My 1999

The first is to reduce costs.

From the perspective of medical care itself, there are two important reasons for the high cost of medical care.

One is that the development cycle of drugs is long and the cost is too high;

The second is that the cost of training medical personnel is too high.

Now, most of the new drugs in the world come from the United States.

It takes 20 years for a new drug in the United States to go from research and development to market, and $2 billion needs to be invested in this process.

According to the US patent law, the entire patent period is only 20 years from the date of application.

However, the application for a patent does not start from the day the drug is launched, but more than ten years before the drug is launched.

That is to say, after the drug is launched, the patent protection period is only a few years.

Through our comprehensive investigation of pharmaceutical companies such as Johnson \u0026 Johnson, Roche, and Pfizer, a drug can usually enjoy patent protection for only 7 years.

That is to say, if a new drug can be successfully developed, it will only have seven years of exclusive sales to earn back the cost.

Therefore, each new drug is very expensive.

It takes 13 years to train a specialist in the United States.

In China, it takes 10 years for college students and undergraduates to become a chief physician, 9 years for a master's degree in medicine, and 8 years for a doctor of medicine.

From the perspective of return on investment, since the investment of time and money is so huge, they must have high income to be cost-effective.

So how can we use artificial intelligence to change the medical industry?

Let's take an example.

We naturally think that we should find experienced doctors to see a doctor.

The accumulation of their experience is a process of learning through cases, and no matter how fast people learn, they can't be faster than computers.

It is difficult for a radiologist to read and study more than 100,000 cases in his lifetime.

But computers can easily learn from millions of cases.

Compared with doctors, computers have three major advantages in diagnosis and surgery.

First, they are very unlikely to miss judgments and make mistakes, which means that they can successfully discover some situations that doctors ignore.

Second, their accuracy is very high, and it improves very quickly as the amount of data (cases) increases.

Finally, what humans do not have, these intelligent programs are very stable, and they are not affected by emotions like humans.

And the cost of these intelligent programs is usually less than one percent of manual labor.

With powerful medical artificial intelligence programs, combined with medical robots, patients far away in remote counties can enjoy top medical services.

Thus solving the problem of uneven medical resources.

Finally, let's talk about the changes that artificial intelligence has brought to the pharmaceutical industry.

Today, humans have invested much more money in cancer than in Apollo's moon landing or voice recognition.

But why is cancer still so difficult to cure?

Because cancer cells are the result of errors in the genes of animals and human cells during replication, not from outside the body.

In other words, cancer is a genetic disease, not a virus.

Today, the most effective way to treat cancer is to use anticancer drugs developed using genetic technology to treat it.

From a mechanistic point of view, it is to find the diseased genes and kill the corresponding cancer cells.

However.

Even if different people have the same cancer, the genes of their cancer cells may not be the same.

Therefore, an anticancer drug may work for some patients, but not for other patients.

In fact, when most doctors give cancer patients drugs, they need to compare the patients' genes to determine whether a certain anticancer drug can be used.

The second difficulty in treating cancer, and the most fundamental difficulty, is that the replication of cancer cells themselves can also go wrong.

This is not difficult to understand. If a gene goes wrong once during replication, it will go wrong a second time.

In this way, the anticancer drugs that were originally effective become ineffective.

When anticancer drugs kill cancer cells, they may not kill all of them.

Even if there is only one cancer cell left that is not killed, it can still reproduce rapidly and may have new gene mutations.

So we often hear this kind of story.

A relative or friend with cancer has controlled the disease for a long time, but suddenly relapses overnight, and the drug does not work, and he soon passes away.

The reason is that the changes in genes make the original anticancer drugs ineffective.

Because the mutations in cancer cell genes are related to people, and they may change again and again.

Therefore, in order to completely solve the problem, it is necessary to design specific anticancer drugs for different patients, and to develop new drugs based on every new change in the patient's cancer cells.

In other words, as long as the speed of developing new drugs can keep up with the changes in cancer cells.

Then even if all cancer cells cannot be completely killed, patients can still coexist with cancer for a long time.

In theory, this method is feasible.

But the cost of doing so is too high.

First, there must be a dedicated R\u0026D team to develop drugs around each patient, and the R\u0026D speed must be fast enough.

Second, it costs at least $1 billion per person.

Therefore, this seemingly possible method is not worth promoting.

So where is the way out? "

Three words appeared on the projection screen:

Big Data.

"At present, we know that the various genetic errors that may cause tumors are only in the order of 'ten thousand', and the known cancers are only in the order of 'hundred'.

In other words, even if all possible combinations of malignant gene replication errors and various cancers are taken into account, there are only millions to tens of millions of types.

This order of magnitude is very small in the IT field, but it is almost infinite in the medical field.

If we can use big data technology to find various combinations that actually cause cancer among these no more than tens of millions of combinations, and find corresponding drugs for each of these combinations, then all possible lesions of all people can be treated.

For different lesions of different people, just choose a drug from the drug library.

In this way, cancer can be controlled.

Although the total R\u0026D cost of thousands of drugs is not low, if it is distributed to every cancer patient in the world, it will not be that high.

The same principle applies to other diseases. "

Hearing his words, the eyes of the audience who were originally expectant soon dimmed.

They are all smart people. Although they think Xu Liang's strategy is feasible, it is still unknown when such a database will be established.

Even the day they die may not come.

So the expectation quickly faded.

Xu Liang saw it, but did not explain much.

He did not see this day before his rebirth, so he could not explain it at all.

But he came to talk about big data.

As long as the big guys understand the importance of big data, the purpose will be achieved.

Everything else is secondary.

Of course, if the big guys in the audience get excited and take the initiative to build this database, the people of the whole country will benefit.

He, Xu, will definitely help.

But it is highly unlikely.

The big guys also emphasize efficiency and return on investment.

"According to the data survey of Hanhua and Hongmeng Bing, only about one-seventh of the clinically proven effective drugs in the United States can eventually pass the entire approval process of the "Food and Drug Administration" and finally be listed.

The remaining six-sevenths of the drugs, although they do have good effects on some patients when used in a small range.

But when used on a large number of patients, the average effect is not significant, so they are rejected by the "Food and Drug Administration".

So, if we can find a specific group of people, we can reuse these "waste drugs".

In the future, there may be different drugs to treat a disease, and different people will have different special drugs.

But to achieve this goal, the country needs to step forward and establish a national drug database.

I believe that after this database is truly built, my country's medical expenses will be significantly reduced, and life expectancy will be greatly improved.

Even cancer is no longer a terminal illness. "

Looking at the distracted bosses in the audience, Xu Liang is ready to give them a little future shock.

"Speaking of medical care, let's talk about something off topic.

Can humans live forever? "

As soon as these words came out, the faces of the big guys changed instantly.

They looked straight at him.

Standing on the stage, Xu Liang could clearly see their dilated pupils and feel their suddenly rapid breathing.

The higher the status and the more wealth one has, the more he desires longevity.

This is also the reason why so many emperors in ancient times pursued immortality.

"We talked about cancer before. Even if humans solve cancer, they can only extend the average life expectancy by 4 years.

The significance of treating cancer is not as great as the public imagines.

The biggest challenge facing human longevity is the problem of aging.

As long as people live long enough, they will eventually face the trouble of Alzheimer's disease, without exception.

According to research by the Massachusetts Institute of Technology, in the past decade, the mortality rates of cancer, AIDS, heart disease and stroke have all declined worldwide, but the mortality rate caused by Alzheimer's disease has increased by 40%.

So the key to extending life is to find the aging gene.

How to find it?

The most fundamental solution is big data and artificial intelligence.

We can use the method of solving cancer to study the aging gene, find the real pathogenic mechanism, and change it through gene repair and editing technology.

Although it seems that this method is not yet of practical significance.

But I believe that with the development of big data and artificial intelligence, this goal will be closer and closer to us. ”

Paused.

“My speech today ends here. Thank you very much for listening to my speech. Thank you.”