Healthcare is one of those fields constantly and exponentially evolving. From the beginning of civilization, we have been looking for ways to save lives and lengthen our lifespan. Each historical moment had its own diseases and problems to deal with: from infant mortality and war injuries to chronic diseases and aging. What we are able to treat today, thanks to the latest advancements of medical technology, would have been inconceivable only a few decades ago. Yet, we are only at the very beginning of a digital revolution. Artificial intelligence and robotics are about to change several aspects of our lives. How is AI changing the healthcare system? What will the future of medicine look like?
Medical training is so complex that it can take up to 10 or more years. This is because medicine is a science that accumulates centuries of discoveries and technological advances. Our broad knowledge of the human body allows us to understand and treat many diseases. But, at the same time, it requires a great deal of knowledge that can only be accumulated through many years of study. This and other factors are leading us to a considerable shortage of medical personnel. Could artificial intelligence help us solve this problem?
Artificial intelligence and machine learning could have several applications in medical training. One of them regards interactive real-life simulations. Imagine: people studying to become doctors or nurses learn the complex anatomy of the human body on 2D anatomical pictures (and on bodies donated to medical science). With the use of digital simulations, they could see functioning body parts in 3D. They could interact with them and see their reactions, recreated by drawing information from the AI’s large database of scenarios. They could learn from making mistakes without real life consequences. The AI-powered simulated training could adjust to the student’s knowledge, adjusting the challenges according to the learning needs. Finally, students could train from anywhere and at any time.
A good example of medical training moving towards this new technology can be found in the Oxford Medical Simulation VR platform. With the use of VR glasses, students can apply their learning to practice, examining, diagnosing and treating virtual patients.
The use of artificial intelligence could also help doctors have better drugs to treat their patients with. As we probably have noticed with the recent pandemic, it may take years to create a drug or vaccine that meets a medical need. On average, it may take from 10 to 15 years for a drug to go from a lab to pharmacy.
Let’s see how drugs work. Usually, a drug is a small chemically synthesized molecule that can bind to a target molecule (usually a protein) involved in the disease. To find these molecules, researchers usually have to analyse large screens of libraries of molecules, until they find one that could potentially work. After that, the drug made from this molecule has to undergo a series of tests before it gets the green light and is sold to the public. This process is both expensive and time-consuming.
How can AI help researchers? Artificial intelligence is capable of analyzing these libraries and getting insights out of it. This leads to three main advantages:
Not only can artificial intelligence help doctors do their job in an easier and more effective way, it is also beneficial from a patient’s perspective. More accessibility, more education, more personalisation - there is a lot to gain out of these future technologies.
Chronic diseases (such as heart diseases, cancer or diabetes) have many causes and we don’t know yet how we can completely avoid them. Yet, there are some behaviours we can adopt that lower the risks. Artificial intelligence could suggest personalized behavioural adaptations by comparing changes in our vital parameters with a wide database of diseases history. In the future, we may be able to have a personal virtual assistant checking on us and keeping us well.
At the moment, there are several apps that help us live a healthier life. Some analyze sleep patterns, some give dietary recommendations and some help us cope with feelings of depression or anxiety. There is still a long way to go, but the direction is the right one: to educate first, in order to avoid avoidable diseases.
Since living standards have been steadily improving, we live much longer than the previous generations and we suffer more often of diseases connected to aging: dementia, heart failure, osteoporosis. In England for example, the most common age of death is 85 years old. This should make us reflect on the importance of institutions and systems for the care of the elderly.
Robots (more specifically AI combined with humanoid design) have the potential to revolutionise end of life care. The use of a robot can allow people to remain independent for longer, reducing the need for hospitalisation and care homes. In fact, the scarcity of caretakers makes them very expensive. The use of humanoid robots could therefore lower costs.
A company that is going in this direction is Devanthro. Rafael Hostettler and his team are working on humanoid robots that can be controlled remotely via wearable devices. Human caregivers will thus be able to take care of several elderly people at the same time, controlling these robots when necessary.
Another essential field in which artificial intelligence could make a difference in the years to come is the detection and therefore prompt treatment of diseases.
One of the aspects of the healthcare system where artificial intelligence is being used the most nowadays is definitely disease detection and diagnosis. AI is very accurate where the human eye can be sometimes fallible. Let’s have a look at three examples.
In the UK, researchers are working on an AI-powered software to detect breast cancer. This program can interpret mammograms and diagnose 30 times faster than a doctor, with 99 percent accuracy. Stephen T Wong, one of the researchers leading the project, said: “This software intelligently reviews millions of records in a short amount of time, enabling us to determine breast cancer risk more efficiently using a patient's mammogram. This has the potential to decrease unnecessary biopsies."
Another AI software, created in Illinois US, by the biomedical engineer Mozziyar Etemadi, is able to detect lung cancer earlier and with more accuracy than a trained radiologist could do. Lung cancer is the deadliest cancer in the world, for both men and women. Because the symptoms (persistent cough and fatigue) are frequently underestimated, the tumour is often found too late. About 75% of those who have it die within five years of diagnosis. Yet, when the cancer is detected early, the prognosis is much better. Etemadi’s system relies on a deep-learning approach: over time, it gets better and better at flagging early signs of cancer.
Finally, the artificial intelligence company DeepMind, together with Google Health, is currently working on several open source projects that will enable us to detect a number of diseases through deep-learning softwares. In 2018 they developed a clinical decision support (CDS) tool for identifying 50 different eye diseases. Recently, they have been producing great results in determining one of biology’s grandest challenges: a protein’s 3D shape from its amino-acid sequence. The human body uses tens of thousands of different proteins. Knowing these shapes could, to name one, help researchers (and AI) devise more efficient drugs. According to the evolutionary biologist Andrei Lupas: “It’s a game changer. This will change medicine. It will change research. It will change bioengineering. It will change everything.” The protein structure database they created is called AlphaFord.
After helping to detect a disease, artificial intelligence can help treat it. Through computational models and machine learning, AI can assist doctors in deciding the best course of treatment. For example, for patients with cancer, researchers can use predictive analytics to determine how an individual will respond to a certain treatment. Not every patient has the same response to medications. Predictive analysis can avoid patients going through a cancer treatment (with its side effects) that may not work for them.
Finally, robotic-assisted surgery has been strongly developing in the last decades. While the presence of the surgeon for decision making is still crucial, these innovative techniques allow, in some cases, a more precise and minimally invasive surgery. Today, it is used in gastrointestinal, gynecology, bone, spine and transplant surgery, to name a few.
Artificial intelligence applied to medicine is changing not only the work of medical professionals but also our relationship with health in general. We expect a future in which AI will work alongside doctors, enabling them to do their jobs better and save more lives. A future in which we will be healthier and live longer.
With cryonics aka biostasis, you have a chance to see this future with your own eyes. The possibilities of medical technology are increasing day by day. Looking at all these advancements, we are optimistic that revival will become a standard practice in the future. Will you be there to see it happening?