A Cerebellar Neuroprosthetic System

Popular (secondary) literature piece:

Neuroprosthetics: wearable tech inside the brain


Even though it may sound like something out of a dystopian book or movie plot, brain implants, also known as neuroprostheses, can facilitate miracles. In this alternative system, an external device obtains sensory information which can no longer be accessed naturally and converts it into electrical signals. These are received by the implant, which then passes the information on to the brain. Said implants can be attached either to a nerve (commonly auditory or optic) or directly to the cortex.

Cochlear implants are the most widely used neuroprostheses today, aiding nearly 300,000 people with ear malfunctions. Retinal implants are far less common. In fact, only one is currently available on the market to help people suffering from retinitis pigmentosa. Another application of neuroprosthetics is even rarer than retinal implants, though it has a very promising future. Researchers are working towards restoring lost learning functions in humans through implants placed directly on the brain.

A group of researchers led by the SPECS group at Pompeu Fabra University, Barcelona conducted a study which demonstrated how a chip implanted into a living rat’s brain can restore a disabled function of the cerebellum. The cerebellum is an area of the brain which is greatly responsible for the attainment of motor memories. In the experiment, this area was anaesthetized and the rat was conditioned to develop an eye-blinking response.

In short, although there are a lot of obstacles to overcome before brain implants are considered safe and are accepted in the society, it appears that they will play an invariably large role in our future.

Primary literature article:

A cerebrellar neuroprosthetic system: computational architecture and in vivo test


The research done in this study demonstrates the functional pairing of an artificial system and the central nervous system in the context of classical conditioning, which is one form of associative learning. This was implemented by interfacing a prosthetic cerebellum with a rat brain. The neuroprosthesis was trained to associate a tone to a puff of air, and to subsequently trigger a closure of the rat eyelid.

The cerebellum is critical to creating a response in such conditioning. The signal from the conditioned stimulus (in this case, the tone) reaches the cerebellum through the fibers originating in the Pontine Nuclei, while the signal from the unconditioned stimulus (the puff of air) is projected through the fibers from the Inferior Olive. These two signals converge onto the cerebellar Purkinje cells that control deep nuclear neurons, which in turn synapse with the motor neurons responsible for producing conditioned responses.

However, motor conditioned responses are not expressed naturally when the cerebellum is anaesthetized. This proves that the observed anticipatory blinks from the rat were conducted by the neuroprosthesis.

In addition, it was shown that the acquired conditioned responses could be abolished by extinction training, as demonstrated by Pavlovian classical conditioning. This also indicates that the response of a learning process is induced by the contingent correlation of the two distinct stimuli.

Interestingly enough, a repetition of the initial training after extinction resulted in the same response being more rapidly acquired: a phenomenon coined as “savings.”

Of course, issues were faced in the execution of this experiment, some of which included the stochastic nature of biosignals as well as a learning stability problem. Despite these, this study demonstrates the possibility of efficient and robust neuroprostheses which would aid the acquisition, preservation, and extinction of behaviors even after the biological system has lost its learning capability.


I think that the popular piece was very accurate in reflecting the layout and the success of the conducted scientific experiment. Of course, the secondary article was extremely concise and ignored more complex aspects of the study. The 30+ pages of the primary source were reduced into one paragraph because this science blog focused on what the researchers did rather than how they did so, what their results were, and what implications those results had. In addition, the blog writers summarized all of the issues faced in the study by categorizing them as a “whole bunch of bioengineering obstacles that need to be addressed”. While this is not false nor is it under-exaggerating per se, I think that the message would be put in a better perspective by explaining at least a few of the obstacles faced in terms that are easy to understand for the general audience.

3 responses to “A Cerebellar Neuroprosthetic System”

  1. Carlos Aizenman says:

    I often find the writing in science blogs to be less sensationalistic and more balanced than news media, although as you point out, sometimes they may reduce the complexity too much.

  2. Isabella DeCarlo says:

    I agree that the secondary source gets straight to the point about the main idea without explaining the methodology and complexity actually involved. Very interesting articles, though!

  3. Aroosa Cheema says:

    Your articles were very interesting! Also, I have to agree that the secondary source seems to oversimplify the actual challenges faced by the researchers in order to appeal to their audience.

Leave a Reply

Your email address will not be published. Required fields are marked *