Eric Topol is the founder and director of the Scripps Research Translational Institute in La Jolla, Calif. His books include The Patient Will See You Now: The Future of Medicine is in Your Hands and Deep Medicine: How Artificial Intelligence Can Make Healthcare Human Again.
For decades, there has been a steady erosion of the practice of medicine, with progressively less time between patients and doctors, a global epidemic of physician burnout that has now reached a crisis, a doubling of medical errors when doctors have symptoms of depression and most serious errors attributable to bad clinical judgment.
Concurrently, each patient’s cumulative data, such as prior history, laboratory tests, scans and sensor output, keeps growing, as has the doctor’s relegation to the role of data clerk. The limited time to think has led one leading physician to conclude: “Modern medical practice is a Petri dish for medical error, patient harm and physician burnout."
It would seem counterintuitive, at the least, to think that technology could come to the rescue for this dire situation. But there are many early signs that a subtype of artificial intelligence called deep learning may be just what the doctor ordered.
The pioneering work on deep learning was performed at the University of Toronto by Geoffrey Hinton, Yann LeCun and Yoshua Bengio, and was recognized in 2019 by the Turing Award – the equivalent of the Nobel Prize for computer science.
Deep-learning neural networks have been trained with large, labelled data sets for high-level performance, often matching or exceeding humans for a variety of tasks such as games, images, voice recognition and self-driving cars. In particular, deep-learning medical algorithms have been gaining considerable attention because of their breadth and potential impact.
We have seen the accuracy of almost all types of medical scans increase with deep-learning algorithms, from X-rays to mammograms, CT and MRI images. Diagnoses of the eye (such as diabetic retinopathy and glaucoma) and skin (especially melanoma and other cancers) have had their accuracy substantially improved by these neural networks. The same extends to gastroenterologists for picking up polyps in real-time during a colonoscopy or the correct determination of cancer type and status from pathology slides.
The advances are also occurring across the lifespan, ranging from improved embryo selection for in-vitro fertilization to use of machine vision in the hospital setting to facilitate patient safety. These tools are symbiotic add-ons, not replacements for clinicians. As Italian academic neurosurgeon Antonio Di Ieva nicely summed it up, “Machines will not replace physicians, but physicians using AI will soon replace those not using it.”
Perhaps the principal short-term impact for deep learning and AI tools would be to liberate doctors and patients from keyboards, their common enemy that markedly detracts from real human interaction. In a review I led of the future work force for Britain’s National Health Service in 2019, our team’s health-care economists calculated that eliminating just one minute of keyboard entry time for doctors represented the equivalent of 400,000 hours of consultation time a year, or 230 full-time physicians.
The improvement in diagnostic accuracy carries over to patients, too, with smartwatch algorithms that can detect and classify abnormal heart rhythm, a drugstore kit to diagnose a urinary tract infection or a smartphone app to detect whether a child has an ear infection. This list of common, non-serious conditions that can be diagnosed by patients with algorithms will grow, cultivating more autonomy for willing patients and, as a result, less requirement for doctor visits.
But there is a more far-reaching objective than improving diagnostic accuracy and speed: giving doctors and patients the gift of time – to get back to where medicine was decades ago, when the relationship was characterized by a deep bond with trust and empathy.
During a clinic visit, there was real listening, presence and a thorough physical exam. All of that can start to be restored if we give back time for the patient-doctor connection to be nurtured. The same physicians who suffer burnout and symptoms of depression today are the ones who pursued a medical career so they could provide care for patients, which is markedly compromised today.
Instead of the big business that health care has become, we have a unique chance to bring back its essence – the humanity of medicine.
To confront the overriding force for increasing productivity and revenue, the medical community will need to stand up and strongly advocate for patient interests, which fully aligns with their ability to provide care.
Surely, it is ironic that we may depend on artificial intelligence and machines to promote emotional intelligence, the time for humans to think and be more human. But over the next decade, I hope we’ll see exactly that, whereby health systems and practices actually compete on the basis of how much time they give to their patients.
Keep your Opinions sharp and informed. Get the Opinion newsletter. Sign up today.