Simulation, identification and visualization-based analysis of multi-parameter distributed capillary-tissue transport models.

摘要

In order to successfully use a model for parameter identification it must be carefully analyzed. Current analysis methods, however, are ad-hoc and provide only partial information. These methods were extended through the application of a scientific visualization method, stacked dimensions. The end result of these extensions are multi-dimensional parametric model-images. These images depict, in a single graphic, a model as a function of all its parameters. We applied parametric model-images to model verification (behavioral analysis), sensitivity analysis, and identifiability analysis. In addition, a method for carrying out parameter identification visually (visual regression) was developed. The new methodology was applied to the evaluation of pulmonary vascular capillary-transport models. Results have shown that the visualization-based method provides a more complete view of a model's behavior and its other characteristics, yielding a greater confidence in the model then previously possible. Furthermore, these methods have also proven to be more computationally efficient than the traditional approaches.; To carry out the visualization-based model analysis two software packages were developed. EN-DIMAN (Environment for N-DIMensional data ANalysis) was written for displaying the data as parametric model-images. GLANSE (Graphical Lung ANalysiS Environment) was written as the model simulation and analysis environment. It was used as the simulation engine to generate the data used for visualization.