4th Annual Mountain West
Biomedical Engineering Conference
September 5-6, 2008
Abstract Details
Presented By: | Millington, Jared |
Affiliated with: | University of Utah, Biomedical Engineering |
Authors: | Jared Millington, Youssef T. AL-sheikh, Joseph Andrade |
From: | Department of Biomedical Engineering, University of Utah |
Title
Abstract
Multi-analytical data generated from biomedical diagnostic and monitoring devices can be displayed via simple iconic patterns to ease interpretation and enable fast diagnostic decision-making. We developed and employed a software program MACROPatterns (Multi-Analytical Chemistry-Recognizer of Optical Patterns) for this purpose. MACROPatterns is a multi-dimensional visualization program that enables simultaneous interpretation of measurements along with the recognition of correlated or uncorrelated patterns. Employing visualization concepts and approaches in MACROPatterns required knowledge in various fields, including visual cognition and recognition, art, scientific visualization, and software design. MACROPatterns provides easy navigation through different pathologic panels and various patterns of each panel. This could be very useful especially in clinical education and training. MACROPatterns can be utilized in different analytical fields that require interpretation of multiple measurements along with the recognition of correlated or uncorrelated patterns. Visualization concepts and approaches include: 1) Visual perception: human visual perception performs best at one-dimensional space, and the larger the dimensionality of space is, the weaker human visual perception becomes; 2) Visual recognition: displaying and adding attributes to useful data; and 3) Display efficiency: projecting useful data onto a two-dimensional physical medium, clustering data into simple visual patterns, adding attributes (color, brightness, transparency, or shape distortion) to patterns, and displaying multiple patterns on one screen to ease comparative interpretation. Multi-analytical clinical chemistry data visualization via simple iconic displays is demonstrated in the diagnosis of galactosemia and hyperlipoproteinemia.