October 26, 2010, 8:56 PM — Researchers at Columbia University Medical Center are turning to data analytics techniques to more quickly spot symptoms of a potentially fatal complication called delayed ischemia that can develop in patients who have suffered a ruptured brain aneurysm.
The hospital on Monday said is using IBM's InfoSphere Streams streaming analytics technology to sift through large volumes of existing patient data to search for patterns that may help spot symptoms of the complication up to 48 hours sooner than is possible now.
Delayed ischemia is a life-threatening condition that sometimes develops in patients who suffer a bleeding stroke. The condition results from a lack of adequate blood supply to the brain and often presents no outward symptoms until it is too late to treat.
The goal of the Columbia University effort is to uncover early warning signs of the condition by analyzing more than 200 elements that are collected from patients in neurological intensive care units. The data analyzed by the IBM tools includes heart rates, temperature, blood pressure, brain and heart activity readings. The tools also analyzes the results of lab tests and observed symptoms.
Michael Schmidt, director of neuromonitoring and informatics at the medical center, said the rapid analysis of such data can help researchers uncover patterns in symptom progression that can't be detected manually. "Hopefully what we will see is that [the data compiled from] patients who have had the complication will look different" from that compiled from patients that didn't develop it, he said.
The detection of early warning signs will help doctors respond sooner and potentially stop the progression of delayed ischemia, he said.
As part of the effort, researchers are comparing data compiled from multiple patients. Schmidt said the research team hopes to have a list of early indicators of delayed ischemia during the first half of next year.
Once the list is compiled, the researchers will determine the validity of the early-warning signs determined by the analysis tools -- a process that could take up to three years. "Then we can set up a clinical trial to see if earlier intervention can help" patients that contract delayed ischemia, Schmidt said.
The IBM streaming analytics technology is designed to enable real-time analysis of data in motion -- such as data captured from patients in neurological intensive care units. The software can handle data from multiple sources and formats including databases, applications, voice, video, databases, and live feeds.
The technology is part of a growing IBM portfolio of data analytics technologies that the company has assembled over the past few years. InfoSphere Streams was developed in-house by IBM, but many of its data analytics products are from its 24 acquisitions in this space over the past few years.
Jaikumar Vijayan covers data security and privacy issues, financial services security and e-voting for Computerworld. Follow Jaikumar on Twitter at @jaivijayan , or subscribe to Jaikumar's RSS feed . His e-mail address is jvijayan@computerworld.com .
Read more about bi and analytics in Computerworld's BI and Analytics Topic Center.
