Needless to say, medicine is progressing very successfully these days. Gone are the days of questionable anesthesia and experimental surgery. We are learning more and more about what diseases do to the body and how to treat them. But can we get better at this? Sure. Some diseases still confuse doctors. Patients are suffering. In some places it has improved, in some places it has not.
“If your doctor visits you, he probably uses a stethoscope. He measures blood pressure with a cuff. Things that were 100 years ago are still the main diagnostic tool when they first meet,”said Vinod Kosla at a conference on exponential medicine hosted by Singularity University.
An entrepreneur, inventor and technologist, Kosla founded Sun Microsystems. Today, his company, Khosla Ventures, invests and helps technology companies grow.
When Kosla looks into the future of healthcare for 10 or 15 years, he sees a medical landscape seething with data-hungry algorithms like Google's AlphaGo, not the doctors we're used to.
“Medicine has improved in practical terms,” says Kosla. "But now I think it's time to turn this practice into a medical science."
To do this, we will have to transfer medical expertise to the machines.
Humans are not designed to handle the vast and growing amounts of health data that the latest technology generates. From health sensors to complete genome sequencing, our world is awash with data, and there will soon be more data points than there are stars in the universe.
What kind of brain can process all this information? We are quickly losing our grip on this position.
Promotional video:
More people in the United States are dying from misdiagnosis, surgical error, and fatal drug abuse in the United States than in road traffic accidents, Cosl said. In both cases, human error often solves.
In the past, a lack of information was a problem. We relied on training, experience and intuition to fill this gap. Gradually, however, excessive information also becomes a problem.
Kosla says diseases can be diagnosed with just one biomarker - the chemical signature of the disease - or by looking at 300 biomarkers. You can look at the patient in front of you and compare it to several other patients you have seen, or you can scan a database of 100 million patients looking for the last hundred or thousand with similar symptoms.
Harvard, Stanford - it doesn't matter. No doctor can handle this.
“For the most part, why we go to doctors - for a diagnosis, Dr. House’s brilliant insight, a prescription - will increasingly fall on the shoulders of machines, leaving the person with a completely different role in relation to patient and doctor,” Kosla says.
But we all would like that. Freeing doctors from the impossible task of collating stacks of medical information means they can focus more on patient care and technology advancement.
At the moment we have specialists in everything. None of them communicate with the other the way we would like. But the patient has an average of seven major symptoms. Wouldn't it be better if these symptoms were consistently studied by artificial intelligence, and one doctor - not seven - transmitted these results to the patient?
This new model will turn healthcare upside down. Instead of wandering from specialist to specialist, patients can get an initial consultation from a doctor or even a nurse. This single point of contact will allow the patient to stay in their comfort zone and receive all the help they need.
“We choose not the smartest, but the most caring. If you are going to treat people, take care of them, why not choose the most humane approach?"
ILYA KHEL