Improving the success rate and access to medical diagnosis has always been one of the most important priorities in the history of medicine. And today it’s clear that artificial intelligence (AI) is a powerful ally on this front. The field of Al has been experiencing an extremely fast development over the last few years. IBM has been one of the main forces behind this trend for over two decades: from beating the world’s strongest chess player 20 years ago with Deep Blue, the tech giant now possesses a supercomputer capable of performing successful medical diagnosis on its own.
The Watson supercomputer has already given very impressive displays of its intellectual talents, but IBM wants to lead it into a much more important aspect in the real world by putting it at the service of medicine. Over the last few years, Watson has been endowed with the ability to learn the basics of medical diagnosis by scanning exam books and analyze confusing data in health data centers.
For this purpose, IBM has collaborated with medical institutions like Cleveland Clinic and the Institute for Applied Cancer Science at MD Anderson. The initial goal was to make Watson improve the success rate and access to medical diagnosis in hospitals. This was done by helping to summarize and organize the recorded medical data, giving doctors a summary of the patient’s history. In theory, this should help doctors treat more patients more effectively by tracking information directly to their source.
The first glimpses of automated diagnosis
IBM’s project hasn’t stopped there. By creating a conceptual map and analyzing medical examinations, facts and test results, Watson can also analyze and create theories that could explain patient’s symptoms. Soon the supercomputer would prove its ability to perform medical diagnoses on its own. A prime example is the successful diagnosis of a rare form of leucemia that Japanese doctors had failed to detect on a female patient.
Another very similar tool is Isabel Healthcare’s Diagnostic Engine, which has been in use for several years at several health institutions with very impressive results.
Patients’ histories in most advanced societies are increasingly digitized. Thus, the ability of supercomputers to track so much information is gaining importance. Consequently, one of the most important benefits of medical Al is the ability to offset the increasing shortage of doctors in many countries. With the number of patients per doctor increasing, it is becoming humanly difficult to manage so much information.
In the future, the power of medical data analysis could be sufficiently leveraged and refined to be distributed as software that people could install and use on their personal mobile devices. At which point, there would be no obstacles to truly fair and easy access to automated diagnosis and other medical services. This would revolutionize the efficiency and operational paradigms of health institutions.
Another potential consequence is a shift in the pertinent skill sets of hospital professionals. Data mining and analysis could become a very important task in health institutions owing to the increasing need to collect and pre-arrange all the data that must be fed to Al machines.