Steve Tolle, VP and Chief Strategy Officer, Watson Health Imaging, addresses the audience at Merge Live 2016.
Given the explosion of medical images in recent years, what’s expected of radiologists – from a data processing standpoint alone – has become untenable.
“They’re being asked to do the impossible by getting it right every time,” said Steve Tolle, VP and Chief Strategy Officer of Watson Health Imaging (WHI).
To provide some perspective, in 2015 alone, there were roughly 800 million multi-slice exams performed, generating approximately 60 billion medical images in the U.S. Further complicating matters, these images are disconnected from other relevant data (EMRs, lab results) that would provide critical context and patient history. Given this complex reality, one can hardly expect the nation’s 31,000 radiologists to examine and extract all of the potentially life-saving information hidden in a handful of images within this rising sea of data – but work underway with Watson Health is poised to help ease the burden.
During one of the most highly anticipated sessions of Merge Live 2016, held in Chicago last month, Tolle addressed hundreds of medical imaging experts, IT specialists, and radiologists, laying out the vision of what Watson Health is striving to do in order to help overcome the challenges of too much siloed data and too little time for healthcare providers.
“We’re bringing the power of cognitive computing to healthcare to expand the physician’s view so they have greater confidence in their diagnostic and treatment decisions for their patients,” said Tolle. The goal is to accomplish this in three important and inter-related ways.
- WHI seeks to help physicians see expansively by harnessing massive cognitive power to search and quickly interpret billions of data points – both text and image – within the patient’s EMR, across other patient similar cases, and across the most up-to-date medical research.
- WHI strives to help physicians recognize reliably by unlocking previously hiding information in disconnected data by presenting a wide variety of information within the context of patient history.
- And WHI intends to help physicians communicate objectively by searching and analyzing vast amounts of structured and unstructured patient, population and research data, presenting unambiguous evidence, and assigning probability scores to aid consistency across physicians’ diagnoses.
Of course, in order to make this vision a reality, there’s a great deal of work to be done. IBM is leveraging the data and expertise of the newly-formed medical imaging collaborative
to “train” Watson with thousands of studies per modality and organ. The collaborative is a growing network of academic and private hospital systems, across ambulatory and acute healthcare settings, whose diversity helps to avoid the bias and limitations of a one-system strategy.
“Our approach is unprecedented, and we are setting the bar very high,” said Tolle, adding that there are plans to seek FDA clearance for WHI products.
The first wave of commercialized offerings, designed to reduce practice pattern variation and assist in EMR patient data summarization, will be commercially available soon to address some of the toughest problems in cardiology and oncology, with other medical specialty applications to follow.
“Given the enormity of what we’re creating, and the degree of difficulty, we are pacing our progress carefully and taking a very disciplined approach,” said Tolle. “But make no mistake, what we are doing is real, it’s rigorous and it’s revolutionary.”
And based on the enthusiastic reaction from the Merge Live audience, the medical imaging community is ready to put Watson to work.