The technology is a method for creating a personalized multivariate time series representation (MTSA) of physiological data that physicians can visually interpret. Patterns in the MTSA are used to differentiate between medical conditions such as renal and respiratory failure. The method first inputs time-series data for each parameter (heart rate, temperature, etc.). It then fills in time gaps in the measurements through linear regression. Next, it uses a Symbolic Aggregate Approximation (SAX) method to normalize the data, create time segments of the data, and assign discrete values to each segment. Radial representations of the various discrete values can be displayed for all parameters at once. In addition, multiple radial representations can be overlaid to display changes over time. Patterns can then be detected at both the parameter level and over time.
Technical Details:
Medical professionals can capture up to 350 different types of vital signs and laboratory reports on a patient in intensive care. However, this data is under-utilized. Current methods for analyzing this data focus on measurements at a single point in time, rather than considering the vitals over time. Examining the vital signs and laboratory results together in a multivariate time series provides greater insight into how the body and its vital organs function as a whole and a means by which to better diagnose and treat a patient.
Looking for Partners:
The technology represents the next step in vital sign analysis where real time data is visually presented to a medical professional in a manner that combines multiple time series parameters. Applications include Intensive Care Unit monitors, health record analytics, and mobile patient care.
