IBM officials struck a bold tone about artificial intelligence, cognitive computing and machine learning at Health Datapalooza.
“In 10 years cognitive will be as mainstream as transactional systems are today,” said Jane Snowdon, director IBM Watson health partnerships. “Today we are very focused on healthcare, we are training on different lexicons and domains so Watson can be very specialized for those fields.”
Snowdon explained that with 8,000 new medical journal articles published every day it’s impossible for, say, an oncologist to keep pace with state of the art data on clinical trials, treatment regimens and other medical literature but Watson can translate all that data into information relevant to a specific patient case.
She added that IBM worked with doctors to test Watson on 1,000 patients and the supercomputer not only agreed with humans 99% of the time but it also offered suggestions about care pathyways that clinicians had not thought of approximately 30 percent of the time.
The ultimate goal is to be able to provide Memorial Sloan Kettering levels of cancer care in far-flung locales such as rural China, she said.
“Cognitive and artificial intelligence have the ability to level the playing field, to democratize medicine,” said Jose Morey, MD, senior medical scientist, IBM Watson research. “That’s the real potential in cognitive and augmented intelligence.”
Amar Das, MD, director of IBM’s healthcare effectiveness research said that getting insights we have right now is a major challenge because the health IT system is as fragmented as the healthcare system and that means doctors have to search through a maze of information and interfaces to find useful data.
“Instead of having more data, we need to be able to glean insights from it,” Das said. “We need to augment the intelligence of physician so they can find exactly what they need.”
When asked by an attendee whether AI, cognitive computing and machine learning might one day be so efficient that human clinicians and physicians become the bottleneck inhibiting care delivery, Morey said that we will never get to a point where a computer can hold a loved one’s hand and explain why the patient is dying or help them understand what’s happening.
“But it could make more doctors efficient,” Morey said. “It’s an augmentation to the human to enable what the human does best and what the computer does best.”