Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4...
Transcript of Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4...
![Page 1: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/1.jpg)
Simple Neuron Models:
FitzHugh-Nagumo and Hindmarsh-Rose
R. ZillmerINFN, Sezione di Firenze
• Reduction of the Hodgkin-Huxley model• The FitzHugh-Nagumo model• Phase plane analysis• Excitability (threshold-like behavior), periodic spiking (Hopf bifurcation)• The Hindmarsh-Rose model for bursting neurons
![Page 2: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/2.jpg)
Neuron models (sketch)
Single Neurons
Hodgkin−Huxley, 1952− current based
detailed, specific models− compartmental (structure)− more currents− adaptive (state−dep. prop.)
low−dimensional models− FitzHugh−Nagumo, 1960’s− Hindmarsh−Rose, 1980’s
Networks* effective numerical simulation* allow for most common features − excitability − spiking, different time scales
integrate−and−fire modelsstochastic models
Hopfield network, 1980’s− on−off neuron, learning, stat. physics
experiments
reduction
simplification
abstraction
1
![Page 3: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/3.jpg)
Hodgkin-Huxley model
• neuronal signals are short electrical pulses: spikes or action potentials on msecscale
• intracellular: incoming spike modifies membran potential
Hodgkin-Huxley (1952): Semirealistic 4-dimensional model for the dynamics of themembran potential by taking into account Na+, K+, and a leak current. Dynamics of ionchannels highly nonlinear ⇒ emergence of chaotic evolution.
membran potential:dVdt
= CNam3h(ENa−V)+CK n4(EK−V)+Cleak(Vrest−V)+ Iinj(t)
sodium INa, fast:dmdt
= αm(V)(1−m)−βm(V)m
slow:dhdt
= αh(V)(1−h)−βh(V)h
potassium IK , slow:dndt
= αn(V)(1−n)−βn(V)n
2
![Page 4: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/4.jpg)
Dynamics of currents m, h, n
General form:dxdt
=− 1τ(V)
[x−xs(V)]
Solution for constant V : x(t) = (x0−xs)exp(−t/τ)+xs
⇒ exponential relaxation to steady state value xs
For varying V(t) : x(t) follows varying steady state value xs(t)
small τ : fast relaxation ⇒ x(t) ≈ xs(t)large τ : slow dynamics
3
![Page 5: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/5.jpg)
Reduction to two-dimensional model
fast sodium dynamics:
approximate by steady state value: m(t) ≈ ms(V)
similar dynamics of slow sodium and potassium:
replace h(t) , n(t) by one effective current w(t)
⇒ two equations for temporal evolution of V(t) and w(t)
4
![Page 6: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/6.jpg)
FitzHugh-Nagumo model
FitzHugh (1961) and Nagumo, Arimoto, Yoshizawa (1962) derived 2-dimensional modelfor an excitable neuron:
membran potential:dvdt
= v− v3
3−w+ I
current variable:dwdt
=1τ(v+a−bw)
typical values: a = 0.7, b = 0.8, τ = 13
⇒ vw∼ 10 ⇒ w slow , v fast
For constant input I = constno chaotic evolution
5
![Page 7: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/7.jpg)
Phase plane analysis
Two-dimensional flow field:
~F(v,w) =ddt
(vw
)=
(v− v3
3 −w+ I1τ(v+a−bw)
)
(numerical) solution:(
v(t)w(t)
)⇒ trajectory in 2-D plane
Characteristics:
• trajectories cannot cross (uniqueness of solutions)• nullclines define lines in the 2-D plane:
v = 0 ⇒ w = v− v3
3 + I
w = 0 ⇒ w = (v+a)/b
• crossings of the nullclines correspond to fixed points (stable for I = 0)
6
![Page 8: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/8.jpg)
Phase plane portrait of FitzHugh-Nagumo model for I = 0
arrows indicate flow field (v, w)
-3 -2 -1 0 1 2v
-0,5
0
0,5
1
wv=0w=0
. .
7
![Page 9: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/9.jpg)
Subthreshold pulse injection
injection of weak pulse I(t) = I0δ(t− t0) : fast return to FP
-3 -2 -1 0 1 2v
-0,5
0
0,5
1
wv=0w=0
. .
8
![Page 10: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/10.jpg)
Subthreshold pulse injection
no action potential
0 10 20 30 40 50t
-1,2
-1
-0,8
-0,6
-1,2 -1 -0,8-0,63
-0,6
-0,57
-0,54
v
w
9
![Page 11: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/11.jpg)
Suprathreshold pulse injection
stronger pulses: large excursion in phase plane
-3 -2 -1 0 1 2v
0
wv=0w=0
. .
10
![Page 12: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/12.jpg)
Suprathreshold pulse injection
spike response – action potential generation
0 10 20 30 40 50 60 70t
-2
-1
0
1
2
-2 -1 0 1 2
-0,5
0
0,5
1
v
w
11
![Page 13: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/13.jpg)
Refractory period
Immediately after spike the neuron is indifferent to further input
0 10 20 30 40 50 60 70t-3
-2
-1
0
1
2
v(t)
-2 -1 0 1 2
-0,5
0
0,5
1
refractory period
12
![Page 14: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/14.jpg)
FitzHugh-Nagumo model for constant I > 0
Phase plane analysis:I shifts nullcline of v, nullcline of w unaffected
v = 0 : w = v− v3
3+ I , w = 0 : w = (v+a)/b
⇒ for large enough I > 0.33 the fixed point, v = w = 0, becomes unstable
⇒ Onset of sustained oscillations (Hopf-bifurcation)
13
![Page 15: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/15.jpg)
Nullclines for constant I > 0
v - nullcline shifted ⇒ for I > 0.33 the fixed point becomes unstable
-2 -1 0 1 2v
-2
-1
0
1
2
3
w
I=0.4I=0
v=0
w=0.
.
14
![Page 16: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/16.jpg)
Below the bifurcation, I = 0.3
Fixed point remains stable ⇒ small damped oscillations
-2 -1,5 -1 -0,5 0v
-1
-0,5
0
0,5
1
w
v=0
w=0.
.
15
![Page 17: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/17.jpg)
Below the bifurcation, I = 0.3
Fixed point remains stable ⇒ small damped oscillations
0 25 50 75 100t
-1,4
-1,2
-1
-0,8
v(t)
16
![Page 18: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/18.jpg)
Above the bifurcation, I = 0.4
Fixed point unstable ⇒ Hopf-bifurcation to sustained oscillations on limit cycle
-3 -2 -1 0 1 2v
-1
0
1
2
w
v=0
w=0.
.
17
![Page 19: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/19.jpg)
Above the bifurcation, I = 0.4
Fixed point unstable ⇒ periodic spiking
0 50 100t
-2
-1
0
1
2
v(t)
18
![Page 20: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/20.jpg)
FitzHugh-Nagumo model for varying I(t)
Recapitulation:
• For I = const> 0.33 onset of stable oscillations with Frequency Ω(I)• Refractory period where system is rather indifferent to external signals
Time dependent input:
• periodic signals: resonance effects• noisy signals: coherence resonance
19
![Page 21: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/21.jpg)
Summary FitzHugh-Nagumo
• two dimensional model that can be derived from Hodgkin-Huxley via reduction ofvariables
• allows effective phase plane analysis• ecitable: spike response to suprathreshold input pulse• refractory period• with increasing input current Hopf-bifurcation to sustained periodic spiking
• reduction of complexity: no self-sustained chaotic dynamics• no bursting• few parameters: difficult to adapt to neurons with specific properties
20
![Page 22: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/22.jpg)
The Hindmarsh-Rose model
Developed 1982-1984 by J. L. Hindmarsh and R. M. Rose to allow for rapid firing orbursting
Idea:Allow for triggered firing, i.e., switch between a stable rest state and a stable limit cycle(rapid periodic firing)⇒ more than one fixed points required: can be achieved by deformation of the null-clines (nonlinear “current” equation)
Basic equations:
dxdt
= 3x2−x3−y+ I ,dydt
= 5x2−1−y
Nullclines:
x = 0 : y = 3x2−x3+ I , y = 0 : y = 5x2−1
21
![Page 23: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/23.jpg)
Phase portrait of Hindmarsh-Rose model
3 Fixed points ⇒ coexistence of rest state and limit cycle
-2 -1 0 1 2x
0
4
8
12
16
20
y
dx/dt=0dy/dt=0
stable
saddle
unstable
22
![Page 24: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/24.jpg)
Adaption variable
Termination of firing via additional adaption variable z that should:
• lower the effective current when neuron is firing• return to zero when x has reached its rest state value xr
Complete equations:
dxdt
= 3x2−x3−y+ I −z ,dydt
= 5x2−1−y ,dzdt
= r [s(x−xr)−z]
23
![Page 25: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/25.jpg)
Bursting of Hindmarsh-Rose model
After repeated firing the dynamics returns to the stable fixed point
-2 -1 0 1 2x
0
4
8
12
y
24
![Page 26: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/26.jpg)
Bursting of Hindmarsh-Rose model
Several spikes with varying interspike-interval (ISI)
0 25 50 75 100t-2
-1
0
1
2
x(t)
25
![Page 27: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/27.jpg)
Features of the Hindmarsh-Rose model
3-D model for neuron with rapid firing
Suitable choice of parameters allows for
• regular bursting• chaotic bursting
Suitable choice of parameters ? ⇐⇒ ? real neurons
26
![Page 28: Simple Neuron Models: FitzHugh-Nagumo and Hindmarsh-Roseatorcini/ARTICOLI/lezione4-zillmer.pdf · 4 8 12 16 20 y dx/dt=0 dy/dt=0 stable saddle unstable 22. Adaption variable Termination](https://reader033.fdocuments.in/reader033/viewer/2022050100/5f3fc3631cf4604b0b15353b/html5/thumbnails/28.jpg)
Further reading
• W. Gerstner and W. M. Kistler, Spiking Neuron Models: Single Neurons, Popula-tions, Plasticity,http://diwww.epfl.ch/ gerstner/SPNM/SPNM.html .
• C. Koch, Biophysics of Computation: Information Processing in Single Neurons(Computational Neuroscience), Oxford University Press.
• J. Hindmarsh and P. Cornelius, The Development of the Hindmarsh-Rose modelfor bursting,www.worldscibooks.com/lifesci/etextbook/5944/5944−chap1.pdf .
27