IBM’s Watson artificial intelligence, which the company is considering condensing into a mobile form, will soon be moving on from defeating human contestants in trivia game shows to diagnosing patients at Memorial Sloan-Kettering Cancer Center in New York.
The AI has spent at least six months being taught the treatment guidelines in use at the Center and soon the doctors there will be using Watson to diagnose patients or at least assist doctors in making their decisions. Watson has been fed large amounts of data on oncology, including medical textbooks, journals, patient histories and treatment guidelines. When the AI is presented with a patient’s medical records and other data, something that has been tested with a custom iPad app used to simulate an actual patient, it will return a suggested treatment or range of treatments for that patient along with a confidence probability for the treatment based on the information Watson is privy to.
Doctors will be able to approve the treatments suggested by Watson with the touch of a button and having the AI perform the diagnosis should speed up the process for oncology patients considerably. An eventual use for Watson’s abilities could be to provide an expert second opinion to doctors in other parts of the world, a relatively simple jump considering that Sloan-Kettering’s physicians are interacting with the AI remotely at the moment.
As the learning AI is taught more, its abilities in specific lines of expertise are expected to grow. IBM’s Manoj Saxena, who is in charge of commercialising Watson, has said “It’s like being able to take a knowledge worker–cancer specialist, nurse, bond trader, portfolio manager, whatever–and equip that person with the best knowledge, and have it available at their fingertips.”
Source: via Gizmodo