Brain-Computer

Interfaces

1

Brain-Computer Interfaces

Introduction

The human brain is a marvel of evolution, a ‘machine’ that enables us to think, reason,

imagine, and perceive the millions of stimuli we receive every day. The brain accomplishes

all of this through connections between neurons, forming neural circuits.

Interestingly, Neural Networks, a type of machine learning algorithm, learn to perform

tasks by analyzing examples and adjusting the connections between artificial ‘neurons’.

Neural networks, which serve as the basis for Artificial Intelligence, stem from a part of the

brain called the visual cortex, which uses similar layers of neurons to decode more

complex levels of visual information (Yamins & DiCarlo, 2016, p. 358).

Brain Computer Interfaces (BCIs) work to integrate the fields of Neuroscience and

Computing, as these devices analyze brain signals and convert them into digital computer

data, blurring the boundaries between traditional computers and our biological

computer— the brain.

BCIs have four principal components of function – Signal Acquisition (Acquiring electrical

action potentials from neurons), Feature Extraction (Converting these brain signals into

digital data), Feature Translation (Analyzing the signals and converting them into

appropriate commands for the downstream device) and Device Output (Shih et al., 2012,

p. 270).

BCIs can augment many activities, including the restoration of function to those who have

been disabled by neuromuscular disorders like ALS, stroke, among others.

Additionally, a recent paper (Tang et al., 2023) on BCIs explores the role of flexible

electronics in these interfaces, claiming that flexible electronics may address some of the

problems that conventional rigid electronic BCIs face, such as a gradual decrease in signal

intensity and the risk of neurodegeneration.

Furthermore, with the advent of advanced AI systems, bi-directional BCIs may become

possible, which could have far-reaching impacts on human intelligence and memory. An

example of this would be for a BCI to “burn [a grocery list] into short-term memory, so that

we remember the milk, the eggs and the apples, making it the holy grail two-way interface”

(Webster, 2025, p. 1046).

Brain-Computer

Interfaces

2

How BCIs Work

BCIs collect signals from the brain (i.e., action potentials from specific neurons) through

electrodes – either non-invasively fitted on the scalp or invasive electrodes residing in the

brain.

Most commonly, signals used to control a BCI are electrical signals produced by

postsynaptic changes in membrane polarity (action potentials) through the opening of

voltage or ligand-gated ion channels in neurons.

Our brains produce sensorimotor rhythms, known as the mu and beta rhythms, which are

associated with movement (Shih et al., 2012, p. 269). BCIs capture signals of these

rhythms from the brain and help the user control devices. Electrical devices, such as small

intracortical microarrays, can be implanted in the cortex. These devices can then record

the action potentials of individual neurons (making it like a micro-EEG) and transmit these

signals for downstream processing. (Shih et al., 2012, p. 269)

Other methods include functional magnetic resonance imaging (fMRI) which measures

blood oxygenation in the cerebral region and correlates that with neural activity – these

signals are then processed and used to control devices.

As far as applications are concerned, researchers have used BCI systems to help

individuals control computer cursors, prosthetic limbs, wheelchairs, and other devices.

Limitations of current BCIs and Flexible Electronics

As BCIs are implanted devices, the body and brain trigger an immune response, leading to

the formation of astrocytes and microglia around the device. Over time, this response can

hamper the signal received by the device.

Additionally, rigid electronics do not account for the volume changes in the brain as it

develops and morphs over time, which can cause unstable or incorrect signaling from the

implanted devices.

Brain-Computer

Interfaces

3

To address these challenges, flexible electronics, which mimic brain tissue in terms of

materials, flexibility, and other key properties, may serve as promising candidates for BCIs.

(Tang et al., 2023, p. 109).

With the current advancements in chip manufacturing and material science, we can

produce smaller chips that better adapt to brain tissue and provide enhanced Signal

Acquisition, requiring minimal calibration from external software. For example, ultra-

flexible nano-electronics with sizes less than 10 µm and thicknesses of 1 µm, featuring

mesh-like designs, may be highly compatible with neurons and axons (Tang et al., 2023, p.

112).

Another challenge that flexible electronics overcomes is the level of invasiveness required

to implant the electrode.

With flexible materials, a stent-like electrode can be easily inserted into the brain using

catheters, thereby reducing invasiveness and the risks associated with implantation.

Bi-directional BCIs and Artificial Intelligence

Future technologies may aim to work on bi-directional BCIs. With Artificial Intelligence, BCI

users can utilize AI to create memories, assist with learning, or even augment human

intelligence.

This area of study is still in its infancy, but it reveals the fascinating world of how AI and the

human brain can work together, albeit there are ethical and policy concerns.

Conclusion and Opinion

BCIs are one of the many fields of technology that demonstrate the significant

advancements in replicating the highly complex biological systems that reside within us.

BCIs hold tremendous potential – from helping patients with motor, learning, and other

neurological disorders to creating a human brain that has AI as a built-in assistant.

Brain-computer interfaces, to me, represent how interdisciplinary the world has become –

where discoveries and revolutions lie in the blurred lines between disciplines.

-Aryaman Deshmukh

Brain-Computer

Interfaces

4

Keywords -

Brain-Computer Interfaces, Flexible Electronics, Artificial Intelligence, Neural Networks,

Signal Acquisition, Bi-directional Interfaces

Bibliography

Shih, J. J., Krusienski, D. J., & Wolpaw, J. R. (2012). Brain-computer interfaces in medicine.

In Mayo Clinic Proceedings (Vol. 87, Issue 3, pp. 268–279). Elsevier Ltd.

https://doi.org/10.1016/j.mayocp.2011.12.008

Tang, X., Shen, H., Zhao, S., Li, N., & Liu, J. (2023). Flexible brain–computer interfaces.

Nature Electronics, 6(2), 109–118. https://doi.org/10.1038/s41928-022-00913-9

Webster, P. (2025). Can AI-powered brain–computer interfaces boost human intelligence?

In Nature Medicine. Nature Research. https://doi.org/10.1038/s41591-025-03641-7

Yamins, D. L. K., & DiCarlo, J. J. (2016). Using goal-driven deep learning models to

understand sensory cortex. In Nature Neuroscience (Vol. 19, Issue 3, pp. 356–365). Nature

Publishing Group. https://doi.org/10.1038/nn.4244

CNET (2023, September 24). Brain-Computer Interface: No Open Brain Surgery Required

[Video] https://youtu.be/3GCS9JqNJB0