Encyclopedia of Computational Neuroscience

Living Edition
| Editors: Dieter Jaeger, Ranu Jung

Dynamics of Disease States: Overview

  • John MiltonEmail author
Living reference work entry

Latest version View entry history

DOI: https://doi.org/10.1007/978-1-4614-7320-6_781-2


Deep Brain Stimulation Temporal Lobe Epilepsy Cochlear Implant Stochastic Resonance Balance Control 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Detailed Description

The evolution of an illness is one of the clues that a bedside physician uses to arrive at a diagnosis and treatment strategy for diseases that affect the nervous system. Is the onset acute or subacute? Is the clinical course self-limited, relapsing-remitting, cyclic, and chronic progressive? The impetus for studying disease dynamics comes from the mathematics and physics communities: their long experience has shown that insights into mechanism often derive from examining how dynamics change. Consequently, the time evolution of a disease is modeled as differential equations, and disease processes are described in terms of the origin, stability, and bifurcations of the model’s dynamical behaviors. In 1977, Michael Mackey and Leon Glass associated changes in physiological dynamics from healthy to unhealthy with changes in underlying control parameters (Mackey and Glass 1977). Subsequently, this concept of a dynamical disease was extended to that of a dynamic disease to account for the possibility that mechanisms other than those associated with changes in parameters may be involved (Dynamic Diseases of the Brain).

Computational neuroscience extends dynamical approaches to neurological disease in two ways. First, by making it possible to include anatomical, physiological, and molecular details, computational models provide insights into how mechanisms acting at the level of molecules and individual neurons translate into phenomena manifested clinically at the bedside (Epilepsy: Abnormal Ion Channels; Modeling of Disease – Molecular Level: Overview). Historically the two neurological diseases which provided the most insight into cortical function were temporal lobe epilepsy and classical migraine. Indeed the study of the geometry of migraine, drug- and flicker-induced visual fortification patterns (Stochastic Neural Field Theory; Visual Hallucinations and Migraine Auras; Flicker-Induced Phosphenes), and their propagation (Migraines and Cortical Spreading Depression) has provided deep insights into the functional architecture of the visual cortex. Surgical approaches for the treatment of patients with medically intractable epilepsy provided the impetus to directly record from the cortex of awake humans. This, in turn, motivated studies into the ability of large populations of neurons to synchronize and generate seizures (LFP Analysis: Overview; Transfer Function, Electrocortical; Epilepsy: Computational Models; Epilepsy, Neural Population Models of). Moreover, computational models are now making it even possible to gain insights into the nature of cognitive, functional, and psychiatric diseases of the nervous system that up to now have largely remained mysterious (Computational Psychiatry).

Second, by having “a disease in a computer model,” it is possible to efficiently evaluate and refine treatment strategies in silico before applying them to humans. Computational challenges arise because of the presence of multistability (Multistability in Seizure Dynamics; Multistability: Stopping Events with Single Pulses; Multistability in Motor Control) and time delays (Time-Delayed Neural Networks: Stability and Oscillations). Indeed time-delayed, multistable dynamical systems have a tendency to generate transient oscillations that can be easily mistaken for limit cycle oscillations, thus causing confusion (Pakdaman et al. 1998). Nonetheless computational approaches have already proven useful to develop therapeutic strategies.

Although the ultimate goal of medicine is cure, one cannot overlook the need to improve the patient’s quality of life when cure cannot be achieved. The electrical properties of neurons make it possible to use electrical stimuli as a treatment modality. Applications range from aborting seizures with electrical stimuli (Multistability: Stopping Events with Single Pulses) to improving the quality of movements of patients with Parkinson’s disease with deep brain stimulation (Parkinson’s Disease: Deep Brain Stimulation; Computational Models of Deep Brain Stimulation (DBS); Deep Brain Stimulation (Models, Theory, Techniques): Overview). Even noisy stimuli can be beneficial by enhancing the detection of weak signals by the sensory nervous system in order to improve the function of cochlear implants in the hearing impaired or balance control in those with peripheral neuropathies (Stochastic Resonance: Balance Control, Cochlear Implants).

As insight increases in our understanding of neural encoding, it is becoming possible to replace broken parts with electronic ones that perform the same function (Neural Prosthesis and Rehabilitation, Biomedical Applications of Motoneuron and Neuromuscular Models). A large number of articles in this Encyclopedia point to the current enthusiasm in this therapeutic approach. Applications include restoring vision to the visually impaired (Vision Prosthesis), hearing to those who cannot hear (Peripheral Nerve Interface Applications, Cochlear Implants), continence to those incontinent, and relief from pain to those who suffer (Peripheral Nerve Interface Applications, Neuropathic Pain). Dramatically, it has become possible to interface the brain directly with electronic devices and make it possible for a patient to move robotic limbs by thought alone (Cortical Motor Prosthesis; Functional Neuroscience: Cortical Control of Limb Prosthesis).

The frontier for dynamic disease is to understand the collective behaviors of the nervous system that emerge over the timescale of years. Can the development of an epileptic focus (epileptogenesis) be halted early so that an individual at risk never experiences a seizure? Can the rate of learning of a complex voluntary skill, such as the golf swing, by a patient with a robotic or stem cell-derived limb prosthesis be sped up to the point that the individual could enjoy the use of these limbs throughout a lifetime? Large, complex physical systems tend to self-organize dissipative structures, namely, dynamical entities whose existence is maintained far from equilibrium by a supply of energy. Already dynamical signatures of this self-organization, including power laws, have been observed in the bursting propagating activities of living neural populations (Neuronal Avalanches) and the dynamics of human balance control (Human Balancing Tasks: Power Laws, Intermittency and Levy Flights).

Computational neuroscience provides the tools for understanding how the nervous system learns to exert control, thereby bringing to many tangible hope of a better life.



  1. Mackey MC, Glass L (1977) Oscillation and chaos in physiological control systems. Science 197:287–289PubMedCrossRefGoogle Scholar
  2. Pakdaman K, Grotta-Ragazzo C, Malta CP (1998) Transient regime dynamics in continuous-time neural networks with delays. Phys Rev E 58:3623–3627CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media New York 2014

Authors and Affiliations

  1. 1.W.M. Keck Science DepartmentThe Claremont CollegesClaremontUSA