|
|
Rubicon Research Repository >
Rubicon Foundation Archive >
UHMS Workshops >
Please use this identifier to cite or link to this item:
http://archive.rubicon-foundation.org/8027
|
| Title: | Survival Analysis & Maximum Likelihood Techniques as Applied to Physiological Model. |
| Authors: | Weathersby, PK
Gerth, WA |
| Keywords: | diving
model
Survival Analysis
Maximum Likelihood Techniques
Physiological Model
decompression sickness
oxygen toxicity
probabilistic model
statistics
Breathing Resistance
Altitude
Log-Logistic Survival Model
Flying After Diving
Cold Exposure Survival Model
Methodology
Predictive Modeling
Experimental Design |
| Issue Date: | 2002 |
| Publisher: | Undersea and Hyperbaric Medical Society |
| Citation: | Weathersby PK, Gerth WA (eds). Survival Analysis & Maximum Likelihood Techniques as Applied to Physiological Model. 51st Undersea and Hyperbaric Medical Society Workshop. UHMS Publication Number WD650. Kensington: Undersea and Hyperbaric Medical Society; 2002; 176 pages. |
| Abstract: | Preface - Wayne A. Gerth: In 1984, Dr. Paul Weathersby, Dr. Lou Homer and Dr. Edward Flynn published a seminal paper in which they introduced survival analysis into the study of decompression sickness (DCS). The approach they outlined, and continued to develop with colleagues Shalini Survanshi, Erich Parker and others at the Naval Medical Research Institute (NMRI) in a subsequent series of published papers and reports, gave new direction to the way we reconcile theory with experience in this field. First, it explicitly recognizes that a given physiological outcome is not an inevitable result of a particular environmental history, but instead is only a probabilistic function of that history. Second, the approach includes rigorous means to make one or more candidate expressions of that probabilistic function each provide its best possible, or optimum, correlation of observed outcomes in actual experience. The optimized models that emerge from such work are consequently quantitative generalizations of that experience, which renew the analyst's focus on the data he or she has in hand. Finally, the approach allows quantitative assessments to be made of how well a given model accounts for observed behavior in specific sets of data, so that the best of a collection of candidate models can be selected. This selection process allows models that are more complex than the data warrants to be identified and deselected, helping to separate necessary theoretical complexity from speculation. Workers interested in other undersea and aerospace physiological problems soon recognized the analytic advantages of survival modeling. Adoption of the techniques in these areas has lead to development of application-specific functions describing responses to ever more complex patterns in the independent variables, and to use of meta-analytic approaches to build data sets with analytically tractable numbers of occurrences of the adverse events of interest. As these applications have ventured farther from those described in standard statistical tests, there has been a growing need to pause and distill their underlying principles, critically evaluate their merits, and outline directions for further development and application. The present Proceedings of the Workshop on Survival Analysis and Maximum Likelihood Techniques as Applied to Physiological Modeling is both an attempt to meet that need, and a salute to the NMRI workers who originally introduced us to this promising line of inquiry. TABLE OF CONTENTS: Preface - Wayne A. Gerth; Welcome - Paul K. Weathersby; Workshop Origin - Edward D. Thalmann; Overview of Survival Functions and Methodology - Wayne A. Gerth; NMRl Models of CNS Oxygen Toxicity Modeling - Paul K. Weathersby; Diver Tolerance to Breathing Resistance - John Clarke; A Log-Logistic Survival Model Applied to Hypobaric Decompression Sickness - Johnny Conkin; Testing of Hypotheses About Basic Mechanisms with Risk Functions - Hugh D. Van Liew; Survival Models for Altitude Decompression Sickness - Nandini Kannan; Multinomial Bubble Score Model - Peter Tikuisis, Keith A. Gault; Probabilistic Models of DCS During Flying After Diving: Motivation for Mechanism - Wayne A. Gerth; Improving on a "Good" Model - Erich C. Parker, Shalini S. Survanshi, Paul K. Weathersby; Meta Analysis of Diver Decompression Data - Paul K. Weathersby, Diana A. Temple, Erich C. Parker; Cold Exposure Survival Model - Peter Tikuisis; Critique of Methodology - Frank E. Harrell, Louis D. Homer; Promising Approaches to Experimental Design - Louis D. Homer; Directions in Statistical Methodology for Multivariable Predictive Modeling - Frank E. Harrell, Jr; General Discussion; Close - Wayne A. Gerth; List of Participants |
| Description: | Copyright Undersea and Hyperbaric Medical Society. |
| URI: | http://archive.rubicon-foundation.org/8027 |
| ISBN: | 0-930406-19-2 |
| Appears in Collections: | UHMS Workshops
|
Files in This Item:
| File |
Description |
Size | Format |
|---|
| 51UHMS.pdf | | 81210Kb | Adobe PDF | View/Open |
|
All items in DSpace are protected by copyright, with all rights reserved.
|
