Christopher Thompson. Executive Director of the AMIIF
We usually talk with great enthusiasm about health and artificial intelligence, but in many aspects, health systems have not even come out of the previous revolution, electronics. Just go for an example: the electronic clinical record still cannot be consolidated and in the rare institutions that have one, it is common to find that these medical records are not compatible beyond the software of each institution. And I give this example, not to depress you, but, as the American cardiologist Eric Topol says, to realize that we are in the infancy of the era of artificial intelligence and that allows us to choose the path we want to take forward.
In his book “Deep Medicine: How artificial intelligence can make medical care human again”, Topol talks about three fundamental pillars: deep phenotyping, which will allow us to better understand each individual, from their genetic history to the factors environmental factors that influence it; deep learning, to take advantage of artificial intelligence based on pattern recognition; and deep empathy, to reprioritize the relationship between doctors and patients. In Topol’s vision, the future of medicine can be one of true collaboration between new technology, doctors and patients. Artificial intelligence will protect us from our biases, logical fallacies, and limited vision, giving doctors time back to do what machines cannot imitate: listening to patients with empathy, inspiring trust, achieving complex, deep communication. and human.
Perhaps one of the most important warnings of this book is that it reminds us that the future of medicine cannot be written from scratch, but rather will have to be built on a health system full of complexities and inconveniences that must be overcome. For example, one of the data that stands out is the little time that doctors spend with each patient: only 12 minutes in the first consultation, and 7 in the subsequent ones. Topol compares this to data from 1975, when a doctor spent an average of one hour with a first-time patient and thirty minutes on subsequent visits. This results in the practice of “superficial medicine” due to lack of data, time, context and presence. Topol proposes to use artificial intelligence to move from surface to deep medicine. Technology known as deep learning can help clinicians more accurately recognize pathology, analyze patients globally, and sift through scientific literature for up-to-date information.
Artificial intelligence gives us the opportunity to create more robust and more humane health systems. We still have time to design them with this in mind.