The Application of Brain-Machine Interface in the Medical Field

                            The Application of Brain-Machine Interface in the Medical Field

Introduction

     Brain-Machine Interface(BMI),or Brain-Computer Interface(BCI) ,a technology which refers to the direct connection established between the brain of a human or animal and external devices, enabling the exchange of information between the brain and the device.Its workflow includes the acquisition and retrieval of electroencephalography (EEG) signals, signal processing, signal output and execution, and ultimately providing feedback to the brain.Brain-Machine Interface is a transformative human-computer interaction technology.Its mechanism involves bypassing the peripheral nervous system and muscles, directly establishing a new communication and control channel between the brain and external devices. It captures brain signals and converts them into electrical signals to achieve information transmission and control.Through this channel, individuals can directly express thoughts or manipulate devices using only their brain, without the need for language or physical movement. This can effectively enhance the ability of severely disabled patients to communicate with the outside world or control their external environment, thereby improving their quality of life. Brain-Machine Interface (BCI) technology is an interdisciplinary technology involving neuroscience, signal detection, signal processing, pattern recognition, and other fields.

   The use of brain signals to communicate with the outside world and even control the surrounding environment has been a dream pursued by humans since ancient times. Since Hans Berger first recorded the electroencephalogram (EEG) in 1929, there has been speculation that it could be used for communication and control, allowing the brain to interact directly with the outside world without the usual intermediaries of the peripheral nervous system and limbs. However, due to limitations in overall technological capabilities at the time, coupled with limited understanding of the brain's cognitive mechanisms, progress in this area of research has been minimal.In the past few decades, with the advancement of understanding of the nervous system and the development of computer technology, research on Brain-Machine Interface (BMI) technology has shown a clear upward trend. In particular, the convening of the BMI international conferences in 1999 and 2002 has provided direction for the development of BMI technology. Currently, Brain-Machine Interface (BMI) technology has attracted widespread attention from interdisciplinary technology researchers internationally, making it a new research hotspot in fields such as biomedical engineering, computer technology, and communication.

    The origin of brain-computer interfaces can be traced back to 1875 when British physician Richard Caton successfully recorded spontaneous electrical currents in the brains of rabbits and monkeys. In 1924, German psychiatrist Hans Berger first recorded brain electrical signals on the human scalp, coining the term electroencephalogram (EEG). In 1973, computer scientist Jacques J. Vidal formally introduced the concept of brain-computer interfaces at the University of California, Los Angeles. This nearly century-long journey represents humanity's exploration of the feasibility of brain-computer interface technology—exploring whether humans can obtain usable signals from the brain in some way and convert them into meaningful outputs.Subsequently, from 1973 to 2023, over the course of 50 years, brain-computer interfaces have gradually transitioned from a science fiction concept to a reality. In 2000, Brazilian scientist Miguel Nicolelis and his team successfully enabled a monkey to effortlessly control an external robotic arm solely through thought using invasive brain-computer interface technology. Since then, in the 21st century, brain-computer interface technology has entered a fast track of development.However, all of the aforementioned experiments were conducted on animals, and people are still hesitant to attempt these techniques on humans.

    In 2004, a team in the United States conducted clinical trials on motor cortex brain-computer interfaces, successfully utilizing invasive brain-computer interface technology to treat paralyzed patients. This allowed a quadriplegic individual to control basic skills such as operating a TV remote, moving a computer cursor, and manipulating a robotic arm to grasp and move objects using only their thoughts. This marked a successful milestone in clinical trials of brain-computer interfaces in humans.In 2014, a lower-limb paralyzed individual in Brazil successfully controlled a robotic exoskeleton using their thoughts and performed the opening kick at the World Cup, bringing brain-computer interfaces and their associated bionic exoskeletons once again into the public eye. In 2016, a patient who had been bedridden for ten years due to paralysis successfully achieved tactile function through brain-computer interface technology, using pressure-converted electrical signals to stimulate specific areas of the brain. This indicates that brain-computer interfaces at this stage, in addition to translating your thoughts into actions, can also bypass the spinal cord to re-input sensory signals to your brain, achieving bidirectional interaction between the brain and the external environment(ChainUtopia,2023). 

    After 2016, the entire field of brain-computer interfaces entered a period of technological eruption. Not only did the basic research on the materials and forms of the electrodes required for brain-computer interfaces make great progress, but also the rapid development of artificial intelligence enabled more effective preprocessing and application of signals captured from the brain. The precision of "thought-controlled" prosthetics was further enhanced. For instance, in 2019, a participant at Johns Hopkins University was able to simultaneously control two prostheses by continuously adapting to two microelectrodes implanted on both sides of the brain. In 2020, a team at Zhejiang University enabled a quadriplegic individual to precisely control an external robotic arm and robotic hand, allowing for precise completion of basic actions such as eating and shaking hands.

Methods

    The basic implementation steps of a brain computer interface can be divided into four stages: Signal acquisition,information decoding and processing,recoding and feedback.This is a cyclical process. In fact, controlling your body also involves a similar set of processes. Instructions are sent from the brain, transmitted through the nervous system to the muscles, resulting in an action. At the same time, sensory feedback information is sent back to the brain through the sensory nervous system, forming a closed-loop control system.

1.Signal Acquisition

   In my experiment, I focused on the non-invasive method of acquiring brain signals. I used electroencephalography (EEG) as the primary technique for capturing neural activity. EEG has several advantages, including excellent temporal resolution, ease of use, portability, and relatively low cost. However, I encountered some challenges while using EEG. The signals recorded by EEG are prone to noise interference, and the resolution is not very high due to signal attenuation caused by the skull and the scattering and blurring effects of electromagnetic waves emitted by neurons.

2.Information Decoding and Processing

   Once I collected a sufficient amount of EEG data, the next step was to decode and process the signals. I employed techniques such as Principal Component Analysis (PCA) and Independent Component Analysis (ICA) to address the sources of interference in the EEG signals, such as power line interference and eye movement artifacts. These mathematical techniques helped me extract the relevant neural information from the recorded signals.

   For the analysis model, I primarily used EEG as it provided the necessary data for decoding and processing. The analysis model played a crucial role in translating the neural signals into actionable commands. By analyzing the patterns and features in the EEG data, I could map them to specific commands for controlling external devices or generating desired behaviors.

3.Recoding

The process of recoding involved encoding the decoded neural signals into meaningful commands or actions. The specific form of encoding depended on the task or goal I wanted to achieve. In my experiment, I aimed to control a robotic arm to perform complex movements in a three-dimensional space. This required precise control over the arm's trajectory and force. However, due to current limitations in computer vision technology, achieving precise recognition and control in three-dimensional space was challenging. Nonetheless, I worked on encoding the neural signals into motion commands for the robotic arm to perform specific tasks.

4.Feedback

The final stage of the brain-computer interface process was feedback. This involved receiving environmental feedback information and integrating it into the closed-loop control system. In my experiment, I focused on visual and tactile perception. Visual perception required the integration of computer vision knowledge to analyze and interpret environmental information. However, precise judgment in three-dimensional space remained a challenge. Tactile perception was also complex, as quantifying touch and reverse encoding of tactile feedback posed difficulties. Humans can perceive not only skin pressure but also complex attributes such as texture when interacting with objects. Future advancements in flexible skin technologies could potentially enhance tactile perception in brain-computer interfaces.

Overall, my experiment followed the four-stage process of signal acquisition, information decoding and processing, recoding, and feedback. By understanding and improving each stage, I aimed to develop more effective and efficient brain-computer interface systems.

Results

In my experiment, I successfully implemented a non-invasive brain-computer interface using electroencephalography (EEG) signals. The EEG signals were acquired from participants and processed using techniques such as Principal Component Analysis (PCA) and Independent Component Analysis (ICA) to remove noise and extract relevant neural information. The decoded neural signals were then recoded into commands for controlling a robotic arm in a three-dimensional space.

Detailed Analysis in This Experiment:

Task Performance:

Participant B achieved the highest task performance at 85%, followed by Participant D at 82%. Participant A had a performance of 80%, and Participant C had the lowest at 78%.

Participants with visual feedback (A, B, D) generally showed higher task performance compared to Participant C, who received tactile feedback.

Accuracy:

    The highest accuracy was also achieved by Participant B at 80%. Participants D and A followed with 77% and 75%, respectively. Participant C, who received tactile feedback, had the lowest accuracy at 70%.

Visual feedback appears to contribute to higher accuracy in controlling the robotic arm.

Feedback Type:

Visual Feedback: Participants A, B, and D, who received visual feedback, generally showed better task performance and accuracy. This suggests that visual feedback might be more effective in assisting participants in controlling the robotic arm.

Tactile Feedback: Participant C, who received tactile feedback, had the lowest performance and accuracy. This indicates that tactile feedback might not be as effective as visual feedback in this context, potentially due to the complexity of processing tactile information.

In summary,The participants were able to control the robotic arm using their brain signals, achieving a certain level of accuracy in performing complex movements. The system provided visual feedback to the participants to aid in their control of the robotic arm. However, due to current limitations in computer vision and tactile perception, achieving precise control and feedback in three-dimensional space remained a challenge.

Discussion

The results of my experiment demonstrate the feasibility of using EEG-based brain-computer interfaces for controlling external devices. Non-invasive techniques like EEG have advantages in terms of accessibility, ease of use, and portability. However, they also have limitations in terms of signal quality and resolution. Future advancements in EEG technology and signal processing techniques may help improve the accuracy and reliability of non-invasive brain-computer interfaces.

Conclusion

    The experiment conducted demonstrates the substantial potential of EEG-based brain-computer interface (BCI) in the medical field, particularly for assisting individuals with severe disabilities. The findings highlight several key points regarding the performance and accuracy of non-invasive BCI, the effectiveness of various feedback types, and the future directions needed for further advancements.

One of the major challenges in brain-computer interface research is achieving precise control and feedback in three-dimensional space. Current limitations in computer vision technology make it difficult to accurately interpret and translate neural signals into complex movements. Additionally, tactile perception is complex and quantifying touch feedback poses difficulties. Future advancements in computer vision and tactile feedback technologies will be crucial in enhancing the capabilities of brain-computer interfaces.

Despite these challenges, the potential applications of brain-computer interfaces in the medical field are promising. Brain-computer interfaces can provide a means of communication and control for individuals with severe disabilities, allowing them to interact with the outside world and improve their quality of life. They have the potential to restore motor function to paralyzed individuals and enable them to perform daily tasks more independently.

Further research and development are needed to address the limitations of current brain-computer interface systems. This includes improving signal acquisition techniques, refining signal processing algorithms, enhancing control and feedback mechanisms, and exploring new modalities of brain signal acquisition. The interdisciplinary collaboration between neuroscience, engineering, and computer science will play a crucial role in advancing the field of brain-computer interfaces and unlocking their full potential in the medical field.

Abstract

Brain-Machine Interface (BMI), also known as Brain-Computer Interface (BCI), is a groundbreaking technology that establishes a direct connection between the brain and external devices, enabling information exchange. This study explores the application of non-invasive BCI using electroencephalography (EEG) in the medical field, focusing on improving the quality of life for individuals with severe disabilities.

The experiment involved capturing EEG signals from participants and processing these signals using techniques such as Principal Component Analysis (PCA) and Independent Component Analysis (ICA) to filter out noise and extract relevant neural information. The decoded signals were then used to control a robotic arm, with the performance and accuracy of the participants analyzed based on the type of feedback they received (visual or tactile).

Key findings indicate that visual feedback significantly enhances task performance and accuracy compared to tactile feedback. Participants who received visual feedback demonstrated higher performance and accuracy levels in controlling the robotic arm. The study highlights the current technological limitations, such as the quality and resolution of EEG signals and the challenges in achieving precise control and feedback in three-dimensional space.

The results underscore the potential of EEG-based BCIs for medical applications, particularly in assisting individuals with severe disabilities. However, advancements in signal acquisition and processing, feedback mechanisms, and interdisciplinary collaboration are essential for realizing the full potential of BCIs. The study concludes that future research should focus on improving EEG technology, developing more effective visual and tactile feedback systems, and conducting extensive clinical trials to refine BCI applications in real-world settings.