Dynamical Systems in Neuroscience:

Dynamical Systems in Neuroscience: Dynamical Systems in Neuroscience:

ib.cnea.gov.ar
from ib.cnea.gov.ar More from this publisher
12.07.2015 Views

496 Synchronization (see www.izhikevich.com)nI Na +I K -modelV(t)Vquadratic integrate-and-fire modelx(t)=-coth 2 (t-T)resetspikex reset-1 0 +1xresetresetresetFigure 10.38: Top: Periodic spiking of the I Na +I K -neuron near saddle-node homoclinicorbit bifurcation (parameters as in Fig. 4.1a with τ(V ) = 0.167 and I = 4.49). Bottom:Spiking in the corresponding quadratic integrate-and-fire model.and it can be computed explicitly for some simple p(t).The in-phase synchronized solution, χ = 0, is stable whenH ′ (0) = 1 π∫ π0sin 2 t p ′ (t) dt > 0 .Since the function sin 2 t depicted in Fig. 10.37 is small at the ends of the interval andlarge in the middle, the integral is dominated by the sign of p ′ in the middle. Fast risingand slowly decaying excitatory (p > 0) synaptic transmission has p ′ < 0 in the middle,as in the figure, so the integral is negative and the in-phase solution is unstable. Incontrast, fast rising slowly decaying inhibitory (p < 0) synaptic transmission has p ′ > 0in the middle, so the integral is positive and the in-phase solution is stable. Anotherway to see this is to integrate the equation by parts, reduce it to − ∫ p(t) sin 2t dt,and notice that p(t) is concentrated in the first (left) half of the period where sin 2tis positive. Hence, positive (excitatory) p results in H ′ (0) < 0, and vice versa. Bothapproaches confirm the theoretical results independently obtained by van Vreeswijk etal. (1994) and Hansel et al. (1995) that inhibition, not excitation synchronizes Class1 (SNIC) oscillators. The relationship is inverted for the anti-phase solution χ = π/2(prove it), and no relationships is known for other types of oscillators.10.4.3 Homoclinic oscillatorsBesides the SNIC bifurcation considered above, low frequency oscillations may also beindicative of the proximity of the system to a saddle homoclinic orbit bifurcation, as in

Synchronization (see www.izhikevich.com) 497(a)normalized PRC (Q1( ))0spikedownstroke00sinh 2 ( -T)-n'( )sinh 2 ( -T)numerical0.1T0 0.1T phase of oscillation,T(b)nBphase advancephase delayAVFigure 10.39: (a) Numerically found PRCs of the I Na + I K -oscillator near saddle-nodehomoclinic orbit bifurcation (as in Fig. 10.38) using the MATLAB program in Ex. 12.Magnification shows the divergence from the theoretical curve sinh 2 (ϑ − T ) during thespike. (b) A pulsed input during the downstroke of the spike can produce a significantphase delay (pulse A) or advance (pulse B) not captured by the quadratic integrateand-firemodel.Fig. 10.38, top. The spiking trajectory in the figure quickly approaches a small shadedneighborhood of the saddle along the stable direction, and then slowly diverges fromthe saddle along the unstable direction, thereby resulting in a large period oscillation.As it is often the case in neuronal models, the saddle equilibrium is near a stable nodeequilibrium corresponding to the resting state, and the system is near the co-dimension2 saddle-node homoclinic orbit bifurcation studied in Sect. 6.3.6. As a result, thereis a drastic difference between the attraction and divergence rates to the saddle, sothat the dynamics in the shaded neighborhood of the saddle-node in the figure can bereduced to the one-dimensional V -equation, which in return can be transformed intothe “quadratic integrate-and-fire” formx ′ = −1 + x 2 , if x = +∞, then x ← x reset ,with solutions depicted in Fig. 10.38, bottom. The saddle and the node correspond tox = ±1, respectively. One can check by differentiating that the solution of the modelwith x(0) = x reset > 1, is x(t) = − coth(t − T ), where coth(s) = (e s + e −s )/(e s − e −s )is the hyperbolic cotangent, and T = acoth (x reset ) is the period of oscillation, whichbecomes infinite as x reset → 1.Using the results of Sect. 10.4.1, we find the functionQ(ϑ) = 1/(−1 + coth 2 (ϑ − T )) = sinh 2 (ϑ − T )whose graph is shown in Fig. 10.39a. For comparison, we plotted the numerically foundPRC for the I Na + I K -oscillator to illustrate the disagreement between the theoreticaland numerical curves in the region ϑ < 0.1T corresponding to the downstroke ofthe spike. Such a disagreement is somewhat expected, since the quadratic integrateand-firemodel ignores spike downstroke. If a pulse arriving during the downstroke

Synchronization (see www.izhikevich.com) 497(a)normalized PRC (Q1( ))0spikedownstroke00s<strong>in</strong>h 2 ( -T)-n'( )s<strong>in</strong>h 2 ( -T)numerical0.1T0 0.1T phase of oscillation,T(b)nBphase advancephase delayAVFigure 10.39: (a) Numerically found PRCs of the I Na + I K -oscillator near saddle-nodehomocl<strong>in</strong>ic orbit bifurcation (as <strong>in</strong> Fig. 10.38) us<strong>in</strong>g the MATLAB program <strong>in</strong> Ex. 12.Magnification shows the divergence from the theoretical curve s<strong>in</strong>h 2 (ϑ − T ) dur<strong>in</strong>g thespike. (b) A pulsed <strong>in</strong>put dur<strong>in</strong>g the downstroke of the spike can produce a significantphase delay (pulse A) or advance (pulse B) not captured by the quadratic <strong>in</strong>tegrateand-firemodel.Fig. 10.38, top. The spik<strong>in</strong>g trajectory <strong>in</strong> the figure quickly approaches a small shadedneighborhood of the saddle along the stable direction, and then slowly diverges fromthe saddle along the unstable direction, thereby result<strong>in</strong>g <strong>in</strong> a large period oscillation.As it is often the case <strong>in</strong> neuronal models, the saddle equilibrium is near a stable nodeequilibrium correspond<strong>in</strong>g to the rest<strong>in</strong>g state, and the system is near the co-dimension2 saddle-node homocl<strong>in</strong>ic orbit bifurcation studied <strong>in</strong> Sect. 6.3.6. As a result, thereis a drastic difference between the attraction and divergence rates to the saddle, sothat the dynamics <strong>in</strong> the shaded neighborhood of the saddle-node <strong>in</strong> the figure can bereduced to the one-dimensional V -equation, which <strong>in</strong> return can be transformed <strong>in</strong>tothe “quadratic <strong>in</strong>tegrate-and-fire” formx ′ = −1 + x 2 , if x = +∞, then x ← x reset ,with solutions depicted <strong>in</strong> Fig. 10.38, bottom. The saddle and the node correspond tox = ±1, respectively. One can check by differentiat<strong>in</strong>g that the solution of the modelwith x(0) = x reset > 1, is x(t) = − coth(t − T ), where coth(s) = (e s + e −s )/(e s − e −s )is the hyperbolic cotangent, and T = acoth (x reset ) is the period of oscillation, whichbecomes <strong>in</strong>f<strong>in</strong>ite as x reset → 1.Us<strong>in</strong>g the results of Sect. 10.4.1, we f<strong>in</strong>d the functionQ(ϑ) = 1/(−1 + coth 2 (ϑ − T )) = s<strong>in</strong>h 2 (ϑ − T )whose graph is shown <strong>in</strong> Fig. 10.39a. For comparison, we plotted the numerically foundPRC for the I Na + I K -oscillator to illustrate the disagreement between the theoreticaland numerical curves <strong>in</strong> the region ϑ < 0.1T correspond<strong>in</strong>g to the downstroke ofthe spike. Such a disagreement is somewhat expected, s<strong>in</strong>ce the quadratic <strong>in</strong>tegrateand-firemodel ignores spike downstroke. If a pulse arriv<strong>in</strong>g dur<strong>in</strong>g the downstroke

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!