|
I'm actually getting ready to speak on this at a conference coming up soon and am pretty excited about the future of healthcare, specifically in the United States because that's where my research has mostly been focused (and where I live). One of the first things I'm interested in the future of is more in the near term future, and that's AI's ability to connect patients with doctors more appropriately suited to diagnose and care for patients, especially ones with unique/rare illnesses. One other near term benefit from using AI would be that the computer can take over medical billing for the doctor, allowing them more time to meet & treat patients versus filling out paperwork. In the long term, I could see AI being leveraged with medical records to diagnose an illness based on symptoms & medical history. Already, some AI models are outperforming doctors in terms of accuracy of diagnosis, but it's not across the board just yet. This will be a double-edged sword though because, although this may expand access to care for many people (ESPECIALLY if it means lower costs being passed onto patients), it will also likely make meeting with a human doctor much more difficult. Will a computer be better at diagnosing across the board and humans only necessary for edge cases, or will humans still be needed to validate AI decisions into the foreseeable future? Empathy is not programmed into these algorithms, so a doctor may make a choice that is better for the patient overall, but incongruous with what the computer would suggest for a care plan. Another exciting space is the use of visual AI models for early detection of cancer and other illnesses where even a trained human might not notice early indications of illness. Earlier treatment will lead to better patient outcomes, so the sooner a tumor is noticed, the sooner it can be mitigated. The entire field is rapidly developing and things seem to be changing about as quickly as "AI" in general, so I'm excited to see where things go! |