Will there come a time when computers or robots will replace human doctors? I know I’ve thought about it. Have you?
An article in the Sunday, June 4, 2017 Augusta Chronicle (“Researchers use IBM’s Watson to assess tumors’ genetic markers”) addressed the use of IBM’s super computer to help in the diagnosis and treatment of cancer. From that article, we learned that a physician at the Georgia Cancer Center was feeding raw genetic data sequenced from a very rare cancer into IBM Watson for Genomics, a computer system that is endowed with artificial intelligence that has helped it conquer other frontiers such as competing on the game show Jeopardy. Less than one minute after he had finished entering his data, he had a report back from the computer database that not only addressed the particular genetic mutations that he was concerned about, but also told him about ten medication trials that were already underway that might have implications in the treatment of his own patient.
This is astounding. I can remember the days when researching anything meant going to the medical library, looking in card catalogues for physical cards, cross checking references, pulling large, heavy textbooks off of shelves and taking notes or making copies of relevant passages, then taking all of that back to the dorm or apartment or on call room to read and digest and make sense of it all. Even then, there was very little certainty that one had found everything that was known about the topic at hand. This type of research could take hours or even days.
One thing that a computer with artificial intelligence is good at is taking huge data sets and analyzing them, making sense of them and then offering things that are truly helpful to the human who queried the data base in the first place. Another thing that these systems offers is time savings. As I mentioned above, the sheer amount of time that it would take one human, or even a team of humans, to sift through so much data is prohibitive in a normal clinical setting.
You might remember that another activity that Watson and similar computers are good at is playing chess. According to Wikipedia, chess computers were able to beat strong chess players starting in the late 1980s. One of the most famous chess matches between man and machine was when Deep Blue played then world champion Garry Kasparov in 1997, defeating him. Interestingly, what has been found as the years have passed since that match is that man can defeat machine and vice-versa, but the strongest combination of all is what is called a Centaur player, meaning the combination of both man and machine playing together as one. Rather than the mythological half horse-half human, a Centaur chess player combines all the intuition, creativity and feeling of a human player with the brute computational strength of a supercomputer. Centaur chess is not about computers taking the place of humans, but in fact augmenting and strengthening them to the point that they are many times better than before.
Does this play out elsewhere in medicine? You might also be familiar with the da Vinci® machines that are being used in the area to do surgeries of various types. These robotic machines are used not to take advantage of augmented intelligence, but to fine tune minute surgical techniques that a human hand might not be able to accomplish. Through very small incisions, the surgeon, who is in control of the machine at all times, is able to do very complex procedures in very small, tight spaces. The machine is being used not as a replacement for the human surgeon, but as an extension of him.
Are there already automated procedures or robotic elements in use in mental health today? Yes, there are already automated history taking computers and programs, automated treatment planning, and even algorithms that point toward a particular drug or other treatment for a certain disease. Genetic testing is on the rise, both to help make an accurate diagnosis and to target a specific disease with a specific treatment. Computers and computerized diagnosis and treatment are everywhere in medicine, and that includes mental health.
Once again, the role that computers might play going forward is to access and analyze vast databases of information on millions of medical topics, more than a single physician could access and understand in a lifetime. Computers can search this data with lightning speed, and they can test all known hypotheses and possibilities for intervention in mere seconds.
Is there still going to be a need for human empathy, human contact, human touch? Is there still going to be room for the human interactions that we now take for granted to impart hope to patients suffering from illnesses physical and emotional? I would argue yes, as human physicians have a unique perspective, bond and ability to heal that will most likely never be replaced by a machine alone.
Future physicians may be cyborgs, at least in principle, but the computing part of a practicing medical doctor will, like Watson, always be a tool. According to Dr. Kohle in the article above, “the physician will remain responsible for the conduct of patient care and for evaluating the clinical relevance of the information provided by the tool”.
I believe that computers will continue to enhance our ability to provide good care to patients, but that they will not be replacing us any time soon.