A short introduction to Neural engineering, Neurotechnology & Brain-Computer Interfaces
Understand the basic principles underlying the most hyped emerging field of neuroscience: neurotechnology
Neural engineering and neurotechnology are interdisciplinary fields utilizing a vast series of scientific fields including neuroscience, biomedical signal processing, nanotechnology, electrical engineering, clinical medicine, computer science, genetics, and the list go on…
The goal is to discover and engineer new ways of understanding, repairing, or enhancing the properties of neural systems.
Currently one of the most essential applicational objectives of neurotechnology is the restoration of damaged brain functions through interactions between external devices and the nervous system.
Research is being conducted to develop an understanding of the coding and processing of information in the brain: How processing is altered in the pathological state and how these can be manipulated to regain a state of normal function. As we understand more about the underlying mechanisms, we can develop tools such as neuroprosthetics, brain-computer interfaces, or other forms of neural interfaces to obtain the necessary manipulation.
In the not so long future, we will also be witnessing developments in neurotechnology aimed at not only restoring normal brain functions but taking the step further to enhance the normal functioning brain of the average Joe.
Perhaps setting an all-together stop on the anti-vaxx movement…
As of 2021, we have witnessed exciting new developments, but the current scene of applicable and widely spread neurotechnologies can still be divided into only two main categories: (i) brain activity monitoring & (ii) brain activity stimulation (neuromodulation).
Brain activity monitoring
The most potent modalities for monitoring brain activity today are -from least to most invasive- MRI & fMRI scans, fNIRS, PET and CAT scans, magnetoencephalography (MEG), electroencephalography (EEG), intracranial-EEG (iEEG) / electrocorticography (ECoG), microelectrode (e.g. the Utah Array).
Clinical applications of these tools are vast in both patient monitoring and diagnostics. Also, the non-invasive modalities with high temporal resolutions, are used for brain-computer interfaces (BCI), such as spellers (typing by imagination) or environment control applications for patients that are suffering from Locked-In-Syndrome or otherwise brain-damaged (ALS, cerebral palsy, stroke, spinal cord injuries, etc.)
Brain activity stimulation
Stimulating the brain is magnificently more difficult than its activity monitoring sister. Partly due to being a comparatively uninvestigated field of study. Neurostimulation generally requires more invasive methods, to achieve specific results, although non-invasive neurostimulation methods exist. Transmagnetic stimulation (TMS) has shown promise in a variety of use-cases, from treating depression to brain-to-brain communication. Transcranial Direct-Current Stimulation (tDCS) is a portable and wearable form of neuromodulation that delivers a low electric current to the scalp. Currently, tDCS is still at an early stage, with no strong or conclusive evidence for cognitive enhancement or disease treatment. No tDCS treatments have been FDA approved yet.
As for invasive neuromodulation, there are specifically two types of implants that are widely spread and have shown incredible results among patients: cochlear implants that improve hearing (in some cases curing deafness!) and Deep Brain Stimulation (DBS) that among other things helps with tremor symptoms in Parkinson’s disease. DBS has over the past 25 years been approved for the treatment of dystonia, OCD, epilepsy, and clinical trials are now on the way investigating the potential of treating chronic pain for various affective disorders.
What is a brain-computer interface (BCI)?
A BCI converts thoughts into action with no means of muscular control. These actions can be for both medical and non-medical applications. Below I’ve put a visualization of the typical BCI setup & overview:
Brain signals to control BCI systems
The quality of brain signals for use in BCI systems, have two crucial variables beyond availability and invasiveness, mainly the temporal and spatial resolutions. A high temporal resolution allows for a well functioning feedback loop with a minimal delay that has high-speed real-time control, while a high spatial resolution makes it easier to create precise control and commands in the BCI system.
The challenges of building BCIs and neural interfaces
Beyond the obvious anatomical challenge of communicating with an organ that's hiding beneath the skull, the brain presents a series of challenges that makes it challenging to develop neural interfaces:
Processing depends on unknown parameters
· It’s person-specific
· It’s task-specific
· It’s otherwise variable
Reasons for variability
· Folding of brain cortex differs in individuals (even in monozygotic twins!)
· Relevant functional maps differ across individuals
· Sensor locations differ across recording sessions
· Brain dynamics are non-stationary at all time scales
Also, the signal-to-noise ratio is quite challenging, why sensitive measurements are hard to obtain. Useful brain activity recordings are small compared to the interfering artifacts and background brain activity.
Likewise, specific measures are even harder to obtain with coarse-grained sensing. Large collections of neurons are involved in many different activities. Not just a single one.
The underlying phenomena are also highly diverse and rich and derived measures are still poorly understood — it’s not always clear what to look for.
Furthermore, if working with EEG, the signals are computationally complicated to handle since all sensors record almost the same signal. (known as the superposition of all brain activity). Therefore they need to be disentangled (e.g. statistically) for optimal performance.
As a consequence = sophisticated signal processing is a must, all approaches are fundamentally statistical, BCI systems must be calibrated before use, and said calibration should entail as much information as available.
To summarize…
Neural engineering and neurotechnology apply the principles of engineering and neuroscience in developing tools that interact with the nervous system. This can help the development of solutions for problems related to brain injury and malfunction, or what specifically excites me: turning us into awesome transhumanistic cyborgs!
Thank you for reading. I will leave you with the following two videos. The first from 2012 by nature video, showing Cathy Hutchinson, a paralyzed patient, using her mind to control a robotic arm in order to take a sip of coffee.
The second, a more recent video, of Thomas Readon CEO of CTRL-labs, from the BrainMind Summit 2019 at MIT, on neural interfaces for human augmentation.
