Foreword

With the rapid advancement of technology, Brain-Computer Interfaces (BCIs), serving as a vital bridge between the human brain and external devices, have progressively emerged as a research hotspot in neuroscience and artificial intelligence. Researchers worldwide are actively exploring innovations in this cutting-edge field, striving to overcome technical bottlenecks and enable smarter, more efficient human-machine interactions.

In this exclusive interview, we are honored to have Professor Wang Shouyan, Director of the Center for Neural Modulation and Brain-Computer Interface Research at Fudan University. Professor Wang provides an in-depth analysis of the current state, development advantages, and future trends of BCI research in China. He also shares forward-looking insights and invaluable experiences from his work in the field.

The following is the interview with Professor Wang Shouyan. We have condensed the text to make it easier to read.

Question: What is the current state of Brain-Computer Interface (BCI) development in China?

Wang Shouyan: I am from the Institute of Brain-Inspired Intelligence Science and Technology at Fudan University and am responsible for establishing the Center for Neural Modulation and Brain-Computer Interface Research. First, I’d like to thank you for organizing this interview, which provides an opportunity to help the public better understand the true meaning, development status, and key contributors of BCI, rather than focusing solely on Musk’s promotional activities.

In fact, BCI research extends far beyond Musk’s endeavors. Recently, I proposed a four-stage developmental roadmap for BCIs*: (1) Brain Reading, (2) Brain Writing, (3) Brain-Computer Interaction, and (4) Brain-Intelligence Integration. This roadmap encompasses processes from basic signal decoding to advanced human-machine interaction. Currently, research is largely focused on decoding brain signals, a field in which both Musk and Tsinghua University in China are actively engaged.

*Professor Wang emphasizes that the scientific development of BCI will follow four phases: from Brain Reading (decoding motor, speech, memory, and consciousness) to the currently thriving field of Brain Writing (neural function modulation and reconstruction) and Brain-Computer Interaction (brain interfacing), eventually reaching Brain-Intelligence Integration (brain-inspired intelligence and digital life).

From the perspective of decoding brain signals, could using MRI scans also count as “Brain Reading”? International studies have already employed MRI and magnetoencephalography to analyze brain functions, which represents one direction of BCI research. Simultaneously, techniques such as electromagnetic stimulation are being explored to regulate brain activity.

The essence of BCI lies in achieving two-way information exchange between the human brain and external machines. We aim to enable this interaction through more intelligent means, such as using virtual reality to modulate brain activity. Therefore, I proposed BCI 3.0 (Brain-Computer Interaction), which emphasizes deep interaction and integration between the human brain and machines. The more advanced BCI 4.0 (Brain-Intelligence Integration) focuses on the coexistence and collaboration between human brains, intelligent agents, artificial intelligence, and future intelligent societies.

In these research directions, many Chinese neuroscientists, engineers, and physicians are actively exploring. For example, Huashan Hospital and Tiantan Hospital have achieved significant progress in this field.

In addition, we must strengthen science communication. The public needs to recognize the importance of BCI and the national commitment to its development. Relying solely on Musk’s research cannot represent the strategic development of a country. True strategic development involves core foundational technologies, which have profound implications for the health and well-being of society and its people.

Overall, BCI is advancing by integrating multiple factors, driving societal progress, forming the prototype of future industries, and becoming a vital part of national strategic planning.

Question: What are the advantages of China’s BCI development?

Wang Shouyan: First, we have a significant international advantage in terms of talent and disciplinary foundation. China’s BCI research began in the 1980s, and after decades of accumulated experience, we now possess a wealth of experience in BCI research, constituting a major strength.

Second, China has a solid foundation in medical and neurological disease research, which provides essential support for BCI development. We can leverage this foundation to make advances in the medical applications of BCI technologies.

Third, we have a long-standing foundation in technological development, particularly in scientific research and engineering technologies related to BCI, laying a solid groundwork for future progress.

However, we also face challenges, particularly in the integration of interdisciplinary research. We need to enhance collaboration among experts from various fields to jointly address major scientific, technological, and industrial challenges. Both the central government and the Shanghai government are actively planning and addressing these issues to further drive the development of BCIs.

Question: What breakthrough discoveries do you foresee in the field of Brain-Computer Interfaces (BCIs) over the next 5 to 10 years?

Wang Shouyan: The future breakthroughs in BCIs are likely to focus on several key areas. First, significant advancements have been made by researchers at ETH Zurich in spinal cord rehabilitation*, particularly in technologies that enable paraplegic patients to walk again. This represents a cutting-edge direction where artificial intelligence intersects with BCIs. Additionally, the integration of brain-reading and brain-writing technologies, especially innovations in deep brain stimulation for Parkinson’s disease and pain management, is currently a major focus of international research in both basic science and clinical medicine.

*In China, the team led by Jia Fumin at Fudan University’s Institute of Brain-Inspired Intelligence Science and Technology has developed a new generation of implantable brain-spinal interface devices for patients with spinal cord injuries. This innovation won the 2024 National Disruptive Technology Innovation Competition and is expected to undergo its first clinical trial by the end of the year, offering hope for spinal cord injury patients to stand and walk again.

These breakthroughs are not accidental; they are driven by profound underlying forces. By analyzing these driving forces, we can anticipate future trends. China’s early brain initiatives strategically invested in critical technologies such as neural modulation and BCIs, laying a strong foundation for future technological breakthroughs.

One clear trend from these initiatives is the move toward intelligent implantable deep brain stimulation technologies. The national strategic goal is to achieve a leading position in global competition, leveraging China’s strengths in medical industries and technology to achieve leapfrog development.

Another potential breakthrough lies in the decoding of neural activity. China has extensive resources in clinical research and data accumulation. With the launch of upcoming major national projects, progress in brain information decoding is expected to accelerate, paving the way for new advancements in BCI technology.

Question: If we gather enough brain data, is it possible to build a general brain model similar to ChatGPT?

Wang Shouyan: Large language models (LLMs) like ChatGPT have indeed provided us with significant inspiration. However, this question should be approached from two perspectives: the problem and the methodology. While LLMs offer a promising approach for understanding and decoding brain information, they are not the sole solution to the problem.

Although large data-driven models are disruptive and innovative, helping us better interpret and analyze brain data, relying solely on one approach in brain science research often leads to bottlenecks. Therefore, we advocate for a research paradigm that combines mechanisms and methodologies. For instance, optogenetics can help us better understand specific neural phenomena.

Currently, we are conducting research on integrating LLMs with large-scale EEG data modeling. The critical question is whether these models can genuinely help us address fundamental scientific questions. The future direction is likely to involve the convergence of data, mechanisms, and modeling. The ultimate value and significance of this technical approach depend on its ability to solve major scientific challenges.

Question: What research question are you most focused on?

Wang Shouyan: My primary focus is on how to effectively model under limited data conditions. In hospital-based research, we often cannot collect sufficient large-scale datasets, particularly when studying specific diseases. For example, in implantable device research, studies involving dozens of patients are already considered substantial, while those with over a hundred patients are extremely rare.

Does this mean that patients in such circumstances should not be studied? Or that they do not deserve treatment? These questions are not only scientific but also philosophical in nature. We must explore how to conduct effective research with limited data. Specifically, we are currently focused on leveraging large language models (LLMs) for precise modeling with small datasets. This approach aims to develop more personalized and intelligent strategies for neural modulation and brain-computer interaction.

Question: Are there valuable research cases in the application of small data?

Wang Shouyan: I can share some examples from our research on deep brain stimulation (DBS) for Parkinson’s disease. In China, over 100 hospitals are now offering this treatment, and tens of thousands of patients undergo this surgery every year. However, there are still hundreds of thousands of patients who need treatment, but only a small fraction have access to it. Our current focus is on how to use more precise neural modulation technologies to help a broader range of patients, especially those with significantly varied symptoms, benefit from treatment.

Additionally, we are collaborating with Huashan Hospital and Tiantan Hospital to study chronic pain patients. This includes those who have no effective pharmaceutical treatments or even patients in a vegetative state. Our goal is to use disruptive technologies to help these patients reconnect and communicate with their families. The core of these studies is leveraging small datasets to develop more precise and effective treatments.

Question: What challenges and difficulties do you see in the development of the Brain-Computer Interface (BCI) field?

Wang Shouyan: One of the biggest challenges in the broader scientific research landscape is interdisciplinary collaboration. Currently, China’s research evaluation system, with its emphasis on publishing papers and securing funding, often creates conflicts of interest. Consequently, institutional and systemic issues, particularly barriers to interdisciplinary collaboration, pose significant difficulties.

The root of these barriers lies in a value assessment system historically driven by metrics such as the number of publications and performance indicators. Researchers tend to focus on their specialized fields and are often reluctant to engage deeply with experts from other disciplines. Additionally, issues related to data sharing, especially disputes over authorship, frequently complicate collaborations. These evaluation criteria not only affect the publication of scientific results but also directly impact researchers’ year-end bonuses and personal assessments.

We are currently advocating for institutional innovation in the BCI field to break down these barriers. Chinese researchers are highly intelligent and hardworking, but overcoming these remaining challenges requires addressing systemic issues and establishing efficient collaborative teams. Adequate funding support and the creation of interdisciplinary platforms are essential. Other technical challenges can be resolved gradually in accordance with the natural progression of scientific development. Interdisciplinary collaboration is not an insurmountable scientific problem; instead, institutional obstacles represent the real bottleneck.

Question: What advice would you give to young scientists?

Wang Shouyan: BCI is a highly interdisciplinary field, making a broad knowledge base and the establishment of collaborative networks especially important. For young scientists, I have a few straightforward suggestions. First, read extensively and learn continuously. The old saying “read ten thousand books and travel ten thousand miles” remains relevant. Young scholars should visit and learn from different laboratories to broaden their horizons. Second, make friends and actively participate in academic conferences. Don’t hesitate to ask questions and engage in discussions during these events.

For students and postdoctoral researchers in the BCI field, stepping out of their narrow research focus is particularly critical. Understanding the scientific questions that other researchers are exploring will help them gain a broader perspective and achieve greater heights, ultimately making their work more forward-looking and practically meaningful. I hope everyone can gain something valuable from this process. Thank you.