From science fiction to reality: How far are we from a world where “what you wish for can come true”? (Part 2)

Produced by: Science Popularization China

Author: Qian Yu (Center for Excellence in Brain Science and Intelligence Technology, Chinese Academy of Sciences)

Producer: China Science Expo

You must have seen such magic performances, where magicians use so-called "telekinesis" to bend forks and other objects at a certain distance, take objects from a distance, etc. These are all appearances that magicians have carefully designed for us to see. So in reality, can we manipulate objects through "telekinesis" without using our own limbs?

The answer is yes. Brain-computer interface technology can help us directly use the brain's "telepathy" (brain electrical signals) to control external single objects, robotic arms, or even exoskeletons, thereby achieving the purpose of manipulating the corresponding objects.

So how can we easily enable our brain to have the ability of "telekinesis"?

Two ways to achieve "telekinesis"

In order to realize "telepathy" through brain-computer interface technology, it is necessary to collect the neuronal electrical signals of "what your brain wants to do", parse out the information encoded in these signals, and transmit them to external objects, so as to directly control the brain to receive information and perform operations such as movement.

Current brain-computer interface technology can be divided into two categories: invasive brain-computer interface and non-invasive brain-computer interface .

As we all know, our brain is protected by a hard skull. Non-invasive brain-computer interface is a type of interface that can collect activity signals of brain neurons without destroying the skull structure or penetrating deep into the brain tissue. Generally, a head-mounted device like a hat is used to collect signals.

The most common is scalp electroencephalogram (EEG), a non-invasive way to detect brain electrical activity. Electrodes are placed on the scalp to record the relatively slow post-synaptic potentials generated by the current in and around neurons. EEG can detect brain neuron activity directly and in real time without invasive means.

However, since it is only placed outside the skull, the neural activity that generates EEG is transmitted throughout the brain, so the signal accuracy is relatively poor, and it is difficult to determine the relative position of the source of the electrode signal; and the brain is a three-dimensional structure, so it is difficult to record signals from deep brain regions.

Schematic diagram of the EEG device

(Image source: Reference [1])

An invasive brain -computer interface requires partial destruction of the skull structure and the insertion of signal-collecting electrodes through the skull into the soft brain tissue to obtain EEG signals. A semi-invasive brain-computer interface , such as electrocorticography (ECoG), passes through the skull but does not insert the electrodes into the brain tissue but only sticks them on its surface.

Different types of brain-computer interfaces

(Image source: Reference [2])

The benefit of an invasive brain-computer interface is that it can record the brain's electrical signals more clearly and accurately , even down to the level of individual neurons. This is necessary for successfully executing commands after decoding, and thus mastering "telepathy". It can also perform precise electrical stimulation to produce specific sensations.

However, because invasive brain-computer interfaces require the removal of a small piece of skull, this "brain-opening" approach is daunting.

Is there a way to easily master "telekinesis" with less or no harm? This is a problem that the field of brain-computer interfaces is working hard to overcome, and some progress has been made.

How to learn "telekinesis" at a low cost and easily

The mosquito mouthpart bionic flexible electrode demonstrated at the recent World Artificial Intelligence Conference can achieve minimally invasive implantation. The structure of this electrode is similar to the unique mouthparts of mosquitoes, which are hard on the outside but soft on the inside. When invading brain tissue, it is like a mosquito sucking blood and invading the skin. It minimizes damage to the brain while maintaining high signal accuracy, but this method still requires opening a small piece of the skull.

Mosquito mouthpart-like bionic flexible neural probe

(Image source: Chinese Academy of Sciences)

In addition, is it possible to implant the electrodes of the brain-computer interface through injection, just like in "Inception"?

After all, although we are afraid of craniotomy, we are not afraid of injections.

A research team from Australia has developed a minimally invasive brain-computer device called "Stentrode". It is a vascular stent-like brain-computer device that is delivered into the brain blood vessels from the vein at the base of the neck, and 16 electrodes are placed in the blood vessels next to specific brain areas to collect and translate the brain's neural activity.

Stentrode

(Image source: Neuronews)

Similarly, the Stanford University team also developed an ultraflexible micro-endovascular (MEV) probe, which was precisely delivered to the sub-hundred-micron brain blood vessels of rats through the neck vessels, thereby achieving long-term and stable recording of electrical signals in the cortex and olfactory bulb.

Ultra-flexible microvascular probe

(Image source: Reference [4])

These technologies that record brain electrical signals through blood vessels to master "telepathy" can be easily recorded without surgery on the skull. However, the disadvantages of these technologies are the same as other non-invasive technologies. The signal flux and fidelity are not very ideal, and they may only achieve a very general level of "telepathy".

It is hoped that in the future minimally invasive or even non-invasive brain-computer interface implants can be achieved, and the signal-to-noise ratio of the brain's electrical signals can be enhanced, making the way of obtaining "telepathy" simpler and safer.

The future of brain-computer interfaces

What does the future hold for brain-computer interfaces?

It is definitely moving towards the goals of non-invasiveness, high quality (high spatiotemporal resolution), and interactivity (i.e., the brain and brain-computer interface can interact). However, before achieving these goals, a lot of scientific research input and output are still needed.

Just like in "Avatar", lying in an airtight cabin and wearing a device, one can control another body to perform various activities. This should be possible in the foreseeable future. After all, the collection and decoding of motion signals is the most common technology in brain-computer interfaces.

Brain-computer interface remote control

(Photo source: Movie "Avatar")

However, to realize brain-computer interface technology that can completely control human thinking as in science fiction novels, it is still science fiction based on the current level of technology.

Although brain-computer interfaces can currently only achieve some simple "telepathy", this has brought a lot of good news to the rehabilitation of patients with brain damage and given many disabled people the possibility of taking care of themselves again.

It is foreseeable that in the near future, brain-computer interfaces will likely completely change the current situation of people with disabilities, allowing many paralyzed patients to recover more fully. Hopefully, by then, every disabled person who has suffered irreversible physical damage will be able to live a beautiful and dignified life like normal people.

Perhaps one day, brain-computer interface technology will allow everyone who wants to learn "telekinesis" to use it as they please without risk, making operations such as telekinesis no longer a "magic trick."

References:

[1] Sebastian Nagel, Towards a home-use BCI: fast asynchronous control and robust non-control state detection. SEBASTIAN NAGEL

in Gelnhausen, 2019

[2]Shujhat KhanTipu Aziz, Transcending the brain: is there a cost to hacking the nervous system? Brain Communications, 2019

[3]Tang J, LeBel A, Jain S, et al. Semantic reconstruction of continuous language from non-invasive brain recordings[J]. Nature Neuroscience, 2023: 1-9.

[4] Anqi Zhang, Charles M. Lieber, et al. Ultraflexible endovascular probes for brain recording through micrometer-scale vasculature. Science 381, 306–312, 2023.