While the potential uses of artificial intelligence are pondered quite often, few people put much thought into how AI could benefit medical fields. A good example is brought forth in a recent study, showing the AI driven computers either performed as well as doctors or better than them in analyzing CT scans for signs of lung cancer. Of course, this technology isn’t a final product yet, and it still needs more time in the developing oven before it is ready for widespread use, but the results are nevertheless promising.
The greatest part about the potential of AI in medicine is its ability to swiftly recognize patterns that it has encountered before. While a human doctor could of course recognize patterns as well, he wouldn’t be able to do it nearly as quickly or as efficiently as an AI computer could manage. And the best part is, with the learning capability of AI, the machine would become more and more adept at identifying patterns the more often it did it. In other words, it would gain experience just like a real doctor over time.
This will be extremely helpful when it comes to analyzing CT scans, because a computer AI will have a much easier time of identifying small irregularities like hairline fractures, cancer, or pneumonia, things that even experienced medical professionals have trouble noticing sometimes. And of course, with vastly superior processing power compared to humans, a computer AI could go through a lot more data in a lot less time, greatly streamlining a time consuming process.
Better yet, AI can even be taught like a normal person would be. For instance, the AI program that is analyzing these CT scans can be given many scans that have already been accurately diagnosed, like a worksheet in a class, to see how well it is performing. This type of teaching was done with the AI looking for lung cancer, with over 6,500 CT scans being used. Ultimately, the AI was 94% accurate with its analysis of these scans. It was even able to determine locations in a person’s body that were simply ‘likely’ to develop cancer, which would allow doctors to divide patients into risk groups depending on such an analysis.
Ultimately, it might be a while before AI is being used regularly in the medical field, as most people will probably still prefer the tried and true experience of human doctors. But while there may be some minor concerns regarding how much work AI could take off medical shoulders (and by extension how many jobs it could cost to employ) there’s no doubt that use of artificial intelligence in the medical field will help save many lives, as well as a lot of time and effort. Whether or not we’ll be able to see those AI benefits in the next ten years remains to be seen, but with how fast technology advances, it seems likely that computers will be taking more care of us than usual very soon.