Great Country Academician

If he were asked to conduct research on the experimental data of brain-computer interface and bionic robot arm, Xu Chuan thought he didn't have the ability.

Every profession has its own specialty. Even Newton and Leonardo da Vinci, all-round scientists involved in many fields, have fields they don't understand.

The biological field is indeed not within the scope of his academic research.

However, if he was just asked to mine data from these experimental data and experimental images and use mathematical tools to analyze the laws in them, he could still do it.

Time passed by little by little, and the sky outside the window gradually dimmed.

It took an afternoon for Xu Chuan to complete the preliminary arrangement of the data and the ionic current dynamics formula based on the Hodgkin-Huxley model.

After processing these, he sent the preliminary formulas and data in his hand to Chuanhai Network Technology Company.

He took out his mobile phone from his pocket and opened the address book. After finding the familiar name in the special attention, he dialed it.

The phone rang twice and was quickly connected.

"Hello."

The call was connected, and a gentle voice sounded in his ears. Xu Chuan smiled and said:

"I sent you an email. After work tomorrow, I will find a few people to help build a mathematical model according to the content in the email."

"Okay, I'll take a look later."

Xu Chuan smiled and said, "Thank you for your hard work again. I'll treat you to a big meal in two days."

"Then I'll wait."

After chatting for a few words, Xu Chuan didn't think much and hung up the phone.

At the same time, on the other side, in a high-end residence in the Qixiashan New Development Zone, Liu Jiaxin looked at the hung up phone and the black screen, with a gentle smile on her face, and turned on the shower again.

The efficiency of Chuanhai Network Technology Company is very high.

It took only four days, and a complete mathematical model was delivered on the fifth day after Xu Chuan gave the modeling data.

After receiving the model, Xu Chuan directly loaded it into the small supercomputing center at home.

It has to be said that a supercomputer, even if it is only a small one, is extremely fast in processing various data, which is unmatched by ordinary computers, even the most expensive servers.

With the help of Xiao Ling, the AI ​​academic assistant, it took less than an hour to fully complete the calculation of the relevant data.

"Sure enough, the problem is not in the data conversion between the quantum mathematical model and the traditional multi-electrode array reset mathematical model."

Staring at the calculation data sorted out on the screen, Xu Chuan's eyes had a hint of "as expected" and he whispered softly.

As he expected, the quantum mathematical simulation model he had previously built specifically did not conflict with the traditional multi-electrode array reset mathematical model built by Xu Xiao himself.

The data conversion between the two was quite smooth, and there was no situation of enlarging, reducing or modifying the experimental data.

"If the problem did not appear here, what exactly caused the interference?"

Looking at the experimental data, Xu Chuan's face was interested.

He had seen the test experiments of the bionic robotic arm and robotic leg before, and the problem Xu Xiao mentioned did exist.

The brain nerve chip sensed that the brain wave signal was converted into an electrical signal and transmitted to the bionic robot arm, and there was indeed an abnormal situation.

After flipping through the experimental data, Xu Chuan fell into deep thought.

Although brain-computer interface technology is not in his research field, he still knows some general situations.

Apart from the fuzzy boundary between man and machine, the protection of mental privacy and autonomy, the ethical boundaries of neural intervention and other ethical problems.

There are two main problems with brain-computer interface technology.

One is the biocompatibility of implanted materials.

For example, the materials used in implanted brain-computer interfaces may cause brain rejection or brain damage due to movement.

After all, the brain is the most precise of all organs in the human body.

Encountering any external force may lead to serious problems such as brain damage and brain death.

However, this problem does not need to be considered at present, because the biocompatibility of materials theoretically does not lead to abnormal conversion and transmission of neural signals.

"Could it be that the capture of brain wave signals is not comprehensive?"

Flipping through the experimental data in the computer, an idea popped up in Xu Chuan's mind.

For brain-computer interface technology, the limitation of neural signal capture is a big problem.

An average person's brain has about 86 billion neural units, and humans can only capture a part of them.

This means that there are still a lot of neural signals that cannot be effectively used.

In particular, the neural network in the brain is not a simple linear superposition, but involves complex nonlinear relationships.

This makes it difficult to parse the encoding that occurs simultaneously.

And distinguishing the encoding of brain neural signals for specific behaviors from the encoding of other behaviors is still a big challenge.

Is there a problem in this aspect?

Thinking about it, Xu Chuan clicked on another file in the information Xu Xiao gave him, which contained the technology she and the team of Starlight Virtual Technology Company developed specifically for the Starlight brain-computer interface chip.

A two-section RNN architecture, a method of nonlinear dynamic modeling.

This technology uses a recurrent neural network architecture and training method, through nonlinear, dynamic modeling, separation and priority of behavior-related neural dynamics, and continuous and intermittent behavior data modeling.

It can improve the accuracy of neuro-behavior predictions, optimize the recognition of original local field potentials, and other areas that traditional neural signal simulation technology cannot achieve.

However, even he would find it difficult to find problems from these algorithms and experimental data.

After all, on the one hand, this is not a field he is familiar with, and on the other hand, the amount of experimental data of neural signals is a bit large.

Apart from other things, the frequency of the beta wave (beta wave) related to thinking, conscious problem solving, and attention to the outside world in the normal awake state of brain rhythm alone is as high as 14-30Hz.

It sounds like this data is very small. After all, 14-30 fluctuations per second are nothing for human research and development technology.

But if it is combined with the feedback and processing of various external signals by the brain nerves, the data generated is an extremely large amount.

Fortunately, for the brain nerve model, most of the data can be classified by different indicators.

Otherwise, it is simply unrealistic to process such a large amount of data through the brain-computer interface chip.

In the study, Xu Chuan took a sip of the tea that had already cooled in the porcelain cup to moisten his throat and stretch his tired eyes.

"Xiao Ling, help me keep an eye on the data analysis work of the SAS data platform. If there is data that is more than 5% different from the data that has been completed before, remind me."

"Okay, master! Leave it to Xiao Ling!"

In the study, Xiao Ling's voice sounded, and Xu Chuan pulled out the chair and walked outside, ready to take a shower.

It must be said that this is indeed a more difficult problem he has encountered in applied mathematics.

Almost all the brain nerve model data and the converted electrical signal data have no problems or abnormalities from a mathematical point of view.

Even if the entire data is analyzed and processed through the SAS data platform, no problems are found.

After eliminating the possible errors and problems in the data conversion between the two mathematical models, there has been basically no new progress in the problems in the brain-computer interface technology for several days.

After taking a shower and getting rid of his fatigue, Xu Chuan took out a bag of yogurt from the refrigerator, held it in his mouth and walked towards the study.

The problem with the brain-computer interface chip has consumed his time for more than ten days. If he can't find the problem in the next two days, he will put it aside for now.

Although failing to solve this problem will affect his image of "omnipotence" in Xu Xiao's mind.

But he has a lot of other work on hand, and it is impossible to spend all his time on this.

Just as he was thinking about how to restore his image in Xu Xiao's mind after suspending his research, the voice of AI academic assistant Xiao Ling sounded in the study.

"Master, the experimental data analyzed by the SAS data platform is abnormal!"

Hearing this voice, Xu Chuan became alert and asked quickly: "Abnormal, what data is wrong?"

This damn problem has been tormenting him for a long time.

More importantly, no problem was found at all, and the feeling of no progress was too uncomfortable for him.

"Comparison of EEG event-related potential signal data shows that there is currently a phase-locked constant waveform data that exceeds the average value, reaching 207.76%."

Hearing this, Xu Chuan quickly walked to the computer and said, "Pull it out, let me see!"

"Here, this is it!"

On the computer screen, Xiao Ling quickly extracted the abnormal data from the analysis data.

At first glance, an EEG brain wave image came into view. Staring at the experimental data in front of him, Xu Chuan's eyes were a little strange.

"If I remember correctly, this seems to be the ERP potential signal data in the EEG brain wave signal? The fluctuation of 2-10 microvolts, if I remember correctly, this fluctuation seems to be related to basic low-level perception?"

"Yes, master."

Xiao Ling's voice sounded in the study, and said with some anthropomorphic emotions: "I just checked the brain wave signal data you provided. This type of electrical signal fluctuation is recorded in the data as irregular periodic brain wave fluctuation data spontaneously generated by the human subconscious."

"It is extracted from continuous EEG data and is a subconscious stimulus response neural signal to specific stimuli, such as pictures or text seen on the computer screen."

Hearing Xiao Ling's words, Xu Chuan subconsciously touched his chin.

He seemed to know where the problem was.

However, this still requires a lot of data analysis of ERP potential signals to verify its idea.

After thinking for a while, he quickly said, "Xiao Ling, stop other analysis work and focus on the 2 to 10 microvolt data of EEG neurological characteristics."

"I need a comprehensive judgment data on the constant waveform and latency waveform of the ERP potential signal. If nothing goes wrong, I may find the problem!"

"Received! (ω)"

The fluctuation of 2 to 10 microvolts is the fluctuation signal of the ERP event-related potential in the EEG brain wave signal.

It is weaker than spontaneous EEG, usually only 2 to 10 microvolts. When collecting information data, because the signal data is weaker, it is usually buried in the spontaneous EEG brain wave.

Therefore, it is generally necessary to use specific technical means to analyze ERP signals.

However, for Xu Chuan, this was not a difficult task. He did not even have to do it himself. Xiao Ling could directly use various tools in the supercomputer to complete it.

With a specific analysis direction, the relevant data analysis work was quickly completed with the support of the small supercomputer.

Looking at the analysis results on the screen, Xu Chuan's mouth curled up a little.

Sure enough, his guess was right. The problem lies in the weaker ERP event-related potential signal.

The activities of the brain and the human body are far more magical than he imagined.

Under normal circumstances, the medical community generally believes that the neural signals that control trunk movement are dominated by active signals such as 8-100Hz alpha waves, beta waves, and gamma waves.

For example, 8-13Hz alpha waves generally control brain electrical activity in a relaxed state.

For example, when people sit quietly, relax, and close their eyes, the alpha waves will gradually increase.

At the same time, alpha waves are also related to the execution of cognitive processing tasks, hand-eye coordination, and emotional regulation.

The 13-30Hz beta waves are mostly related to active cognitive processing tasks, such as thinking, decision-making, and attention.

Especially in the execution of muscle movement, beta waves can provide information about the speed of different muscle segments and grip control.

There are also gamma waves that can reflect muscle tension and muscle shortening rate.

These active brain waves are generally considered to be the main electrical signals that control human activities.

And those lower-level natural weak brain waves usually control the subconscious activities of the human body.

But from the current situation, the relationship between the two may be more subtle.

Movement is not entirely completed by active brain wave signals. In the intense activities of the human body, weak-frequency brain wave signals will also make a certain degree of command to the muscles!

Quickly sorting out the analysis data on the computer, Xu Chuan raised a curve at the corner of his mouth.

Although the research on brain waves is not very in-depth, his mathematical intuition tells him that the problem lies here!

PS: I don’t understand biological things very well. If there are any problems, or if there are big guys in this field who are reading, don’t mind if I write it wrong.

I was also confused by this. I read a lot of information, but I was confused.

I won’t make it so complicated later. Just enjoy it directlyψ(｀)ψ