Accession Number : AD1005103


Title :   PREDICT: Pattern Representation and Evaluation of Data Through Integration, Correlation and Transformation


Descriptive Note : Technical Report,01 Feb 2014,31 Jan 2015


Corporate Author : Trustees of Columbia University in the City of New York New York United States


Personal Author(s) : Schmidt,Michael J


Full Text : https://apps.dtic.mil/dtic/tr/fulltext/u2/1005103.pdf


Report Date : 01 Feb 2015


Pagination or Media Count : 32


Abstract : The objective is to demonstrate that interactive data visualizations designed to transform and consolidate complex multimodal physiological data into integrated interactive displays will reduce clinician cognitive load and will result inreductions in medical error and improvements in patient care, safety, and efficiency. Research accomplishments include: 1) Devised methods to both filter and interpolate physiological data, 2) Created physiologic indices for clinically meaningful variation among several parameters using a version of empirical orthogonal functions, 3) Applied the above physiologic indices to create a patient overview visualization to help clinicians identify patients that are changing and inserted these indices into the sepsis specific decision support visualization, 4) Created a sepsis identification visualization tool to help clinicians identify patients headed for septic shock, and 5) Generated a visualization tool to help clinicians assess intracranial pressure and generate patient specific physiologic set-points for goal directed therapy, and 6) Demonstrate a real-time data feed to a visualization tool that contains elements from each of the three prototype visualizations. In the next period a cognitive experiment will be conducted to determine the effectiveness of these visualization tools.


Descriptors :   MEDICAL RESEARCH , TEST AND EVALUATION , DATA MANAGEMENT , HUMAN FACTOR ENGINEERS , INTENSIVE CARE UNITS , SEPSIS , DECISION MAKING


Distribution Statement : APPROVED FOR PUBLIC RELEASE