Rush using ML, analytics on images and unstructured data
Rush University Medical Center is adopting machine learning and analytics technologies from two companies to process patient information, including from imaging studies and other sources, with hopes of customizing patient treatment and delivering precision medicine.
The Chicago-based academic medical center is using a combination of technology from Cloudera and MetiStream, which are working together on products that providers can use to improve patient outcomes. Cloudera offers a platform for machine learning and analytics optimized for the cloud, while MetiStream develops healthcare analytics solutions.
MetiStream offers an interactive analytics platform for healthcare and life science industries built on Cloudera's machine learning platform. By combining machine learning and analytics from Cloudera Enterprise and Cloudera Data Science Workbench, MetiStream contends its Ember product can deliver insights across massive volumes of handwritten clinical notes as well as genomic data.
If successful, such a combination could provide a way for healthcare organizations to cost-effectively improve genomic research and accelerate the pace at which insights could be used to influence patient care.
Much of healthcare data, particularly imaging studies, is unstructured, Rush and the companies contend, and that makes it difficult for legacy data storage and analytics platforms to process, analyze and correlate data across a patient population. Mixing Cloudera’s platform with MetiStream’s healthcare analytics solution can help organizations capture relevant information from diverse datasets, such as imaging studies, clinical notes, genomics and EHR data, then correlate it to gain improvement in patient care and work processes.
The technology is being employed at Rush University Medical Center, an academic health system composed of Rush University Medical Center, Rush University, Rush Oak Park Hospital and Rush Health. The system needed a healthcare analytics platform to process a backlog of clinical notes. Using solutions from both Cloudera and MetiStream, the medical center was able to process 7.2 million records in less than 36 hours. As a result of using the healthcare analytics platform, the medical center improved the standard of care by identifying patients with certain disease risks.
"With Cloudera and MetiStream on Microsoft Azure, we can quickly spin up and down resources as our data processing needs change and evolve, and we can load huge volumes of data in days that would have taken weeks on premises," says Bala Hota, MD, chief analytics officer at Rush. “We have also been able to apply machine learning to discover new insights from our data, and by using Cloudera technologies, we are working to make development of new models easier and faster for our data scientists.”
The combination of machine learning and analytics technologies offers provider organizations more opportunities to use a wide gamut of unstructured healthcare data, including imaging studies, to make care decisions more quickly, company executives contend.
"Machine learning and analytics are powerful tools for understanding diseases, improving outcomes, containing costs and delivering better care," said Mike Olson, founder and chief strategy officer at Cloudera. "Today, healthcare organizations can do what was previously impossible. They can integrate complex data sets from imaging, EHR and genomics with machine learning and analytics at massive scale."