Chapter 912 The complex brain wave problem
If he was asked to conduct research on the experimental data of brain-computer interfaces and bionic robotic arms, Xu Chuan believed that he did not have the ability.
There are specialties in science, and even all-round scientists like Newton and Leonardo da Vinci, who cover many fields, still have areas they don't understand.
The biological field was indeed outside the scope of his academic research.
However, if he is just asked to mine data from these experimental data and experimental images and use mathematical tools to analyze the patterns, he can still do it.
Time passed little by little, and the sky outside the window gradually dimmed.
It took Xu Chuan an afternoon to complete the preliminary compilation of the data and develop the ion current dynamics formula based on the Hodgkin-Huxley model.
After processing these, he sent the initially completed formula and data to Chuanhai Network Technology Company.
He took out his cell phone from his pocket and opened the address book. After finding the familiar name from the special attention list, he dialed it.
The phone rang twice and was quickly connected.
"Hello."
The phone was connected and a soft voice rang in his ears. A smile appeared on Xu Chuan's face and he said:
"I sent you an email. After work tomorrow, I will find a few people to help build a mathematical model based on what is in the email."
"Well, okay, I'll take a look later."
Xu Chuan smiled and said: "I'm going to have to work hard for you again. I'll treat you to a big dinner in two days."
"Then I'll wait."
After chatting for a while, Xu Chuan hung up the phone without thinking much.
At the same time, on the other side, in a high-end residence in the Qixia Mountain 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 head again.
Chuanhai Network Technology Company is very efficient.
It only took four days, and a complete mathematical model was delivered on the fifth day after Xu Chuan provided the modeling information.
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 a small one, can process all kinds of data extremely quickly. This is something that ordinary computers and even expensive servers cannot match.
With the help of Xiaoling, an AI academic assistant, it took less than an hour to complete the calculation of the relevant data.
"Sure enough, the problem is not the data conversion between the quantum mathematical model and the traditional multi-electrode array reset mathematical model."
Staring at the calculation data compiled on the screen, Xu Chuan murmured something softly with a look of 'as expected' in his eyes.
As he expected, there was no conflict between the quantum mathematical simulation model he had built previously and the traditional multi-electrode array reset mathematical model built by Xu Xiao himself.
The data conversion between the two is quite smooth, and there is no need to enlarge or modify the experimental data.
"If the problem doesn't appear here, what exactly is causing the interference?"
His eyes fell on the experimental data, with an interested look on Xu Chuan's face.
He had seen the previous test experiments of bionic robotic arms and robotic legs, and the problems Xu Xiao mentioned did exist.
After the brain wave signal sensed by the brain nerve chip was converted into an electrical signal and transmitted to the bionic robotic arm, something abnormal did occur.
After looking through these experimental data, Xu Chuan fell into deep thought.
Although brain-computer interface technology is not in his research field, he still understands some general situations.
Put aside ethical difficulties such as the blurring of human-machine boundaries, the protection of mental privacy and autonomy, and the ethical boundaries of neural intervention.
There are two main problems with brain-computer interface technology.
One is the biocompatibility issue of implant materials.
For example, the materials used in implantable brain-computer interfaces may cause brain rejection or cause brain damage due to movement, etc.
After all, the brain is the most sophisticated of all the organs in the human body.
Encountering any external force may lead to serious problems such as brain damage and brain death.
However, this issue does not need to be considered at present, because in theory, the biocompatibility of materials will not cause abnormalities in the conversion and transmission of neural signals.
"Could it be that the capture of brainwave signals is not comprehensive?"
Looking through the experimental data in the computer, an idea came to Xu Chuan's mind.
For brain-computer interface technology, the limitations of neural signal capture are a considerable problem.
An average person's brain has about 86 billion neural units, and currently humans can capture only a part of them.
This means that there is still a large amount of neural signals that cannot be effectively used.
In particular, neural networks in the brain are not simple linear superpositions, but involve complex nonlinear relationships.
This makes simultaneous encodings difficult to parse.
Distinguishing the coding of brain neural signals for specific behaviors from the coding of other behaviors remains a challenge.
Could there be a problem in this regard?
Thinking about it, Xu Chuan clicked on another file in the information Xu Xiao gave him, which contained technology developed by her and the team from Starlight Virtual Technology Company specifically for Starlight brain-computer interface chips.
A two-section RNN architecture, a method for nonlinear dynamic modeling.
This technique uses recurrent neural network architectures and training methods through nonlinear, kinetic modeling, separation and prioritization of behaviorally relevant neural dynamics, and modeling of continuous and intermittent behavioral data.
It can improve the accuracy of neural-behavior prediction, optimize the identification of original local field potentials and other areas that are difficult to achieve with traditional neural signal simulation technology.
However, even if he wants to find problems in these algorithms and experimental data, it will be difficult for him to do so for a while.
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 on neural signals is a bit large.
Not to mention other things, the frequency of beta waves (beta waves) related to thinking, conscious problem solving, and attention to the external world is as high as 14-30 Hz in the normal waking state of the brain rhythm.
It sounds like this data is very small. After all, 14-30 fluctuations per second is nothing to the technology developed by humans.
But if it is combined with the feedback and processing of various external signals by the brain nerves, the data generated will be an extremely huge amount.
Fortunately, for cranial nerve models, most data can be classified by different indicators.
Otherwise, it would be unrealistic to process such huge data through a brain-computer interface chip. In the study, Xu Chuan picked up the already cold tea in the porcelain cup and took a sip to moisten his throat and move his tired eyes.
"Xiao Ling, please help me keep an eye on the data analysis work of the SAS data platform. If there is data that exceeds 5% of the previously completed data, let me know."
"Okay, master! Leave it to Xiaoling!"
In the study, Xiao Ling's voice sounded. Xu Chuan pulled out his chair and walked outside, preparing to take a shower.
I have to say that this is indeed one of the more difficult problems he has encountered in applied mathematics.
Almost all cranial nerve model data and converted electrical signal data have no problems or abnormalities from a mathematical point of view.
Even if the entire data was analyzed and processed through the SAS data platform, no problem was found.
After eliminating possible errors and problems in data conversion between the two mathematical models, for several days, there was basically no new progress in the problems that arise in brain-computer interface technology.
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 taken him more than ten days. If he can't find the problem in these two days, he will put it aside.
Although failing to solve this problem will affect his 'omnipotent' image in Xu Xiao's mind.
But he has many other tasks on his hands, 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 heart after suspending his research, the voice of AI academic assistant Xiao Ling rang out in the study.
"Master, there is an abnormality in the experimental data analyzed by the SAS data platform!"
Hearing this voice, Xu Chuan became energetic and asked quickly: "Abnormal, what data has a problem?"
This damn problem has been bothering him for a long time.
More importantly, no problem was found at all, and the feeling of not making any progress was too uncomfortable for him.
"Data comparison of EEG event-related potential signals, there is currently a phase-locked constant waveform data that exceeds the average, reaching 207.76%."
Hearing this, Xu Chuan quickly walked to the computer and said, "Pull it out and let me take a look!"
"Now, this is it!"
On the computer screen, Xiaoling quickly extracted the abnormal data from the analyzed data.
Upon entering, an EEG brain wave image came into view. Staring at the experimental data in front of him, Xu Chuan had a strange look in his eyes.
"If I remember correctly, this seems to be the ERP potential signal data in the EEG brain wave signal, right? A fluctuation of 2-10 microvolts. If I remember correctly, this fluctuation seems to be related to basic low-level perception?"< br>
"Yes, Master."
Xiao Ling's voice sounded in the study room, and said with some anthropomorphic emotions: "I just checked the brain wave signal data you provided. According to the data, this type of electrical signal fluctuation is spontaneously produced by the human body's subconscious mind. Periodic brain electrical changes fluctuation data”
"It is a subconscious stimulus response neural signal extracted from continuous EEG data to a specific stimulus, such as a picture or text seen on a computer screen."
Hearing Xiao Ling's words, Xu Chuan subconsciously touched his chin.
He seemed to know what the problem was.
However, this still requires a large amount of data analysis of ERP potential signals to verify its idea.
The thoughts in his mind circulated for a while, and he quickly said: "Xiao Ling, suspend other analysis work and focus on the EEG neurological data of 2 to 10 microvolts."
"I need to have comprehensive judgment data on the constant waveform and latency waveform of the ERP potential signal. If nothing else happens, 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 brain electricity, generally only 2 to 10 microvolts. When collecting information data, because the signal data is weaker, it is usually buried in spontaneous EEG.
Therefore, analyzing ERP signals generally requires the use of specific technical means.
But for Xu Chuan, this is not a difficult task. It is not even his turn to do it himself. Xiao Ling can directly use various tools in the supercomputer to complete it.
With a specific analysis direction, and with the support of small supercomputers, the relevant data analysis work was quickly completed.
Looking at the analysis results on the screen, Xu Chuan's mouth curled up slightly.
Sure enough, his guess was right, and the problem lay in the weaker ERP event-related potential signal.
The activities of the brain and human body are far more magical than he imagined.
Under normal circumstances, the medical community generally believes that the nerve signals that control trunk movement are dominated by brain waves such as alpha waves, beta waves, and gamma waves of 8-100 Hz of active signals.
For example, alpha waves of 8-13Hz generally control brain electrical activity in a relaxed state.
For example, when people sit quietly, relax, and close their eyes, 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 beta waves of 13-30Hz are mostly related to active cognitive processing tasks, such as thinking decision-making, attention, etc.
Especially in the execution of muscle movements, beta waves can provide information about the movement speed and grip force control of different sections of the muscle.
There are also gamma waves that reflect muscle tension and the rate of muscle shortening.
These active brain waves are generally considered to be the main electrical signals that control human activity.
Those lower-level natural weak brain electrical signals usually control the subconscious activities of the human body.
But judging from the current situation, the relationship between the two may be more subtle.
Exercise is not all accomplished by active brain wave signals. During strenuous activities of the human body, weak frequency brain wave signals will also command the muscles to a certain extent!
After quickly sorting out the analysis data on the computer, Xu Chuan's mouth curved.
Although the research on brain waves was not very in-depth, his mathematical intuition told him that the problem appeared here!
PS: I don’t know much about biological things, so if there are any questions, or there happens to be a big shot in this field reading it, don’t mind if I make a mistake.
I was also dizzy and confused. I read a lot of information, but I was confused.
It won’t be so complicated later, just enjoy it ψ(｀)ψ
(End of chapter)