Predictability of Burn Depth: Data Analysis and Mathematical Modeling Based on USAARL's Experimental Porcine Burn Data
Abstract
The working hypothesis of this study was: given the output of physical thermal sensors and the actual burns experienced by pigs, it should be possible to develop a mathematical model which can convert the output of the sensors to predicted burn depths in reasonable agreement with the observed burn depths. The specific objectives accomplished are as follows: (1) the previously collected data base was reviewed for error both in coding and in content; (2) the biopsy specimens were re-read using a new set of standards so that corrections for thermally induced tissue shrinkage could be calculated; (3) graphical representations of the data contained in the revised data base were made as a help in understanding the relationships embodied in the data; (4) an empirical, multidiscriminant model was written which predicts either gross (clinical) grade or histopathologic grade given parameters such as heat flux, exposure time, skin temperature, and the like; (5) an analytical model was developed to circumvent the problems associated with the multidiscriminant model regarding the inability to use a flux-time profile; (6) optimization of the analytical model was accomplished using data in the data base, intraskin time- temperature profiles and such data as could be found in the literature, e.g. University of Rochester, Moritz and Henriques and Stoll, as performance criteria. Abstracts of 10 additional reports covering various aspects of the project in more detail and a listing of BRNSIM, the analytical model are appended.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 1979
- Accession Number
- ADA091676
Entities
People
- Francis S. Knox
Organizations
- Louisiana State University