Invasive & Non-Invasive Brain-Machine Interfaces
Brain-Machine Interfaces have so much potential to be used in a variety of use-cases from neurorehabilitation to replacing or restoring useful functions. Regardless, the technologies that could allow these BMIs to perform these functions in a sustainable and safe manner while also taking into account ethics and long-term effects are yet to be fully developed and effectively tested.
BMIs can be classified into three main categories — invasive, non-invasive, and semi-invasive. In invasive procedures, the device is surgically implanted directly into the human brain. These procedures are typically extremely expensive and experimental, which can pose a number of risks.
On the other hand, non-invasive, which is considered to be the safest of the three, is generally very low-cost and affordable but with many limitations such as weaker brain signal reception because of obstruction from the human skull. In the middle of invasive and non-invasive comes semi-invasive devices, which are surgically implanted into the skull on top, not in, the brain of a human.
However, the most substantial problem in all three classes of brain-machine interfaces is the acquisition of signals. The brain is a highly complex nonlinear system in which the neural signals are extremely chaotic and nonstationary. Factors such as external noise, user fatigue, and varying concentration levels all play into this huge roadblock of BMI technology. In addition to these factors, the main one is movement — the slightest even, which can throw off data reception.
As technology progresses and new techniques are developed and perfected around solving existing problems, many new problems will undoubtedly arise. This is no different from modern-day brain-machine interfaces — it’s had significant advances since Richard Caton first discovered electrical signals in animal’s brains. Realistically, non-invasive BMIs are much closer to being implementable and available commercially in everyday life, but again, the thing is, they aren’t effective enough compared to high-bandwidth fully invasive interfaces.
With invasive technology, user acceptance to the neurosurgery and the actual implant is much lower than other BMI counterparts and procedures. Due to this major issue, invasive BMI technology has ways to go before it becomes a new normal in medical & commercial use.
BMI technology has significant potential for the coming years, and there is still much gray area to be explored which will require significant testing to ensure that these key problems in all three subclasses of BMIs are solved or worked around.
Resources:
