https://ift.tt/3BWO2hz When technology from science fiction movies becomes reality Photo by Jesse Martini on Unsplash The connection ...
When technology from science fiction movies becomes reality
The connection of the human brain and computers (or machines in general) sounds like science fiction — like a technology from a utopian (or dystopian) future. However, the development of modern brain-computer interfaces (or BCIs) started almost 100 years ago, when Hans Berger discovered electrical activity of the human brain and measured these signals through a method that later became known as electroencephalography, or simply EEG. Nowadays, BCIs already have many different applications, but we are only at the beginning and might see some impressive advances in the near future.
Before getting to the current applications of BCIs and some speculation of their future uses, we will first introduce different approaches to “read the mind”, or more scientifically, to measure brain activity. We will finish with a discussion of ethical issues connected to BCIs.
Brain-Computer Interfaces
First things first: No, it is currently not possible to read people’s thoughts — at least not directly. However, what we can do, is measuring people’s brain activity. We can get information from the gathered data, such as reactions to events in the environment or intentions, like a planned movement. The analysis of the data depends on the type of brain activity that is collected.
Brain-computer interfaces can be classified into three categories: invasive, partially invasive and non-invasive BCIs. While invasive and partially invasive BCIs require some kind of surgery, we will focus on non-invasive BCIs because they are easier to set up, cheaper and have the obvious advantage that no surgery is required. (Although there are also stunning examples of invasive BCIs — like a monkey trained to play pong by Elon Musk’s Neuralink).
The most common approach to quantify brain activity is to measure electrical signals on the scalp that result from neural activity in the brain. This is called Electroencephalography or EEG and more specifically measures the voltage between electrodes and a reference.
Similar to EEG is the EOG (Electrooculography), which uses electrodes placed around the eye to measure the voltage between different points. Because of an electric potential between the front and the back of the eye, changes in the measured voltages can be used to calculate and track eye movements.
Other approaches to measure brain activity are the so-called fNIRS (functional near-infrared spectroscopy) and fMRI (functional magnetic resonance imaging) that measure haemodynamic activity and blood flow respectively, which both stem from neural activity.
The different technologies have different advantages and disadvantages. While fMRI has a high spatial resolution, recording a single sample takes several seconds and MRI scanners are (very) large and (very) expensive. In contrast, EEG, EOG and fNIRS devices have a lower spatial resolution (depending on the number of electrodes/optodes) and only measure the activity on the scalp, but have a higher sample frequency, are portable and much cheaper.
Most modern BCIs are based on EEG while also combinations of EEG and EOG or EEG and fNIRS can be found.
The bottom line is that neither of the above methods is measuring neural activity but some sort of (noisy) proxy such as the electric potential between electrodes, which are influenced by background noise from the environment.
Applications of BCIs
Although recording EEGs has a long-standing history, their efficient analysis is far from straightforward. A common approach to analyze EEG data is to investigate the power spectral density of certain frequency bands in specific brain areas. This however, only allows an interpretation of brain activity at a rather low-level. Less technically this mean that it is easier to measure the level of concentration but difficult (if not impossible with current state of the art EEG devices) to distinguish if someone is thinking about the colors red or blue. (The basics of EEG analysis, like event-related potentials, frequency bands or common spatial patterns will be the topic of a future article). This level of insights that we currently get from EEG recordings points into the direction of applications.
Applications in Medicine
One possible application of EEGs is based on the idea that we can measure changes in the state of consciousness in the human brain. This allows to monitor the level of sedation/anesthesia during surgical procedures.
Diagnoses of epileptic seizures, sleep disorders, Parkinson’s disease or other diseases are often based on the analysis of EEGs (either through experts or AI based systems).
Further, EEGs are frequently used in rehabilitation (especially in connection with VR or prothesis) where the aim is to restore motor functions via neuroplasticity.
Besides diagnosis and rehabilitation, the use of EEGs in medicine is manifold, and EEGs are even used to assess the functioning of the auditory system through auditory evoked potentials.
Other Applications
Outside of labs and clinics, brain-computer interfaces are becoming more popular but the applications of consumer-grade BCIs are currently rather limited.
Measuring the state of consciousness is not only relevant in medicine. There are BCIs available that are supposed to facilitate meditation by monitoring the concentration of their users. Similar systems might be of interest in fields such as transportation and education as drivers and students need to be focussed as well.
In the future, consumer-grade BCIs can become a game changer in various fields. In marketing, marketers will get direct reactions of costumers and can adapt their strategies and products accordingly. In connection with IoT devices, BCIs can create the next generation of smart environments — at home, at work and in transportation. In combination with VR, computer games will be more immersive and intuitive. BCIs have the potential to allow seamless authentication and change our understanding of security. Finally, BCIs will revolutionize the connection between humans and machines, more specifically cobots (collaborative robots).
On the downside, technologic developments might allow (military) use cases from dystopian novels.
Ethical Issues
Above we tackled specifically non-invasive one-way BCIs that record brain activity. However, in the mid to long-term we might see BCIs that allow signals to travel the other way around — from computers to human brains. At this point mind control becomes a pressing ethical issue to consider. Although it is neither clear what free will means nor if it even exists on a physical level, the concept needs some rethinking after emotions and movements of a human being are controlled by a computer. The brain, and subsequently the personality of BCI users might change through deep-brain stimulation.
But even before it is possible to go both ways, at some point it might be possible to measure brain activity from a distance, which effectively would allow mind reading and some serious violation of privacy. A famous German song says “Die Gedanken sind frei” (Thoughts are free) — let’s hope this remains true in the future.
Conclusion
Although we are far from widespread consumer-grade BCIs and there are some serious societal issues (as with every new technology), we might see some amazing, helpful and positive inventions in the near future that can improve many lives.
Brain-computer interfaces are an active field of research with breathtaking developments and serious opportunities and threats to society. We are at a crossroads and need to make sure that the technology is used for good (whatever this means to you).
Similar to AGI, I personally believe, that decentralized research and development are the key to avoid a misuse of BCIs. Some time after this realization, I changed career to develop brain-computer interfaces. What will you do?
Introduction to Brain-Computer Interfaces was originally published in Towards Data Science on Medium, where people are continuing the conversation by highlighting and responding to this story.
from Towards Data Science - Medium https://ift.tt/2Z5iCqp
via RiYo Analytics
No comments