biomedical systems modeling &...
TRANSCRIPT
![Page 1: Biomedical Systems Modeling & Simulationmeche.mit.edu/sites/default/files/20120625_Konstantinos.pdf · 2015-08-10 · Modeling Chaos in Physiology •Chaos (χᾶος) = preponderant](https://reader034.vdocument.in/reader034/viewer/2022042111/5e8c6dd8cab12058af64c85f/html5/thumbnails/1.jpg)
Biological Systems Modeling & Simulation
Konstantinos P. Michmizos, PhD
June 25, 2012
②
![Page 2: Biomedical Systems Modeling & Simulationmeche.mit.edu/sites/default/files/20120625_Konstantinos.pdf · 2015-08-10 · Modeling Chaos in Physiology •Chaos (χᾶος) = preponderant](https://reader034.vdocument.in/reader034/viewer/2022042111/5e8c6dd8cab12058af64c85f/html5/thumbnails/2.jpg)
Previous Lecture
• Biomedical Signal examples (1-d, 2-d, 3-d, …)
• Purpose of Signal Analysis
• Noise
• Frequency domain (1-d, 2-d)
• Filtering unwanted frequencies
• Sampling
• Kalman Filter
• Recursive Algorithm
Konstantinos Michmizos
![Page 3: Biomedical Systems Modeling & Simulationmeche.mit.edu/sites/default/files/20120625_Konstantinos.pdf · 2015-08-10 · Modeling Chaos in Physiology •Chaos (χᾶος) = preponderant](https://reader034.vdocument.in/reader034/viewer/2022042111/5e8c6dd8cab12058af64c85f/html5/thumbnails/3.jpg)
Biological Signals
• Amplitude limited
• Distorted by colored noise
• Limited length
• Non-stationary
• The underlying system is unknown
Konstantinos Michmizos
![Page 4: Biomedical Systems Modeling & Simulationmeche.mit.edu/sites/default/files/20120625_Konstantinos.pdf · 2015-08-10 · Modeling Chaos in Physiology •Chaos (χᾶος) = preponderant](https://reader034.vdocument.in/reader034/viewer/2022042111/5e8c6dd8cab12058af64c85f/html5/thumbnails/4.jpg)
Why do we need models?
Insight • Understanding the underlying system • Testing bed for several hypotheses Description • Design verification • Explore linear behavior / non-linear nature of the
biological physics (going beyond human senses) Control • Predictions • Functional limits
Konstantinos Michmizos
![Page 5: Biomedical Systems Modeling & Simulationmeche.mit.edu/sites/default/files/20120625_Konstantinos.pdf · 2015-08-10 · Modeling Chaos in Physiology •Chaos (χᾶος) = preponderant](https://reader034.vdocument.in/reader034/viewer/2022042111/5e8c6dd8cab12058af64c85f/html5/thumbnails/5.jpg)
Definition of a Model
Konstantinos Michmizos
Difficulties 1. No access to the
unconditional truth 2. No unbiased observer 3. Personal opinion on a
hypothesis (abduction) 4. Abduction is not an
infallible way for discovering truth
abduction
![Page 6: Biomedical Systems Modeling & Simulationmeche.mit.edu/sites/default/files/20120625_Konstantinos.pdf · 2015-08-10 · Modeling Chaos in Physiology •Chaos (χᾶος) = preponderant](https://reader034.vdocument.in/reader034/viewer/2022042111/5e8c6dd8cab12058af64c85f/html5/thumbnails/6.jpg)
Modeling a System
• System
(a collection of interconnected processes)
• Model
(a representation that approximates the behavior of an actual system)
• Simulation
(puts the model to work)
Konstantinos Michmizos
![Page 7: Biomedical Systems Modeling & Simulationmeche.mit.edu/sites/default/files/20120625_Konstantinos.pdf · 2015-08-10 · Modeling Chaos in Physiology •Chaos (χᾶος) = preponderant](https://reader034.vdocument.in/reader034/viewer/2022042111/5e8c6dd8cab12058af64c85f/html5/thumbnails/7.jpg)
Goodness of a Model
Konstantinos Michmizos
Accuracy Difference between predicted and true value Precision Reproducibility of the results
![Page 8: Biomedical Systems Modeling & Simulationmeche.mit.edu/sites/default/files/20120625_Konstantinos.pdf · 2015-08-10 · Modeling Chaos in Physiology •Chaos (χᾶος) = preponderant](https://reader034.vdocument.in/reader034/viewer/2022042111/5e8c6dd8cab12058af64c85f/html5/thumbnails/8.jpg)
Model refinement
Konstantinos Michmizos
![Page 9: Biomedical Systems Modeling & Simulationmeche.mit.edu/sites/default/files/20120625_Konstantinos.pdf · 2015-08-10 · Modeling Chaos in Physiology •Chaos (χᾶος) = preponderant](https://reader034.vdocument.in/reader034/viewer/2022042111/5e8c6dd8cab12058af64c85f/html5/thumbnails/9.jpg)
Classification of Models
Konstantinos Michmizos
Sub-cellular Cellular Intercellular Tissue Organ Organism
Dynamic StaticBehavior
Deterministic Stochastic
Level
Chance
ParametricNon-
parametric
Single Compartment
Multiple Compartments
Parameters
Compartments
LinearNon-LinearLinearity
Time FrequencyDomain
![Page 10: Biomedical Systems Modeling & Simulationmeche.mit.edu/sites/default/files/20120625_Konstantinos.pdf · 2015-08-10 · Modeling Chaos in Physiology •Chaos (χᾶος) = preponderant](https://reader034.vdocument.in/reader034/viewer/2022042111/5e8c6dd8cab12058af64c85f/html5/thumbnails/10.jpg)
Linear Modeling of Physiological Systems
Konstantinos Michmizos
Why linear? • Simple to implement • Powerful analysis tools
Transforming input into output • Time variant • Time invariant
Linear Time-invariant Systems 1. Superposition 2. Impulse Response of a system
• Testing for linearity may be done using the principle of superposition (e.g. same input at different amplitudes) • Response of a linear system to a sinusoidal input is a sinusoid at the
same frequency • Linearity of a system means nothing without its range
![Page 11: Biomedical Systems Modeling & Simulationmeche.mit.edu/sites/default/files/20120625_Konstantinos.pdf · 2015-08-10 · Modeling Chaos in Physiology •Chaos (χᾶος) = preponderant](https://reader034.vdocument.in/reader034/viewer/2022042111/5e8c6dd8cab12058af64c85f/html5/thumbnails/11.jpg)
Impulse Response Function
System
Konstantinos Michmizos
Negative Values = System anticipation Positive Values = System memory
Living systems frequently demonstrate predictive behavior (e.g. visual pursuit)
![Page 12: Biomedical Systems Modeling & Simulationmeche.mit.edu/sites/default/files/20120625_Konstantinos.pdf · 2015-08-10 · Modeling Chaos in Physiology •Chaos (χᾶος) = preponderant](https://reader034.vdocument.in/reader034/viewer/2022042111/5e8c6dd8cab12058af64c85f/html5/thumbnails/12.jpg)
Coherence
If Coherence is greater than 0 but not 1: 1. Extraneous noise is present in the measurements. 2. The system is not linear. 3. Output is due to the input as well as to other inputs.
Konstantinos Michmizos
![Page 13: Biomedical Systems Modeling & Simulationmeche.mit.edu/sites/default/files/20120625_Konstantinos.pdf · 2015-08-10 · Modeling Chaos in Physiology •Chaos (χᾶος) = preponderant](https://reader034.vdocument.in/reader034/viewer/2022042111/5e8c6dd8cab12058af64c85f/html5/thumbnails/13.jpg)
Modeling Chaos in Physiology
• Chaos (χᾶος) = preponderant of all the others
• Physiological Systems are extremely complicated deterministic systems for their observers
• Stochastic determinism observed in biological signals (extraordinary sensitivity to internal conditions)
• Presence of order under the absence of periodicity (strange attractor)
• A chaotic system is not necessarily complex
Konstantinos Michmizos
![Page 14: Biomedical Systems Modeling & Simulationmeche.mit.edu/sites/default/files/20120625_Konstantinos.pdf · 2015-08-10 · Modeling Chaos in Physiology •Chaos (χᾶος) = preponderant](https://reader034.vdocument.in/reader034/viewer/2022042111/5e8c6dd8cab12058af64c85f/html5/thumbnails/14.jpg)
Chaos in Physiological Systems
Chaos in the heart
• Sinoatrial Node
• Sympathetic & parasympathetic fibers & respiration rate heart rate variability
Chaos in the brain
• Neuron doctrine
• Feedbacks internal uncertainties are amplified over time
Konstantinos Michmizos
![Page 15: Biomedical Systems Modeling & Simulationmeche.mit.edu/sites/default/files/20120625_Konstantinos.pdf · 2015-08-10 · Modeling Chaos in Physiology •Chaos (χᾶος) = preponderant](https://reader034.vdocument.in/reader034/viewer/2022042111/5e8c6dd8cab12058af64c85f/html5/thumbnails/15.jpg)
Why Chaos?
• Better adaptation capabilities
(e.g. heart pumps blood in various circumstances)
• Adaptation in the brain = learning
A never seen before stimulus in the brain, moves the brain to an un-patterned chaotic state.
• Chaos = normal function
Disease = acute attack of order against chaos
(e.g. the amount of chaos in the Parkinsonian brain decreases as neurons become more synchronized)
Konstantinos Michmizos