MK sports Korea a member of Chinese People's Political Consultative Conference National Committee and vice president of Tianjin University Photo: Courtesy of Ming Dong" src="https://www.globaltimes.cn/Portals/0/attachment/2025/2025-03-06/75fdb62f-fce1-4052-bab3-6334e15c4909.jpeg" />

Ming Dong, a member of Chinese People's Political Consultative Conference National Committee and vice president of Tianjin University Photo: Courtesy of Ming Dong

China should establish undergraduate programs in brain-computer interface (BCI) at top universities and develop a multi-disciplinary BCI industry-education integration community to enhance long-term, integrated training for top-tier innovative talent in related fields, Ming Dong, a member of the Chinese People's Political Consultative Conference (CPPCC) National Committee and vice president of Tianjin University, told the Global Times on the sidelines of this year's two sessions.

Ming pointed out that the field still faces significant challenges, including a shortage of high-level talent, an underdeveloped academic system and training mechanism, a lack of interdisciplinary industry-education integration frameworks, and an immature talent development ecosystem. He therefore called for the "establishment of a systematic talent development framework."

The comment was made as China's non-invasive BCI technology has reached an advanced global level, achieving major breakthroughs in fields such as motor neurorehabilitation and manned spaceflight. In parallel, invasive BCI technology has also entered clinical trials for applications such as spinal cord injury treatment and speech decoding, according to Ming. 

In February, Ming's team -- in collaboration with Tsinghua University -- developed a novel non-invasive BCI system based on a memristor-based neuromorphic chip. The system for the first time revealed the synergistic enhancement effect between electroencephalography (EEG) evolution and decoder adaptation during brain-machine interaction, enabling mutual adaptation and co-learning between biological and machine intelligence.

Ming said that he believed that the development of the BCI industry - a forward-looking and disruptive strategic field that is a representative of China's new quality productive forces - will be "facing unprecedented opportunities and a favorable policy environment" amid China's industrial upgrade as well as more measures to cultivate new quality productive forces and advance technological innovation. 

"The year of 2025 is expected to be a key year for the application of BCIs, with the technology poised for large-scale breakthroughs and a richer variety of application scenarios," Ming said. 

He called on the industry to ride on the momentum and pour in more efforts for achieving technological self-sufficiency, building a comprehensive industry chain that spans from fundamental research to product applications. Relevant measures could include strengthening underlying technological innovation to overcome key technical bottlenecks, and driving the transformation of scientific and technological achievements based on the needs of major engineering applications, Ming suggested.

For example, his team at Tianjin University is dedicated to addressing the core issues in the development of non-invasive BCIs. In terms of hardware interfaces, the team will focus on developing reliable, flexible, and comfortable sensor electrodes, as well as portable, user-friendly, and aesthetically pleasing hardware forms. On the software interface side, the team will develop natural and continuous human-computer interaction coding methods and construct a universal electroencephalography (EEG) model by integrating prior knowledge of EEG, Ming noted.

"However, the application of BCIs still faces numerous challenges, such as safety issues of invasive technologies and low signal-to-noise ratios and insufficient spatial resolution of non-invasive technologies. Only by overcoming these technical bottlenecks and establishing comprehensive ethical standards - clarifying red lines regarding data privacy, life safety, and human experimentation - can BCIs truly achieve large-scale application," he said.

Looking forward to the future, Ming expected deep integration of BCIs and artificial intelligence (AI) to promote a high degree of unity between machines and humans.

"We could imagine that BCIs can integrate with AI large models like DeepSeek to fully leverage the stable command bandwidth of brain-computer decoding, guiding DeepSeek in providing auxiliary support for executing brain intentions, thereby promoting innovation in BCIs' application scenarios," he explained. 

He also stated that the current large language model frameworks are not entirely suitable for establishing large models for BCIs. And the BCI field lacks a large amount of standardized, high-quality EEG data, making it extremely challenging to build high-quality foundational EEG models. As such, from a technical perspective, developing a large model similar to DeepSeek for BCI - without addressing these two issues - will require further breakthroughs in key technologies.