Peng, a serial entrepreneur in China, is pursuing brain–computer interface development through two ventures: NeuroXess, which develops implantable BCI systems, while Gestala is dedicated to non-invasive interfaces that rely on ultrasound technology. The efforts come as U.S. companies including Elon Musk’s Neuralink and OpenAI-backed Merge Labs continue advancing similar technologies.

Investor demand significantly exceeded the amount sought, with commitments totaling more than $58 million, Peng said.

The deal marks the biggest early-stage funding round so far in China’s brain–computer interface industry. The capital will help grow the team from 15 to about 35 employees by the end of the year, support R&D, and fund the construction of a manufacturing facility in China. Gestala  expects to complete its first-generation prototype later this year.

Globally, investment in BCI technologies has been rising, especially when it comes to ultrasound technology. This is the first ultrasound BCI company in China.

According to Peng ultrasound may become the next generation of brain–computer interface technology, offering the possibility of broader access to brain activity and new ways to interact with neural signals. Its possible that the non-invasive ultrasound could reduce the risks associated with brain surgery. The ultrasound technology can monitor larger areas of the brain, including deeper neural circuits. Also using phased-array ultrasound, the system can precisely stimulate or inhibit neural activity without requiring surgical procedures, he explained.

The company is also investigating potential uses in mental health conditions such as depression, autism and obsessive-compulsive disorder, PTSD, and even stroke rehabilitation. Some of the longer-term targets include Alzheimer’s disease, essential tremor and Parkinson’s disease.

Gestala is collaborating with major Chinese hospitals to fast-track clinical trials at substantially reduced costs, estimated at just 20% to 33% of comparable research in the U.S. or Europe. The company is also building  “Ultrasound Brain Bank” — a large clinical dataset aimed at training AI models to decode brain activity and aid future neurological diagnostics.

Despite increasing geopolitical tensions, Peng said he hopes researchers in the U.S. and China will collaborate on deep-technology development.

Both countries bring different strengths,

Peng said.

China offers large-scale clinical research capacity and efficient supply chains, while the U.S. has world-class scientific talent.

He added that joint initiatives could also focus on developing large clinical datasets to advance global neuroscience research.

Image: Phoenix Peng LinkedIn profile, Edited – 12.03.2026

How does the technique work?

It spots hidden patterns in the data and learns who is most at risk. The algorithm is suggested to have the power of helping reduce the number of people who may have been falsely diagnosed with dementia.

Data from people who have attended a network of 30 National Alzheimer’s Coordinating Centre memory clinics in the US, was analyzed by the study’s researchers. At the beginning of the study, the attendees didn’t have dementia even though they have had problems with their memory or some other brain functions.

The study started in 2005 and finished in 2015. For that period of time one in ten attendees was diagnosed with dementia within two years after they visited the memory clinic.

So, what did the research find?

According to the study, the machine learning model is able to predict these new dementia cases with up to 92 per cent accuracy. Also, it is even more accurate than two existing alternative research methods.

According to Professor David Llewellyn, who is  from the University of Exeter, they were now able to teach computers to accurately predict who will go on to develop dementia within two years.

Also, he added:

“We’re also excited to learn that our machine learning approach was able to identify patients who may have been misdiagnosed. This has the potential to reduce the guesswork in clinical practice and significantly improve the diagnostic pathway, helping families access the support they need as swiftly and as accurately as possible.”

Dr Janice Ranson, who also worked on the project, commented:

“We know that dementia is a highly feared condition. Embedding machine learning in memory clinics could help ensure diagnosis is far more accurate, reducing the unnecessary distress that a wrong diagnosis could cause.”

Machine learning was found to work effectively, using patient information which is routinely available in clinics such as memory and brain function, performance on cognitive tests and specific lifestyle factors.

Now the team has a new plan which includes conducting follow-up studies in order to evaluate the practical use of the method in clinics, as well as to assess whether it can be rolled out to improve dementia diagnosis, treatment and care.

The head of research at Alzheimer’s Research UK, Dr Rosa Sancho shared:

“Artificial intelligence has huge potential for improving early detection of the diseases that cause dementia and could revolutionise the diagnosis process for people concerned about themselves or a loved one showing symptoms. This technique is a significant improvement over existing alternative approaches and could give doctors a basis for recommending lifestyle changes and identifying people who might benefit from support or in-depth assessments.”

Besides, two years ago, VR was found to help people who are suffering from dementia recall past memories and tackle behavioral issues in 2019.