Real-time Chinese language decoding. Photo provided by the interviewee.  

On New Year’s Day 2025, a neurology patient at Shanghai Huashan Hospital successfully sent the world’s first New Year’s greeting through thought alone. By merely thinking “Happy New Year 2025” in his mind, a computer decoded his brain signals and commanded a mechanical hand to make a heart gesture.

This remarkable achievement is the result of a collaborative project between Shanghai Braintiger Technology Co., Ltd. (hereinafter referred to as Braintiger Technology), the Neurosurgery Department of Fudan University’s Huashan Hospital, and the Tianqiao Brain Science Research Institute. Through indigenous innovative invasive flexible brain-machine interface technology, they conducted high-precision real-time motion and language decoding clinical trials, successfully achieving “brain-controlled” intelligent devices and “thought communication”. This milestone marks China reaching an advanced global level in brain-machine interface technology.

Global brain-machine interface research is flourishing. On January 28, 2024, Elon Musk’s brain-machine interface company Neuralink successfully performed its first human brain chip implantation, signaling the transition of brain-machine interface technology from theoretical research to practical application.

“Represented by Elon Musk’s Neuralink invasive hardcore technology, motion decoding has achieved numerous surprising breakthroughs in the United States, Europe, and Asia. Examples include enabling paralyzed patients to control mechanical arms to drink water, operate mice for gaming, or control exoskeletons to restore walking. The next anticipated breakthrough is language decoding,” multiple overseas scientists expressed this view at the global brain-machine interface academic conference, BCI Society International Forum, held in Shanghai previously.

Compared to decoding English’s 26 letters, decoding Chinese with its “418 syllables and 4 tones” presents a higher difficulty. The information conversion during Chinese communication involves more brain regions, requiring the development of neural encoding and decoding mechanisms and information processing methods specific to Chinese characteristics.

Real-time mind control of the XessOS system. Photo provided by the interviewee.  

“We have achieved the world’s first real-time Chinese encoding and decoding,” said Tao Hu, founder and chief scientist of Braintiger Technology. He believes that Braintiger Technology aims not only to decode motion but also language, and specifically the notoriously challenging Chinese language.

“Successful language decoding will inject infinite imagination into brain-machine interfaces, not only restoring language function for patients with speech disorders but potentially enabling direct connection and interaction between the human brain and AI large models, shaping the most powerful brain,” related industry experts stated.

In December 2024, Braintiger Technology collaborated with Professor Wu Jinsong’s team from Huashan Hospital’s Neurosurgery Department to conduct China’s first high-throughput implantable flexible brain-machine interface for real-time Chinese language synthesis clinical trial. The 43-year-old patient was suffering from a language area tumor-induced epilepsy. The project team implanted a soft, thin electronic membrane—Braintiger’s self-developed 256-electrode high-throughput brain-machine interface—to help locate the lesion and protect critical language-related brain function areas.

Two days post-surgery, the patient began related training. By the seventh day, they achieved a 71% decoding accuracy for 142 common Chinese syllables, with a single character decoding latency of less than 100 milliseconds.

Previously, the team had also achieved significant results in motion decoding. In August 2024, Braintiger Technology collaborated with Professor Mao Ying and Professor Chen Liang’s team from Huashan Hospital’s Neurosurgery Department to complete a clinical trial of motion synthesis brain-machine interface for patients with movement disorders. Motion synthesis involves using “brain-controlled” intelligent devices to help patients with movement disorders rebuild motor functions. In the clinical trial, a patient with a motor area tumor-induced epilepsy successfully controlled mobile phone apps for communication and shopping after two days of training following electrode implantation.

“This language decoding clinical trial was a short-term in vivo test of approximately one month. These two trials demonstrate the enormous potential of brain-machine interface technology in reconstructing motor and language capabilities. We will gradually conduct long-term in vivo clinical trials,” Tao Hu revealed. The company plans to complete clinical multi-center registration for three categories of medical devices within three years.