Symptoms of some diseases may appear almost imperceptible. Conversely, artificial intelligence can help achieve an early diagnosis
It was a sunny day out and the air had a spring touch. I followed Angela, to whom we changed the name to protect her identity, along the corridor to my office in Melbourne, Australia.
He has been my patient for several years, but it was the morning when I realized that he was dragging his legs as he walked. The expression on his face seemed a little unstable, and I noticed that he had a slight tremor.
I sent him to a neurologist in a week treatment of Parkinson's, but I felt really bad I did not find their symptoms before.
Unfortunately, this situation is repeated with patients all over the world.
They are only diagnosed when they begin to show remarkable signs of disease when the body radiates warning signs that something is wrong.
If they can be detected earlier, patients will be able to initiate treatment in the early stages, even with the possibility of stopping the disease before it develops.
New technologies are beginning to offer some hope in this respect.
With artificial intelligence, patients and doctors can learn about possible changes in their health months before, or even years that the symptoms appear.
Future Exploration Network futuristic teacher Ross Dawson predicts that in the next few years we will see a shift from the current "attention to disease" model towards a new ecosystem of medical care focusing more on preventing and monitoring possible health problems before are developing.
The change in the attitude of the society, which now has higher expectations of doing complete and healthy lifemoves these changes, "he says.
"During this decade, the explosion of new technologies and algorithms allowed artificial intelligence to have deep knowledge that has become more effective in recognizing patterns than humans."
By using artificial intelligence to monitor heart rhythm, breathing, movement, and even chemicals in our breath, technology can detect potential health problems individually long before the obvious symptoms occur.
This can help doctors or allow patients to change their lifestyle alleviating or preventing diseases,
Perhaps the most interesting thing is that these systems can differentiate models invisible to the human eye that reveal surprising aspects of our future health.
Look at your health
Dawson mentions, for example, studies in which artificial intelligence is able to predict possible heart attacks of people most likely to suffer from them by continuously monitoring the impulse.
One study even finds variables that cardiologists have not given a predicted value.
Another recent Google researchers study has shown that Artificial Intelligence algorithms can also be used to predict whether someone will suffer heart attack p Look at it in your eyes
The team showed a program of artificial intelligence retinal scanners on 284,335 patients. The machine has learned to detect the signs of cardiovascular disease by looking at the blood vessels in the eye.
If Dina Katabi becomes a way, delays in diagnosing genetic diseases and diseases such as Parkinson's, depression, emphysema, heart problems or dementia will remain in the past.
She is responsible for designing the device she is transmitting low power wireless signals in house. These electromagnetic waves take data from the patient's body.
Every time we move, we change the electromagnetic field that surrounds us.
The Katabi device detects these reflections and traces them using it automatic training follow the patient's movements through the walls.
Katabi describes wireless signals as "incredible beasts" that go beyond our natural senses.
Deploying a device in the patient's home allows you to constantly monitor their sleep, mobility and gait.
You can record your rhythm of breathing, even with a few people in a room, and keep a watch on your heart or provide information about your emotional state.
"We do not see them, but they can complement our knowledge almost magically," she says.
"Our new device is able to pass through its walls extract vital information this can increase our limited capacity to perceive change. "
This ability to find changes in patients' daily behavior can provide early clues that something is wrong, perhaps even before they know it.
Many of us already use different devices for self-control of our constants, from caloric intake to the number of steps we take each day.
Artificial Intelligence can play a vital role to help make sense of all this information.
This ability to predict changes in health can be particularly important as our population ages.
According to the UN, people over 60 will account for one fifth of the world's population by 2050.
The value of the person
Artificial Intelligence can also use the way we look to help us predict future illnesses.
A recent study shows that subtle differences in our face may be the characteristics of a disease.
FDNA, a startup Based in the US city of Boston, he has developed an application called Face2Gene that uses something he calls "deep phenotyping"Identify possible genetic diseases from the patient's facial features.
He uses a technique of artificial intelligence, known as deep learning, which teaches his algorithms to discover the features and forms of the face that are usually found in rare genetic diseases, such as youNoonan syndrome,
The algorithm has been trained with more than 17,000 photos of patients affected by one of the 216 different genetic conditions,
Some of these diseases develop patients some features of the face from his illness, as well as in intellectual disabilities like Bain, where children have characteristic almond eyes and a small chin.
The FDNA algorithm has learned to recognize those facial features that are often undetectable to physicians.
Inside your brain
Given that rare diseases affect approximately 10% of the world's population, it is likely that artificial intelligence tools change the drug.
However, not all diseases are obvious to identify from outside.
Doctors and surgeons have long relied on X-rays and scanning to diagnose the cause of their patients' symptoms.
But what would happen if it were possible to use these disease detection tests before it begins to reveal symptoms?
Ben Frank is not an ordinary radiologist.
Professor of Clinical Radiology at Stanford University, USA, is underway to find a secrets hidden in millions of tomographies Whole-body PETs are performed regularly every year in each oncology department.
Frank works with a team to investigate whether changes in brain metabolism that occur in these PET scans can be used.
You want to predict if anyone can develop Alzheimer's diseasea condition that affects 10% of people over 65 years of age.
Using artificial intelligence, they have developed algorithms that are able to detect these subtle changes in metabolism, in this case, the consumption of glucose in certain brain areas thought to occur at an early stage in the development of the disease Alzheimer's.
But it's not just Alzheimer's disease.
Machine training offers new ways to detect early mental illness by detecting hidden signals when choosing words, voice tone, and other nuances of the human language.
Ely, a digital avatar developed by the Institute of Creative Technology at the University of Southern California, analyzes more than 60 points on the face the patient to determine if he or she may be depressed, anxious, or suffering from post-traumatic stress disorder.
The time you take before answer a question, your position, or how much you agree With his head he gives Elli clues about the patient's mental state during the "consultation."
We hope that this way of using machine learning will generate significant progress in. T psychiatric results "Improving the prediction, diagnosis and treatment of mental illness," wrote Nicole Marinez-Martin with colleagues from the Stanford School of Biomedical Ethics in a recent article.
When all this is combined biometric measurements Individuals with the genetic profile, the result may allow predicting individual risk factors that would replace today's expanded medical guidelines.
In the world of precision medicine, artificial intelligence can make annual medical analyzes anachronistic.
But how much trust are we willing to put into an algorithm when it comes to our lives?
A recent article in Ethics Journal of American Medical Association creates a scenario where machine learning is used for decision – making in the predict the end of life per patient.
The authors note that "an algorithm will not lose sleep if it predicts with a high degree of confidence that one would want a machine for sustaining life it goes out.
The question is: do we want something that does not worry about the decisions it takes at such important moments?
When we become ill, we may prefer to have a doctor from our country instead of a machine, but in the near future the AI may be able to detecting problems much earlier by his organic partner.
By being fully adapted to our individualities, behaviors and emotions, they can give us an early warning that can save our lives.
So even if we do not expect a computer to sit next to us in bed, It is wise to ask them to understand what and how we feel.
You can read the original version of this English story of BBC Future.
You can now receive notifications from BBC World. Download the new version of our application and activate it so you will not miss our best content.