context of curing paralysis
Neural bypass is best illustrated by an example. A 2016 study performed by Chad Bouton at the Feinstein Institutes for Medical Research employed neural bypass to restore a patient’s ability to grasp, manipulate and release objects. They performed the bypass via an implant placed inside the patient’s motor cortex connected to a neuromuscular electrical stimulator (NMES). The NMES was a sleeve placed over the forearm with 130 electrodes. Electrical stimulation from the NMES would cause muscles in the arm and hand to contract. Researchers monitored this stimulation and consequential contraction with the hope of developing a machine learning algorithm that could translate motor cortex activity to intended action, thereby restoring movement in the arm and hand. In each training session, the patient would watch a hand on the screen and imagine doing the same movements as the hand while the implant would record his brain activity. Eventually, the patient had enough control to play Guitar Hero! This was the first study to demonstrate the successful translation of intracortical brain activity into functional movement.
While these results are incredibly exciting, the truth is that even if this form of neural bypass became clinically viable, it would still likely miss the mark due to a few limitations. One of the problems with the neural bypass technique described in the above study is the lack of a feedback loop. When an able bodied person grabs an object, they do not need their eyes to confirm that it has been grabbed, but rather they can feel that it has been grabbed. This is why humans can pick up delicate objects like eggs without cracking them. Touch feeds tactile sensory information back into the brain about the nature of the object. In the case of the egg, this additional input prevents us from gripping too hard and cracking it. Unfortunately, when someone loses movement in their hands, they often lose their sense of touch as well. But because neural bypass doesn’t leverage neural circuitry in the hands, patients won’t be able to modulate their grip properly.
Bouton’s later work attempts to address the challenge posed above. In addition to the NMES sleeve on the forearm, Bouton’s lab designed sensors on the hand to mimic haptic feedback. These sensors were connected to another invasive implant placed in the sensory cortex. Unfortunately, the addition of these sensors made this a very difficult control-systems problem. In the earlier study, the control system only went one-way. The patient wanted to grasp something, and the model had to adapt to allow that to happen. In the case of feedback control, not only does the machine learning model have to map the users intentions to movements, but it has to map the resulting movements to sensory feedback. This approach has shown great promise. A 2021 study done at the University of Pittsburgh found that a brain-computer interface implementing bidirectional haptic feedback led to a 50% improvement in grasping time because the subject had more sensory input when grabbing. However, it still took the subject an average of 10 seconds to grab an object, meaning complete feedback has yet to be achieved.
Whether or not this problem is one that can be solved remains to be determined. But, in the meantime their research offers a glimmer of hope for those who have lost their independence due to paralysis. If fine motor control in the hands can be restored, it means neural bypass can be transferred to less complex movement patterns, giving the hope of full movement restoration for people living with paralysis.
