The Limits of the Human Mind, Artificial Intelligence and the Future of Medicine

Acknowledging the limits of our perceptions and striving to learn moresummarize all science and technology. In medicine, this effort found expression through the stethoscope, microscope and electrocardiography. Now we face something that means more than merely a technology: artificial intelligence!

Medical knowledge is constantly updated and increasing rapidly. We cannot continuously follow every development completely, so we use, or should use, computers for help. Although patient decisions require deeper thought, we lack sufficient time. Electrocardiography (ECG), vital for diagnosing heart attacks and rhythm disorders, is currently best interpreted by highly specialized professionals such as cardiologists and emergency physicians. The day is near when computers provide accurate, detailed reports understandable to everyone, allowing doctors to focus better on treatment. Today’s diagnostic tools also provide snapshots: chest pain or palpitations may resolve before arrival at emergency care, leaving no abnormality on the ECG. Instead, if your smartwatch or wristband continuously recorded and analyzed heart rhythm using algorithms, unnoticed heart problems might be predicted earlier. Tens of thousands die each year from sudden cardiac conditions…

These algorithms will clearly shape medicine. The question is whether healthcare professionals will lead the change or watch from outside. Clinical reasoning cannot be excluded. Physicians will apparently teach machines, even at the risk of losing parts of their professional role, and will act as mentors . Examples already exist. IBM’s artificial intelligence Watsonis being developed with clinical oncologists, who teach it diagnostic and treatment algorithms.

Computers can learn how physicians make correct diagnostic and treatment decisions, as well as their mistakes. This could prevent some medical errors, repeat tests and extra treatment costs. They can rapidly analyze someone’s entire health history, including diseases, medications and genetics, and lifestyle, including diet, exercise, smoking and environment, to predict future health problems or treatment responses.

With machine learning, medicine will become a new team endeavor. This team will need good players, particularly physicians knowledgeable in statistics and computer science and skilled in developing and adjusting algorithms.

Medical education must adapt to these needs because graduation requirements are outdated. Future doctors should also learn data science, statistics and applied clinical algorithms.

Integrating data science with medicine is not as difficult or distant as imagined. The same was once thought of cell biology and genetics , but these now form the center of medical research, and all doctors graduate with training in them. Similarly, if work to unite medicine and informatics begins urgently and continues, 21st-century physicians will have tools to process the data needed for decisions and skillfully solve medical problems.

What Is an Algorithm?

An algorithm is a pathway created to solve a problem. Think of “Problem + Algorithm = Solution.” It is not specific to medicine, but algorithms are used particularly intensively in medicine and engineering. Healthcare artificial intelligence is designed to know thousands, even unlimited numbers, of algorithms. To clarify the concept, an example follows:

Management Algorithm for People Presenting with an Attached Tick