It’s simply common sense to use AI in healthcare
A good article in The Guardian, “GPs use AI to boost cancer detection rates in England by 8%”, explains how family doctors in the UK have managed, using a relatively simple AI tool that accesses a patient’s medical history, test results, prescriptions and treatments, along with their zip code, age and family history, enabling them to improve the early detection of cancer cases by 8%, saving a good number of lives as a result.
It makes perfect sense to use algorithms in medicine, but doing so is often made difficult by misinterpreting privacy legislation or by administrative procedures and silos that prevent the proper sharing of data.
The tool being used in the United Kingdom is called “C the signs”, and is based on machine learning processes capable of identifying statistical anomalies in the evolution of a patient’s parameters, and comparing them with cases in which this anomaly corresponds to a given pattern. There is nothing generative or particularly novel in this type of development: it is simply the use of statistics, in ways in which it would simply be impossible for a human observer to draw conclusions as this would involve analyzing a huge number of similar variables.
Medical information must always subject to the highest level of protection. That said, patients are always willing for this…
