IBM Watson Enhances Oncologist Efficiency with AI-Driven Treatment Recommendations (Full Transcript)

Speaker 1: Cancer is the second leading cause of death in the U.S., and in the next decade, demand for treatments is expected to grow by 42%, creating a shortage of 1,500 oncologists. Meanwhile, medical information is doubling every three years, making it more difficult to keep up with the latest treatment options. Memorial Sloan Kettering has been training IBM Watson to help close the gap, as this live system example illustrates. Let's say I'm a clinical oncologist preparing to see a patient. I step into my office, and instead of spending my limited time digging for relevant information, I select my patient from the queue and use Watson to summarize the relevant facts from the electronic health record and identify areas where more information may be helpful. Next, I can match the patient to a clinical trial or review the suggested treatment plans Watson has presented. Green means higher confidence, while red means lower. To make recommendations, Watson has extracted knowledge from guidelines, best practices, 12 million pages of journal articles, and 200 medical textbooks. Watson provides a visual timeline of the treatment plan, as well as the plan's suggested modalities. Again, they're color-coded by confidence level. Each of the blue icons represents an opportunity to dig deeper into evidence. For example, I can review the specific sources behind each component of Watson's recommendation. Watson has also suggested additional publications for my consideration, an overview of the protocols for administering the treatment, and a summary of information of the suggested drugs. When I'm ready, I can select the treatment information to add to the patient's medical record and share with the patient. This demonstration has shown Watson's ability to create an at-a-glance summary of relevant facts from an electronic health record, to suggest confidence-scored, individualized treatment recommendations specific to each patient, and to transparently expose source data that supports the recommendations. Watson is helping oncologists fight cancer.