Predicting an Individual's Physiologic State without a Crystal Ball

Abstract

This talk illustrates two approaches by which biomathematical models can be developed to construct individualized, i.e. subject specific, physiologic predictive algorithms. We describe, in layman's terms, the pros and cons of first-principles physiology-based algorithms and data-driven, autoregressive algorithms. We also discuss how these algorithms may be customized to predict the physiologic state of specific individuals some time into the future. We illustrate the predictive power of these approaches in the prediction of: (i) performance impairment due to total sleep, (ii) body core temperature during physical activity, and (iii) glucose levels of type 1 and type 2 diabetes patients. In conjunction with real-time physiologic monitoring devices, such predictive algorithms may be used to optimize the timing and dosing of fatigue countermeasures, e.g. naps and caffeine, so that performance peaks and is maintained during desired times of day, minimize the occurrence of heat-related injuries, such as heat strokes, and allow for proactive glucose regulatory interventions be for glycemic levels drift from the desired range. [TATRC website, 15 Dec 2008]

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Apr 05, 2008
Accession Number
ADA490361

Entities

People

  • Jaques Reifman

Organizations

  • United States Army Medical Research and Development Command

Tags

Communities of Interest

  • Biomedical
  • Human Systems

DTIC Thesaurus Topics

  • Abstracts
  • Algorithms
  • Army
  • Biomedical Research
  • Coefficients
  • Diabetes
  • Differential Equations
  • Health Care
  • Health Services
  • Heat Stroke
  • Measurement
  • Monitoring
  • Physiological Monitoring
  • Predictive Modeling
  • Reaction Time
  • Sleep Deprivation
  • Wounds And Injuries

Readers

  • Circadian Sleep-Wake Regulation and Chronobiology
  • Computational Modeling and Simulation
  • Gulf War Illness and Chronic Multisymptom Illness in Veterans.