A research team from the Chinese Institute for Brain Research (CIBR) has made significant advancements in brain-machine interface (BMI) technology, successfully implementing a semi-invasive system that aids patients with speech impairments and motor function loss.

At a recent press conference, details emerged regarding the efficacy of the NeuCyber Matrix BMI System, developed collaboratively by CIBR and NeuCyber NeuroTech (Beijing) Co., Ltd. Clinical data from initial patient cases indicate that over 98 percent of implanted BMI channels remained functional post-surgery.

The NeuCyber Matrix BMI System employs a thin, flexible, nano-fabricated film microelectrode, capable of processing electrocorticography (ECoG) signals.

The system, characterized by its 128-channel configuration, includes a compact micro-circuit that aids in signal recording and processing, contributing to its operational reliability.

For the first time, an aphasic patient has demonstrated the ability to output the Chinese language through this semi-invasive BMI system, marking a significant milestone in communication recovery.

Paralyzed individuals have also effectively utilized the technology to operate computers and robotic limbs, thereby compensating for lost motor function.

CIBR Director Luo Minmin highlighted three pivotal technologies essential for the system’s performance: a high-integration micro-host for signal processing, next-generation wireless communication technology for data transmission, and an advanced algorithm capable of real-time decoding of movements and language.

Recently, a surgical procedure led by Zhao Guoguang, president of Xuanwu Hospital, achieved a breakthrough with the world’s first implementation of a wireless Chinese-language BMI for a patient with amyotrophic lateral sclerosis (ALS), who had previously lost the capacity to speak.

Using a semi-invasive surgical technique, the team accurately placed the NeuCyber Matrix BMI System on the dura mater of the left brain hemisphere, which governs language processing, while minimizing trauma.

A control and signal transmission device, roughly the size of a coin, was also integrated into the surface of the patient's skull to facilitate efficient neural signal relay and wireless power supply.

Post-surgery, the patient commenced language decoding training on March 14. Initial assessments indicated a real-time decoding accuracy of 34 percent for 62 common words after three hours, which has since improved to 52 percent.

Through an adaptive error correction algorithm that leverages a large language model, the patient has begun to regain basic communication capabilities, successfully decoding phrases such as "I want to drink water" and "I'm in a great mood today.

I'd like to take a walk with my family.” The latency for decoding a single Chinese character is now below 100 milliseconds.

Li Yuan, business development director at NeuCyber NeuroTech, articulated the potential of BMI technology to facilitate communication for speech-impaired individuals, highlighting the role of synthetic speech technology in making their thoughts audible.

Despite these advancements, Zhao emphasized that ongoing research would focus on employing BMI technology for other neurological conditions, including intractable epilepsy, spinal cord injury, stroke, ALS, and aphasia.