There is a reason why doctors encourage patients to go for regular health checkups. This is because if problems are detected early on, they can either be cured or their progression can be slowed down. But even then, doctors are only humans and can sometimes make mistakes or miss certain things.
This is where AI comes in where it might prove to be an invaluable tool in helping doctors detect certain health problems. Google seems to think so, and Danial Tse, a researcher at Google, actually developed an algorithm that actually managed to beat out several trained radiologists during its testing when it came to detecting lung cancer in patients.
As we have seen in previous AI models that are used in cancer detection, what happens is that the AI is fed thousands of scans of patients, where it eventually is capable of learning what a scan of a healthy patient should look like, and what scans of patients who have, or who are developing, lung cancer should look like.
Based on that, it will be able to tell doctors if a patient might be at risk. However, detection is just one part as there will need to be other tests done to confirm it, plus a treatment plan will also vary from patient to patient, meaning that the role of doctors shouldn’t be discounted just yet.
Filed in . Read more about AI (Artificial Intelligence), Google and Health. Source: technologyreview