It'd be hard to believe how quickly artificial intelligence found its footing in the mainstream — if we weren't already reaping the benefits in industry as well as our personal lives.
If anything, though, the progress we've already made in adapting AI to our existing industries, lifestyles, and workflows only betrays how much more work there is to do. In health care, particularly, we're only beginning to explore AI's potential in making our lives better and healthier.
So, what's the "killer app" for AI in health care? There are several — but for our purposes today, we're going to focus on three of the biggest ones.
1. AI Eases the Administrative Burden of Health Care Systems
For a start, it might be best to stop thinking about AI as a panacea or a one-size "solution" for the current health care landscape. As with applications in other industries, artificial intelligence seems most useful when it adds to the human capacity for reasoning, observations and rigorously attentive diagnoses and troubleshooting. In other words, it complements rather than replaces human intuition and effort.
There's much that needs to change about modern health insurance — like restructuring it into a far cheaper and more efficient single-payer system. But until that happens, health care systems, hospitals, and other facilities that administer care have to work within conventional limits, such as the number of people they can reasonably "process" in each day and the time required to bill all of the associated parties.
One of the most groundbreaking applications for AI in health care is the idea of making social safety nets more efficient and more cost-effective to operate. Hospitals have historically struggled with the labyrinthine billing practices required by a health care system that's one part publicly owned and three parts profit-motivated.
Whether as part of Medicare, Medicaid, or a health plan purchased on a public exchange, hospitals deal with all kinds of patients from all walks of life. Dealing with the administrative burden of managing payments and treatment regimens for hundreds or thousands of patients in a network is a daunting, error-prone challenge that can result in patients being over- or mis-billed or worse: being matched with the wrong medications or even the wrong charts and paperwork.
According to health care experts, some 70 percent of the real-world costs of health care are related to administration. Artificial intelligence has the potential to be a huge time- and money-saver just by adding clarity, accuracy, and oversight to the behind-the-curtains processes that make modern medicine possible.
2. AI Matches Patients with Relevant Treatments and Clinical Trials
Of course, some of AI's applications in health care go far beyond making sure payments and treatments are organized and squared away for every patient. Artificial intelligence might truly shine once it gets better at identifying patients who are a good match with specific care practices and certain types of medical intervention — including relevant clinical trials.
This is increasingly possible thanks to the federally mandated shift to electronic health records, or EHRs, which contain far more information than a single doctor could be expected to internalize. Applying AI to finding treatments and trials for sick patients means getting a top-down view all of the information already at our disposal.
Doctors are some of the most thoughtful and intuitive professionals you'll find anywhere — but AI can greatly cut through the noise and elevate the most important patient characteristics and factors to the top, where they can provide insight into the best next steps to take and which ones to avoid.
Using AI to learn more effectively from medical histories and accumulated patient data allows for more timely interventions, more accurate diagnoses, more patients connected with breakthrough clinical trials and, ultimately, better patient outcomes and fewer hospital readmissions.
3. AI Powers the Next Generation of Non-Invasive Medicine
The technologies that allow us to detect a malign presence in the body, such as cancer, have come a long way. But even MRIs, CT scans and X-rays oftentimes require a physical tissue sample from patients for a full diagnosis. This is a time-consuming process and even represents a possible source of infections. It also doesn't do much to persuade scalpel-averse patients to visit a health care facility when they believe something is wrong with them.
Artificial intelligence, in the near future, will combine with extremely powerful imaging equipment and "learn to look for" some of the signs of disease that would normally require a tissue biopsy and other invasive, unpleasant procedures.
Further Untapped Potential for AI in Medicine
This is far from a complete rundown of how modern medicine stands to benefit from advancements in artificial intelligence.
One more worth mentioning is expanding health care coverage in underserved or geographically remote areas. When time and distance are limiting factors for people seeking medical assistance, AI-powered imaging tools offer algorithmic screening for things like tuberculosis. This can help doctors better prioritize patients according to their condition.
We're already seeing impressive results from apps that use the ever-more-powerful sensors and cameras in modern smartphones to aid these "digital doctors" in their appraisals of patient bodies, even when they can't physically travel to a doctor's office.
This kind of a practice lines up closely with what’s now being called telemedicine. Telemedicine is quickly becoming popular, particularly in the mental health field. Telemedicine currently fulfills needs like connecting counselors with patients via text message. But it also has the potential to connect artificially intelligent responders with patients who perhaps just need someone to talk to.
Something like telemedicine may never necessarily be able to replace certain specialized medical practices, such as the important physical experience former addicts receive from detoxing at a rehab center. However, telemedicine is proving to be very effective in a variety of short-term care situations, particularly with patients suffering from, say, a minor bout with depression. The opportunities available in this area are probably beyond what we can currently imagine.
By now, there are more than enough good ideas floating around that it's inevitable that tomorrow's doctors will regularly rely on artificial intelligence and algorithmic tools to reduce their administrative burdens, reduce errors, improve patient outcomes and even help shape the future of medicine.