Brain Computer Interfaces: Solution to Our Mental Health Crisis?
Have you ever wanted to read someone’s mind? What about moving objects with only your thoughts? Well - it isn’t as far off into the future as you might think. Science and technology are slowly catching up to our imaginations!
A new emerging technology called brain computer interfaces could one day do that. Obviously, the objective of brain computer interfaces isn’t to satisfy your inner desire to gain telepathy or telekinesis. It’s the next big thing.
According to the World Health Organization, an estimated 3.8% of the population or 280 million people in the world suffers depression.
And according to the National Institute of Mental Health in 2013, neuropsychiatric disorders are the leading cause of disability in the U.S. with major depressive disorder being the most common.
Many patients with major depression today are not responsive to current treatments. BCIs could one day solve this.
In this article I will explain what exactly are brain computer interfaces and how they could be used to help people suffering neuropsychiatric disorders.
Let’s jump right into it!
What are BCIs + How do they work?
Brain-computer interfaces (BCIs) put simply are machines that translate electrical signals from the brain into outputs actions to a computer. In other words, BCIs is the bridge connecting desired action coming from you mind to outputs.
For BCIs to work there are a couple of steps:
- Collecting signals ⚡
- Interpreting the signals 🤔
- Making outputs from the signals 💻
First we must collect signals from a brain. To do this we need something to sense brain activity. This means we need some kind of brain imaging. Think of brain imaging as different ways to eavesdrop into the brain or visualize what’s happening in the brain. I will explain the different kinds of BCIs and their pros and cons.
What are Invasive vs Non-Invasive vs Semi-invasive BCIs?
Think of the brain as an orchestra 🎻. Each neuron is responsible for the overall music performance. Each orchestra members are like neurons and the stage is like the brain.
Non-invasive BCIs is like listening to the performance from outside of the concert hall. You can hear the music but it’s a bit muffled and quiet compared to being in the hall.
A few examples of non-invasive BCIs include:
- fMRI, PET, and NIRS, which measure changes in blood flow 🩸
- MEG, which measures the brain’s magnetic activity 🧲
- EEG, which measure the brain’s electrical activity ⚡
Invasive BCIs on the other hand is like enjoying a concert on stage. You can tell hear the sounds of individual musician on stage like collecting signals from single neurons. Invasive BCIs are connected directly to the brain and requires neurosurgery.
Semi-invasive BCIs is like listening to the performance from the concert seats. It’s has a clearer sound than listening from outside of the concert hall but not as clear as being on stage.
How do BCIs work?
Now that we have the signals, we need to make sense of them, which means we must interpret these signals.
In order to interpret the signals are a few sub steps:
- Preprocessing 🧹
- Feature Extraction 📈
Preprocessing cleans up the data. EEG data itself not clean. That is because of the noise and artifacts in the data. Think of it like dust in a picture that needs to be cleaned to get a clearer picture. We do preprocessing to isolate the data from disturbances like electrical signals from in your body (blinking, heart beat) or external signals from electrical devices.
Feature extraction and classification are the two subcategories of dimensionality reduction. The reason to why we need machine learning in signal processing is because brain data is complex, messy, and hard to understand. Feature extraction is the reduction of number of features by creating new features from the pre-existing ones. Feature extraction distinguishes important and meaningful signals to help computers more easily understand them.
After interpreting them, all that’s left is for the computer to decode the signals to outputs in an machine!
What are brainwaves and how could they be utilized?
Brainwaves are electrical voltages that are picked up when collecting brain signals. There are 5 distant brainwaves which are measured in cycles per second. Delta (1–3 Hz) being the slowest, which is present when your mind very calm like in sleep and Gamma(39–42 Hz) being the fastest brainwave, which is present when we are highly alert when we are awake.
One way brainwaves can be used is neurofeedback. Neurofeedback therapy is a way to help people with ADHD and depression without drugs. In short, neurofeedback helps people regulate their brainwaves by monitoring them and rewards patients when they present normal brainwave patterns. One problem area of neurofeedback is standard of “normal.” Because everyone’s brains are unique, there is room for skepticism. So far, neurofeedback therapy is being used but scientists aren’t if they are actually effective.
One example of a company that is taking advantage of neurofeedback is Muse. Their headsets monitor EEG data from our brains during meditation and provide feedback of different sounds through the app. Their goal is for meditators to better control their state of mind.
But what’s the future like for BCIs and mental health? I’ll explain this in the next section.
How can BCIs help regulate mood? What are Affective BCIs?
Affective BCIs are brain computer interfaces that can detect and influence your moods, feelings, and attitudes. So think of a machine that can read your emotions like fear, sadness, happiness, and anger.
So how do they work?
There are two kinds of affective BCIs:
- Affective BCIs that detect emotions 🕵️♀️
- Affective BCIs that influences emotions 🔁
BCIs that detect emotions are open-loop systems🔓 and BCIs that influences emotions are closed-loop systems🔒. Another name for an open-loop system is a non-feedback system. This means that the the output of the continuous system doesn’t influence the input of the system.
Most BCIs and machines today are open-loop systems. For example, printers and coffee machines are closed loop system. What make open-loop systems special is that the output influences the input.
Closed-loop affective BCI automatically simulate an electrical stimulation of a desired mood back to your brain after detecting your mood. Your brain is the plant, which is the object in which the output tries to control.
Closed-loop affective BCIs can possibly one day help people with emotional disorders and amyotrophic lateral sclerosis(ALS) patients express emotions.
Closed-loop stimulation have not been tested on patients with neuropsychiatric disorders yet, but closed-loop stimulation but it have on neurological disorders such as Parkinson’s disease (PD) and epilepsy with promising improvements.
But overall, Affective BCIs is still an emerging technology. It has a long way to go until it can be used.
But wait… wouldn’t there be ethical concerns??
Yes.
There are many concerns with a BCI that can control your mood. Like neurosurgery, they raise concerns of safety, ethics, and privacy. If someone could control your mood, they could also manipulate you and cloud judgment. Moreover, data collected from these BCIs are highly personal.
Moreover, there is a concern if they should even be used. Closed-loop affective BCI systems make people rely on machines for emotional regulation.
Overall, it would be quite a while before BCIs becomes a prominent tool to help those with mental health problems.
Works Cited:
Steinert, S., Friedrich, O. Wired Emotions: Ethical Issues of Affective Brain–Computer Interfaces. Sci Eng Ethics 26, 351–367 (2020). https://doi.org/10.1007/s11948-019-00087-2
S. Wu, X. Xu, L. Shu and B. Hu, "Estimation of valence of emotion using two frontal EEG channels," 2017 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), 2017, pp. 1127-1130, doi: 10.1109/BIBM.2017.8217815.
Shanechi, M.M. Brain–machine interfaces from motor to mood. Nat Neurosci 22, 1554–1564 (2019). https://doi.org/10.1038/s41593-019-0488-y
