Skip to content Skip to main navigation Report an accessibility issue

Improving Intensive Care With Data

The first intensive care unit (ICU) is believed to have been established in Copenhagen, Denmark in 1953. Since that time ICUs have become a standard feature in hospitals, constantly updating their life saving techniques and equipment at their disposal. Xiaopeng Zhao, associate professor of mechanical, aerospace, and biomedical engineering, hopes to add to that list with the results of his current research.

Xiaopeng Zhao, associate professor of mechanical, aerospace, and biomedical engineeringZhao’s Joint Directed Research Development (JDRD) project is focused on taking all the unstructured data generated by ICU patients and organizing it into a scalable infrastructure of patient care management. ICU patients generate a huge amount of data, from the information collected by whatever machines may be used to monitor them, to observations made by hospital staff.

One of the challenges with ordering this data is how different it is. One ICU patient may have data from a cardiac monitor, while another has data from a blood test. Monitors used and tests performed will vary based on the patient’s particular condition. Additionally, there is a great deal of opportunity for human error in the data collection and entry.

Zhao’s team’s goal is to create a tool for structuring this data, which can then provide hospitals and doctors with useful information to inform their practices and individual patient care.

“If we can look at the ICU record to determine which patients have a higher risk of mortality then we can say those patients may need more intensive care or some special treatment,” Zhao said.

Zhao’s  partner at Oak Ridge National Laboratory (ORNL), Georgia Tourassi, has focused her work on big data management. The techniques developed in her corresponding Laboratory Directed Research Development (LDRD) project will be useful for Zhoa’s work, which will in turn provide a test bed for Tourassi’s work.

“The goal is to look at the patient’s data and understand the structure and information in the data and produce something useful so that the information can be used for patient management,” said Zhao.