Artificial intelligence medicine: More and more scientists begin to study the gut and its relationship with the brain in order to treat integrative diseases such as inflammatory bowel disease, which is called brain gut axis. Such research may promote a breakthrough in medical technology, but it is too complex to use gene sequences, and it increasingly relies on algorithm modeling.
At present, the most advanced computational model research laboratory uses DNA sequence confirmation markers, namely “short DNA sequences”. This shortcut does not require the identification of the entire genome. It enables researchers to understand the overall situation of a single intestinal microbiome more quickly. Through this detailed intestinal research, researchers are getting closer to the future of providing truly personalized intervention for patients.
All these will provide immeasurable help for fecal bacteria transplantation, and this technology is the most exciting and eye-catching medical breakthrough in the field of brain intestinal axis. This operation was once regarded as a radical non mainstream treatment, but now more doctors are willing to try this method. They will take a stool sample from a healthy population and then put the sample into the patient’s microbiome, so that the sample can stimulate the development of the system towards a healthier intestinal composition. Fecal bacterial transplantation has been shown to be very successful in treating a fatal infection caused by a bacterium called Clostridium difficile. The researchers hope to use this technology to find more effective treatments for inflammatory bowel disease.
However, without a deeper understanding of the brain gut axis, such attempts can only achieve unstable results at most. Researchers still do not know the reason for the success of fecal bacteria transplantation. The possible reason is that after the bacteria in the feces of healthy people enter the patient’s intestine, they begin to act as a guard against the pathogenic virus, and the more likely reason is that the above situation works together with other unknown factors.
A research article published by Science Translational Medicine on March 1, 2017 pointed out that the feces of rats with inflammatory bowel disease were extracted and transplanted into the microbiome of healthy rats. As a result, healthy rats not only had mild inflammation, but also became anxious. Similar anxiety symptoms are common in patients with inflammatory bowel disease. This symptom contains a key mystery.
If we solve it, we can understand the role of intestinal health on the mental function of the whole human body. This study led by Giada De Palma shows that fecal bacteria transplantation can be used to treat inflammatory bowel disease, but researchers in this field believe that a deeper understanding of the intestine and the relationship between the intestine and the brain is needed.
Only through computing and dynamic modeling tools can we understand these complex and interrelated systems and their relationship with personal health care in more detail. If researchers want to go beyond the trial and error stage of fecal bacteria transplantation, they need to use a more accurate and mechanized microbiota model brought about by the recent development of artificial intelligence.
While researchers are preparing to use algorithmic models to study inflammatory bowel disease, a group of innovative and start-up companies in San Ramon, California, such as healthtell, are planning to use artificial intelligence systems to develop a diagnostic tool for the treatment of another autoimmune disease: lupus. The cause of this chronic inflammatory disease is that the human immune system attacks its own tissues and organs.
Sadly for patients, this disease is often misdiagnosed at the beginning, so it can only be found and treated after serious damage to body organs. I recently talked with bill Colston, founder of healthtell, about the current situation of medical diagnosis technology, and discussed how to enhance diagnosis through artificial intelligence in the future. Healthtell’s technology has evolved from a simple genomic test to a diagnostic test that enables us to understand the overall situation of the immune system.
Colston told me: “researchers have recently learned that over millions of years, the immune system has evolved a very specific and complex disease response mechanism. Its response depends on the type of disease.” in healthtell’s technical research work, he did not focus on the direct detection of disease, but tried to measure the body’s response to disease.
The immune system responds immediately to pathogens. Relatively speaking, a large number of antibodies will appear in the patient’s body, so it is easy to detect the disease and understand the overall situation of the body.
Colston explained: “At present, there are only two kinds of diagnosis. You are either ‘sick’ or ‘healthy’. If you are sick, you need to see a doctor, but we find that this is not the way the disease develops. The disease develops for a long time, and your body will not start to resist until the disease is finally going to win and you feel uncomfortable. I think the artificial intelligence system is really interesting That is, it can predict the process of human body from health to disease. ”
Colston and his team at healthtell bet that measuring immune system response and other biological indicators responsible for human management, such as metabolism, protein and genome, is the best way to understand the overall situation of the human body.
Colston told me: “At present, we have to wait until we are really ill to control the disease. Eventually, we either get rid of the disease or make the disease worse. We hope to develop better therapies in advance or make positive changes to our lifestyle. If you can develop cheap and scalable diagnostic methods for only $10 or $15, and patients can do it regularly by themselves Diagnosis, then you can input these diagnostic methods into the artificial intelligence system so that it can predict the changes of body state. ”
He believes that this knowledge will enable patients to make changes before they get sick. Like many modern diseases, early intervention can more effectively treat autoimmune diseases. The system Colston said may take more than 10 years to appear, but we have been in “Watson” of IBM We can see the initial form of the system: using artificial intelligence to enhance diagnosis. If a system wants to make full use of Colston’s technology, it must remove a major obstacle, that is, the lack of data. Artificial intelligence works normally in driverless vehicles and other fields because there are various embedded sensors all over the vehicle and its surrounding environment, which carry out thousands of real-time measurements at the same time However, we can not deploy enough data points in the human body to predict what will happen in the future, nor do we get enough clear results to train the artificial intelligence system.
Healthtell is where Colston’s technology is ready to be applied. The first challenge it faces is to develop a measurement system. As we can see in the deep learning algorithm, when we have enough data, that is, when we obtain sufficient long-term measurement results from a large enough population, we can obtain enough data depth and breadth to realize machine learning Accurate diagnosis.
In Colston’s view, the future of artificial intelligence does not mean the complete use of computers instead of doctors. It is just a way to enhance human intelligence with more knowledge.
“Artificial intelligence equipped with this more intelligent sensor and treatment technology can provide long-term longitudinal data for physicians to make more accurate diagnosis. Doctors can better determine when patients should be treated. Our diagnosis methods for various diseases are still very general, and doctors hope to get more such information,” Colston said
Colston believes that applying advanced artificial intelligence systems to diagnosis is a feasible way to change this trend. It can fundamentally improve efficiency and enable primary care doctors to spend more time communicating with patients, because artificial intelligence systems can help them interpret more complex data and compare these data with other patients with similar symptoms Line comparison.
Of course, like many other fields, the use of artificial intelligence to improve the ability of professional doctors has also aroused social anxiety. Some people worry that the human side of the medical industry, that is, care, care and healing between people, will be replaced by emotionless technical solutions. Colston at Lawrence Livermore National Laboratory (a competitor of Los Alamos National Laboratory) has been a researcher for 20 years and is committed to developing technologies to combat biological and chemical terrorism. Based on his experience in new anti-terrorism technologies, he believes that there are misunderstandings in these concerns
“Technology has never been evil,” Colston told me, referring to a recent technological breakthrough in prenatal testing for Down syndrome, which can replace the original amniocentesis. “After the advent of this technology, people can detect Down’s syndrome through a drop of blood instead of inserting a needle into the amniotic cavity. Many people say that this is not a good thing because it forces parents to make moral decisions before the child is born, but we have been doing this before, and it is an invasive method that may harm the fetus and pregnant women. With the new detection method, we can obtain this information in a safer way. ”
Colston concluded, “getting more information can’t be a bad thing. What really determines whether it’s good or bad is how we choose to use this important knowledge.”
Artificial intelligence as the decoder of human life
This gene sequencing and diagnostic innovation made by healthtell has transformed artificial intelligence into a human biological decoder. The most typical example is the recent development of gene editing technology, such as the use of CRISPR – cas9. This technology enables scientists to inject modified proteins into the human body, so as to cut genes and recombine them like a pair of scissors On October 28, 2016, researchers from Sichuan University extracted human immune cells and edited them using CRISPR technology. 7 they successfully removed the gene that allows the body to attack healthy cells. By cutting this gene in lung cancer patients and reintroducing it into cells, researchers hope that the modified cells can reduce cancer cells and aggressive behavior.
Each letter of CRISPR stands for “clustered, regularly spaced, short palindromes and repeated sequences” , it is based on the discovery that bacterial cells can recognize invading viruses and cut them into DNA. Later, scientists used this technology discovered 10 years ago to edit DNA sequences far more complex than bacteria. Now, the research results of Chinese scientists enable us to start editing our own DNA.
This is just the beginning of human function enhancement. When we combine artificial intelligence with neural networks, we can use machine learning to interpret signals and use wireless devices to replace the connection between the brain and our limbs and muscles. At the annual meeting of the society of neuroscience in November 2016, scientists announced that women with atresia syndrome can move their computers only by their own ideas Cursor on the computer interface.
Locked in syndrome is a disease in which the patient is completely paralyzed but his mental sensitivity is not affected. This disease is described in the 1997 book the dividing bell and the butterfly, which was made into a film by Julian Schnabel in 2007 and won many awards. This story written by Jean – Dominique Bauby journalist and fashion director tells him in detail It also describes how he wrote a book with only his left eye moving.
This latest invention in computer and cursor has opened up a new way beyond the blink activity and the conversion of human motion information. The woman mentioned above, we only know her name is Hb, and she has implanted electrodes under her skull. Although the electrodes do not pass through the brain tissue, they can make enough contact with Hb’s brain to accurately reflect her brain wave activity. The surgeon will Electrodes under her skull are connected to a device in her chest, which is wirelessly connected to a tablet computer.
The machine learning algorithm is responsible for the final distinction of Hb β Brain wave sum γ Brain waves, and learned which brain activity is related to tiny movements such as pinching fingers. With the help of these algorithms, Hb will soon be able to move the cursor on the tablet only by mind. When researchers put a large alphabet on Hb’s tablet screen, she can input it with the feeling of clicking the mouse and selecting letters. Machine learning can recognize her meaning Chart and record her choices on the screen. Hb can slowly communicate with the world at the speed of one or two letters per minute. She is no longer a “locked person”
Whether we use machine learning to resist the tyranny of unknown diseases, gene editing technology to modify our DNA, or liberate patients from atresia syndrome through neural signals and machine learning, we must use artificial intelligence to further enhance the medical and pharmaceutical industry in the next 10 years.
At present, there are many disputes in this regard, including the fear of designing babies The risk of superhuman species or the unexpected consequences of neural device invasion, and moral discussion is also necessary. Nevertheless, when we discuss whether we need to develop artificial intelligence, we have brought our society into a wrong choice. The development of science and technology is inevitable, the progress of thought can not be controlled, and one breakthrough will lead to another Breach.
As Bill Colston said, “you can’t slow down the development of a technology because you are worried about its specific uses. Anyway, it will always appear in the end, because someone will invent it elsewhere. So the better way is to control it, control it and apply it as soon as possible. There is no good or evil in science and technology itself.”