In any attempt to think ahead to the future, it is worthwhile to examine the small and unusual clues that can point to the changes in perception that are on the way - to the big revolutions that consist of dozens and hundreds of such clues. In this entry, I want to go over three small clues - three developments - that were published in the last week, and try to understand what their greater meaning is for the future of medicine
In any attempt to think ahead to the future, it is worthwhile to examine the small and unusual clues that can point to the changes in perception that are on the way - to the big revolutions that consist of dozens and hundreds of such clues. In this entry, I want to go over three small clues - three developments - that were published in the last week, and try to understand what their greater meaning is for the future of medicine.
When the surgeon works on the other side of town
The first hint was published this week in the respected medical journal E-Clinical Medicine, which is a subsidiary journal of the Lancet - the most prestigious medical magazine in the world. Heart surgeon Tahas M. Patel has successfully performed five heart surgeries using a robot. Apparently there is nothing new in this - doctors have been using robotic systems for heart surgery since 2011. They operate the robot as an extension of their hands and fingers, and the basic artificial intelligence in the robot curbs the small, natural vibrations of the doctor's hands, and sometimes even warns them of possible dangers and problems.
The unusual thing here is that Patel was not in the same room with the patients while he was controlling the robot. He wasn't even in the same hospital. In fact, he controlled the robotic system over the Internet, staying thirty kilometers away from it. This is the first demonstration in history of robotic heart surgery performed at a significant distance from the patient.
This is just one more example of the growing trend of "remote medicine" or Telemedicine. It joins other similar clues: during the Ebola epidemic in Africa, for example, patients were treated by medical teams operating from the University of Virginia. The therapists in the field could even show the doctors on the other side of the world - literally - the results of the tests and get an immediate opinion from them. In France, Italy and Portugal, you can already find dozens of "telemedicine cabins" where patients can contact a doctor who sits hundreds of kilometers away from them, but is able - using the equipment in the room - to check their pulse, temperature and blood oxygen level. He is even able to do X-rays and hearing tests for patients.
telehealth room.jpg
Image of the "remote medicine room" of the H4D company. Dozens of such rooms are currently being tested in several European countries[3].
If there's one thing we can be sure of, it's that medical tools will only continue to improve in the coming years. They will become more connected and networked. The arrival of the 5G Internet will ensure that the connection of these devices to the Internet will be safer and faster than ever before. It is not an exaggeration to think of a future that will arrive in ten years or less, in which many surgeries can be performed by surgeons sitting thousands of kilometers away from the operating room, and with a minimal presence of assisting medical staff.
And why stop here? If the robots themselves become smarter and more sophisticated, it is not an exaggeration to think of autonomous machines that could even perform certain medical procedures on their own. It is not yet clear whether the patients will agree to be treated by robots, but it is certainly possible to argue that in many parts of the world, patients would prefer to undergo cheap and effective robotic surgery, which will be performed in a short time, than to wait for long weeks and months to receive similar treatment from an expert surgeon. And in any case, when the operation is performed by a surgeon sitting in a remote place, and that the patient will never meet him face to face, the whole idea that the operated and the surgeons should know each other personally begins to lose meaning. So why not agree to surgery performed by a robot, if its success rates are the same as those of a human doctor - or higher?
But why stop here?
Let's take for a moment the idea of "remote medicine rooms" which is already starting to take root in Europe. These are, in fact, robotic rooms. Why not add wheels and autonomy to them on the road, and create the super ambulance of the future? I call them super-ambulances because they will be more like a hospital on wheels: long, wide trucks that contain much of the equipment that can be found in a hospital, and will be able to reach any patient quickly - and provide him with autonomous robotic care, or with robots that will be controlled remotely. Such trucks will be of enormous value in disaster areas, or even in the periphery of Israel, where it sometimes takes tens of minutes for a patient to reach the hospital.
This is the first hint, which means that medicine will become more available, more effective, cheaper, and much more equal.
But he is not the only clue.
The machine that develops drugs
The second clue opens with a gloomy tone: one of the most celebrated artificial intelligence engines today - IBM's Watson - failed to help pharmaceutical companies develop new drugs, and IBM chose to stop the service in this area in mid-2019.
The good news is that where IBM failed, someone else is starting to succeed.
Last week, an article was published in the respected scientific journal Nature Biotechnology, which described the results of an intervention between the researchers at the biomedical company Insilico Medicine (Insilico Medicine), and their partners in the field of drug development[5]. A certain pharmaceutical company challenged insilico researchers to identify new molecules that could inhibit the action of a protein involved in tissue scarring.
This may not sound like such a difficult task, but it is one of the biggest limitations today for the development of new drugs. In the early stages of modern medicine, researchers used molecules found in nature: the bark of the chincha tree, for example, contained a molecule called quinine that was useful in fighting malaria. Mushrooms secreted certain molecules that killed harmful bacteria - and these were used as antibiotics. In the second step, researchers took the original molecules and 'tweaked' them to better suit the target. They added different atoms to them here and there to allow them to dissolve more easily in the blood, for example, or they embedded the molecules in tiny cages so that they could not immediately affect the body.
In the third stage, the researchers began to try to develop completely new molecules. It is difficult to explain how challenging this task is. Medicines usually affect (and this is a very gross abstraction and generalization, but let's say) proteins in the body. Each drug is supposed to affect a certain protein and make it stop working, or work in a different way. In this way, certain antibiotics are able to stop the activity of bacterial proteins, and other drugs can bind to protein toxins released by bacteria and neutralize them. The problem is that molecules that affect proteins consist of hundreds, thousands and sometimes millions of atoms that connect to each other in different configurations.
Have you ever tried to put together a XNUMXD puzzle? Did you succeed? Well done. Now think of a XNUMXD puzzle with millions of pieces, where every attempt you make to fit one of the pieces changes the other pieces. Oh, and you're not even sure what the end result should look like.
Do you understand the magnitude of the problem? But here is the big prize: whoever manages to solve this particular challenge and develop new and useful molecules, will save the world. And no, I'm not exaggerating. At least not by much.
Imagine we had machines that could break up oil slicks in the sea. If we could engineer molecules to do a specific action, we could create tiny machines - the size of molecules - to do that too. We could engineer molecules that would break down feces and urine into basic atoms, then reassemble them as nutrients. We could engineer complex giant molecules that would recognize cancer cells and kill them. Or they would cut into the living tissue to perform surgeries without using a knife. And so on, and so on. In fact, engineered molecules would allow us to realize the original nanotechnology vision and achieve complete control over matter. any material.
Nice, right? But again - developing new molecules is hard work. Pharmaceutical companies have been employing biochemistry experts for twenty years trying to develop new molecules, and the successes are few and require a lot of time.
So why not transfer the work to artificial intelligence?
This, in fact, is the business model of Insilico Medicine (which, as the name implies - "medicine on silicon"). After deciding they wanted to find molecules that could bind to a particular protein, they ran their AI engine on all past research on that protein and the molecules that bind to it. They required the AI engine to develop ideas for new molecules, similar to the way a human chemist thinks. And as Adam Renslow, professor of biological chemistry, said in an interview with Wired magazine -
"It's cool to see an artificial intelligence trained to think a bit like a medicinal chemist... this algorithm involves a creative process, not data mining."
Within three weeks, the creative AI engine produced 30,000 new ideas for molecules that should have suited the researchers' goal. The researchers chose to focus on six of these ideas and synthesized them in the laboratory. Four of the molecules proved their effectiveness in tests in vitro. Two were tested on biological cells, and the most promising was tested in experiments on live mice. And guess what? It did exactly what they wanted it to do - at a level of efficiency high enough to make researchers raise their eyebrows and suggest starting the process to register the molecule as an actual drug.
Everyone is starting to understand that artificial intelligence can speed up the drug development process significantly. According to information collected by Bloomberg, investors have poured more than a billion dollars into companies that develop artificial intelligence to discover new drugs. The successes come slowly, but they definitely appear. In mid-2019, the company Exscientia announced that it had produced a new molecule that could help treat severe pulmonary disease (COPD), And the pharmaceutical giant Glaxo-Klein-Smith is currently trying to register the molecule as a drug. Insilico Medicine itself is currently focusing on aging diseases, such as cancer, and is trying to use the same artificial intelligence engine to find solutions for these diseases.
Let's pour some cold water on the enthusiasm for a moment: there is still a long way to go before we have artificial intelligence engines that can develop new 'perfect' molecules. Even the CEO of Insilico admits that he expects it will take a few more years before the molecules identified by the company can reach clinical trials. But the path to success is already clear: all that is needed is superhuman intelligence, at least when it comes to connecting the pieces of a 3D-submolecular-puzzle- shape to each other. And since artificial intelligence only continues to develop and improve, our abilities to produce more effective molecules and drugs, faster, are also expected to advance.
Let's leave aside for a moment the benefit that genetically engineered molecules can bring to the human race. It is equally important to understand the significance of the involvement of artificial intelligence in medical development and research processes. We see, in fact, that artificial intelligence is beginning to become a tool of enormous value in these processes. The medical researchers of the future will use artificial intelligence engines to generate ideas for new drugs in a short time and at low cost. And it is not an exaggeration to expect that a large part of the research processes will become almost completely automated, so that the artificial intelligence engine will be able to propose ideas for new molecules - and then also synthesize them and test them on its own, with minimal human involvement in the process.
Which raises the question - what will be left for humans to do in the field of medicine? After all, we eliminated the need for flesh and blood doctors, and now we also eliminated the need for a large part of the medical researchers. What, then, will be the role of human doctors - if there will be any at all?
And this is where the third clue comes into play.
The system that makes everyone feel better
A week ago an article was published in Fortune magazine describing the way in which the service giant Deloitte tries to help companies retain their employees. It turns out that one of the most annoying things for employees is "time vampires" - a variety of routine, boring and tedious activities that suck the employees' time and do not allow them to do the work for which they were hired.
Two of the biggest time wasters are located in the human resources and technical support departments of companies. A 2019 study showed that about half of employees worldwide "struggle to get answers to basic questions" from these two departments.
worker content.jpg
Chart from ServiceNow's The Employee Experience Imperative report. Link to the source here.
Deloitte's solution? Simple: artificial intelligence. Deloitte offers an artificial intelligence engine that goes through large amounts of information in the company, and provides immediate answers to employees through chatbots. According to the manager responsible for the software, people can get help directly from the bots, "without having to interact with another person."[9]
I admit that at first glance it seems that this is another case where artificial intelligence replaces workers, but here is the important point that is made in the article: Deloitte's bots take on the boring and repetitive tasks in every company. The result, according to the article, is that
"People can invest most of their time and talents in more complex and interesting work, which requires human contact."
Another manager at Deloitte claimed on behalf of his employees that instead of gathering information themselves and having to deal with routine and boring tasks, they could -
"… spend a lot more time in one-on-one meetings with clients, so they can better understand each person's financial goals, concerns and needs. … That's the fun part."
We will probably see a similar pattern of activity in the field of health and medical services as well. As services become more comprehensively automated, people will be able to get answers to simple - and sometimes complex - questions quickly and efficiently from artificial intelligence engines. The doctor's role will be to provide the experience of human care and service. The doctor of the future will not stare at the computer screen, type on the keyboard and read the result to the patient. At least, not if he wants to have a job. The doctors of the future will invest in each of their patients, have in-depth conversations with them, try to understand their concerns and needs - and will also help them choose between the various options for treatment that the computer offers them.
In other words, they will provide sensitive and humane care, which will be even more important than the analytical role they currently play.
And if this is truly the future of medicine - we are blessed.
More of the topic in Hayadan:
2 תגובות
It's fun to read Csena.
As always, Dr.Roey Tsezana's articles are fascinating and interesting. Roi's ability to present a complex idea in a clear language understandable to everyone is the reason for the success of his excellent (and highly recommended) books on the future.
If the special writing is also accompanied by a joking tone, then it only makes it easier and increases the pleasure of the reader.
Continue to delight us…