read-more
Read more about Fred Hutch achievements and accolades.
Kathy Alexion, vice president and chief information officer at Fred Hutchinson Cancer Research Center, gave the closing keynote speech at the Hopperx1 Seattle event on March 23. Coordinated by AnitaB.org and inspired by Dr. Grace Hopper, a pioneering computer scientist and U.S. Navy rear admiral whose work led to the development of the COBOL programming language, Hopperx1 events support women in tech. AnitaB.org’s Seattle chapter is the largest in the U.S., and its two-day conference this year drew more than 1,000 attendees from across the Pacific Northwest.
Alexion addressed how technologists can leverage cloud computing, which is transforming scientists’ ability to exchange data and emerging technologies, to tackle cancer.
Over the course of their lifetimes, Alexion noted, one in three Americans will be diagnosed with cancer. “But we have the ability to change that, we here in Seattle, with technology, with data science and bioscience converging in unprecedented ways — we at Fred Hutch, the leader in cancer research, we as women, we as technologists in this room and beyond,” she said.
With assists from the Hutch’s Emily Silgard and Karma Kreizenbeck, Alexion highlighted three examples of how scientists are collaborating, including with several tech leaders in Seattle, to answer the question, “Can the cloud cure cancer?”
Natural language processing
Silgard, a data scientist with a background in linguistics, spoke about natural language processing, in which computer programs are designed to process and analyze data stored as human language, as in doctors’ notes. Those notes contain data on patient diagnoses, demographics, treatment, treatment responses and outcomes — all critical to aiding researchers.
But, Silgard said, it takes a well-trained human clinical abstractor an hour to annotate these pieces of data in just one to two medical notes. Collaborating with Amazon Web Services on a service called Amazon Comprehend Medical, Silgard and her colleagues have accelerated this process more than 5,000-fold, developing algorithms that can churn through 10,000 notes per hour and dramatically improve researchers’ access to this wealth of information.
Wearables and predictive algorithms
Kreizenbeck addressed how the cloud is helping bring together experts hoping to improve cancer care. Side effects from chemotherapy send many patients on expensive, emotionally draining and potentially preventable trips to the emergency department. To oncologists and researchers, these trips represent unnecessary suffering and cost, said Kreizenbeck, director of research and partnerships at the Hutchinson Institute for Cancer Outcomes Research. Most importantly, they are terrible experiences for patients, she said, and they can get in the way of effective treatment.
Working with Microsoft and using the Azure cloud computing platform, Hutch researchers are developing predictive algorithms that draw on data collected by a wearable biosensor which are then entered into an app to identify patients undergoing chemotherapy who might need support — hours or days before they reach a crisis point. The possibility of intervening early, taking the emergency department out of the equation, and enabling patients to move on with their scheduled treatments could be transformative, Kreizenbeck said.
Data visualization
Alexion gave attendees a glimpse of Oncoscape, an open-source data analysis and visualization tool being developed at Fred Hutch.
She demonstrated an example of how Oncoscape, which uses the latest machine-learning methods and can render millions of datapoints at a time, can map molecular and genetic information from a patient’s tumor, link it with their health history, and place their story within the context of thousands of other patients with similar medical or tumor characteristics. Using Oncoscape, researchers can look for patterns in demographics data, diagnosis and treatment information, survival data and more.
“Data is all about patterns,” she said. “Oncoscape shows us patterns in data like never before.”
‘People cure cancer’
As the keynote ended, Alexion returned to the question of whether the cloud could cure cancer.
“People cure cancer, and the cloud enables people,” she said, emphasizing that her colleagues in the audience help make those cures possible by continuing to devise new ways to utilize the cloud. That innovation, she said, “is what drives me, and all of us, to continue to find cures.”
