Thanks to advances in computer engineering and the proliferation of advanced — and cheaper — machines, everyone from software giants and tech companies to hospitals and university researchers are now looking to capitalize upon the growing power of artificial intelligence as it relates to the world of medicine.

Thus far AI has evolved to such a degree that it can analyze a massive amount of information, producing tailored results and, perhaps more impressively, identifying patterns not otherwise obvious to humans.

In the medical realm, this means that AI can be used to conduct real-time analysis of the information present in the electronic medical records, transforming these systems from digital filing cabinets into vast information repositories from which conclusions, trends and even recommendations can be drawn.

“Electronic health records [are] like large quarries where there’s lots of gold, and we’re just beginning to mine them,” said a managing director with Microsoft who is working on adapting AI in health care settings.

To illustrate how AI is currently being used in the medical realm, consider “homemade” systems currently in place at Vanderbilt University Medical Center and St. Jude’s Medical Center. At both of these renowned institutions, patients’ electronic medical records will alert physicians via a highlighted pop-up notification when a particular drug may not work or even whether the patient is likely to need a certain drug in the future.

This drug-related advice dispensed by the AI-based system, which uses algorithms developed from the electronic medical records of 16,000 patients, is taken into account by physicians roughly 66 percent of the time.

“The algorithm is pretty good,” said one physician of its drug predication capabilities. “It was smarter than my intuition.”

As impressive as all of this is, experts indicate that AI still needs to undergo vast improvements before it can truly be considered anything close to failsafe. Specifically, they say rather than just analyzing data buckets, it needs to be able to interpret things like images, doctor’s notes and reports.

While these developments are fascinating and encouraging, they also raise some important questions. Will all this ultimately lead to physicians becoming too reliant on technology? Is the AI being used today truly safe for use in patient care or can it result in medical mistakes? What steps are in place to protect patient privacy when developing AI technology?

What are your thoughts on this topic?

Source: Wired, “Artificial intelligence is now telling doctors how to treat you,” Daniela Hernandez, June 2, 2014