Modelling Effects of Electrical Stimulation on Seizure BENG/BGGN - PowerPoint PPT Presentation

Modelling Effects of Electrical Stimulation on Seizure BENG/BGGN 260 final project By: Carissa Gunawan

Introduction • Seizure is caused by abnormal excessive synchronous neural activity in the brain • Therapy through drugs, VNS, and DBS • Advantage: • Reversible • More localized area Fisher, Robert S., et al. "Epileptic seizures and epilepsy: definitions proposed by the International League Against Epilepsy (ILAE) and the International Bureau for Epilepsy (IBE)." Epilepsia 46.4 (2005): 470-472. Fisher, Robert S., and Ana Luisa Velasco. "Electrical brain stimulation for epilepsy." Nature Reviews Neurology 10.5 (2014): 261-270.

Purpose • The goal of this project is to find out the effects of electrical stimulation on various neural network during seizure • Expected result: • Stop over excitation during seizure • alter activity of normal network DeGiorgio, Christopher M., and Scott E. Krahl. "Neurostimulation for drug-resistant epilepsy." Continuum: Lifelong Learning in Neurology 19.3 Epilepsy (2013): 743.

Stimulation model (a) 50 40 30 20 10 • Extracellular medium assumptions: I e (t) (mA) 0 − 10 • Homogeneous electrical property − 20 • Homogeneous density − 30 − 40 • Point source − 50 0 2 4 6 8 10 12 14 16 18 20 t (ms) (b) • Biphasic pulse sequence − 1 4.5 − 0.8 • f=120-180Hz, pw= 0.06-0.2ms, V=1-5V 4 − 0.6 • Relationship of current with distance 3.5 − 0.4 3 − 0.2 • 𝐽 𝑢, 𝑆 = ' ( mm 2.5 0 )*+, 0.2 2 R=distance and ρ=resistance/mm 0.4 1.5 0.6 1 0.8 0.5 1 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 mm Monfared, Omid, et al. "Electrical stimulation of neural tissue modeled as a cellular composite: Point Source electrode in an isotropic tissue." Engineering in Medicine and Biology Society (EMBC), 2014 36th Annual International Conference of the IEEE . IEEE, 2014.

Neuron model • Model: Hodgkin and Huxley • Sodium channel, potassium channel, chlorine channel, Na-K pump, leak currents, glia cells, inhibitory synapse, and excitatory synapse • Assumptions: • m is fast as compared to the voltage change • Total amount of sodium ion is conserved • Spherical cell body Cressman, John R., et al. "The influence of sodium and potassium dynamics on excitability, seizures, and the stability of persistent states: I. Single neuron dynamics." Journal of computational neuroscience 26.2 (2009): 159-170.

Neuron model: Ionic Currents Leak Current: 𝐽 -. = 𝑕 -. 𝑛 1 2 ℎ 𝑊 − 𝐹 -. + 𝑕 -.8 𝑊 − 𝐹 -. 𝐽 9 = 𝑕 9 𝑜 ) 𝑊 − 𝐹 9 + 𝑕 98 𝑊 − 𝐹 9 𝐽 ;< = 𝑕 ;<8 𝑊 − 𝐹 ;< Current by Na-K pump, glia cells, and diffusion: 𝜍 1 𝐽 =>?= = × 1 + exp 25 − 𝑂𝑏 I 1 + exp 5.5 − 𝐿 N 3 𝐻 O<I. 𝐽 O<I. = 1 + exp 18 − 𝐿 N 2.5 𝐽 RISS = 𝜗 𝐿 N − 𝑙 1 Barreto, Ernest, and John R. Cressman. "Ion concentration dynamics as a mechanism for neuronal bursting." Journal of biological physics 37.3 (2011): 361-373.

Neuron model: Change in [K] o and [Na] i Change in external potassium and internal sodium concentration 𝜐 𝑒 𝐿 N = 𝛿𝛾𝐽 9 − 2𝛾𝐽 =>?= − 𝐽 O<I. − 𝐽 RISS>ZIN[ 𝑒𝑢 𝜐 𝑒 𝑂𝑏 I = −𝛿𝐽 -. − 3𝐽 =>?= 𝑒𝑢 External Sodium and Internal Potassium concentration: 𝑂𝑏 N = 144 − 𝛾 𝑂𝑏 I − 18 𝐿 I = 140 + (18 − 𝑂𝑏 I ) Barreto, Ernest, and John R. Cressman. "Ion concentration dynamics as a mechanism for neuronal bursting." Journal of biological physics 37.3 (2011): 361-373.

Bifurcation factor • Bifurcation occurs depending on the dynamics of potassium and sodium • This could be because of: • Increase in local K concentration due to drugs or physical injury • Genetic disease that alter the capacity of glia cell 100 50 50 V m (mV) 0 V m (mV) 0 − 50 − 50 − 100 − 100 0 5 10 15 20 25 30 0 5 10 15 20 25 30 10 40 8 [K] o (mM) [K] o (mM) 30 6 20 4 2 10 0 5 10 15 20 25 30 0 5 10 15 20 25 30 19 34 32 [Na] i (mM) [Na] i (mM) 18 30 17 28 16 26 0 5 10 15 20 25 30 0 5 10 15 20 25 30 time (s) time (s)

Result: Bifurcation of one neuron 100 without stimulation with stimulation 50 60 V m (mV) 0 − 50 40 − 100 0 5 10 15 20 25 30 20 10 8 [K] o (mM) 0 V m (mV) 6 4 − 20 2 0 5 10 15 20 25 30 − 40 19 [Na] i (mM) 18 − 60 17 − 80 16 22.9 23 23.1 23.2 23.3 23.4 23.5 23.6 0 5 10 15 20 25 30 time(s) time (s) • Bifurcation start when [K] o in normal condition increase • Surrounding neurons in this environment also conduct action potential at the same time, making synchronous excitation • Stimulation reduce synchrony by changing its frequency

Network formation • Independent neural network • Dependent neural network both inhibitory and excitatory with respect to each other Beverlin II, Bryce, et al. "Dynamical changes in neurons during seizures determine tonic to clonic shift." Journal of computational neuroscience 33.1 (2012): 41-51. • Random neural network

Discussion: Effects on networks (a) • It takes time for electrical 2000 stimulation to unsynchronized the 1000 V sum (mV) 0 excitation − 1000 • It is not guaranteed and all the − 2000 0 5 10 15 20 25 30 t (s) neuron will be not synchronize (b) 2000 with each other 1000 V sum (mV) 0 • Excitation pattern is less − 1000 synchronized when neurons in a − 2000 0 5 10 15 20 25 30 network depends on each other t (s) (c) 2000 • In a totally random network, it is 1000 V sum (mV) hard to find condition that leads to 0 seizure − 1000 − 2000 0 5 10 15 20 25 30 t (s) (a) Independent network (b) dependent network (c) random network

EEG model • Assuming all the affected neurons are the only target • Attenuation and noise • Amplitude of 10-100 µV • Noise consist of noise from other neuron • Random white noise • Sampled at a sampling frequency 200-2000 Hz Noise from other neurons Amplification raw_eeg EEG Attenuation and + Sampling Aurlien, H., et al. "EEG background activity described by a large computerized database." Clinical Neurophysiology 115.3 (2004): 665-673. Hämäläinen, Matti, et al. "Magnetoencephalography—theory, instrumentation, and applications to noninvasive studies of the working human brain." Reviews of modern Physics 65.2 (1993): 413.

Result: EEG model (a) (a) 2000 30 1000 20 V sum (mV) V sum ( µ V) 0 10 − 1000 0 − 2000 − 10 0 5 10 15 20 25 30 0 5 10 15 20 25 30 t (s) (b) (b) 40 2000 1000 V sum (mV) V sum ( µ V) 20 0 0 − 1000 − 2000 − 20 0 5 10 15 20 25 30 0 5 10 15 20 25 30 t (s) (c) (c) 30 2000 20 1000 V sum (mV) V sum ( µ V) 10 0 0 − 1000 − 10 − 2000 0 5 10 15 20 25 30 0 5 10 15 20 25 30 time(s) t (s) Raw EEG Constructed EEG (a) Independent network (b) dependent network (c) random network

Discussion: EEG comparison (a) 30 20 V sum ( µ V) 10 0 − 10 0 5 10 15 20 25 30 (b) 40 V sum ( µ V) 20 0 − 20 0 5 10 15 20 25 30 (c) 30 20 V sum ( µ V) 10 0 − 10 0 5 10 15 20 25 30 time(s)

Conclusion • The simulated EEG model have a similar behavior as the EEG recording in the study. • Electrical stimulation reduce seizure • Part where seizure did not happen have some changes when electrical stimulation was applied • Change in local concentration will most likely cause seizure in area where neuron are less dependent on each other • Synchrony can be reduced if they are dependent on each other

Future direction • Test with more realistic neuron model • The one with Ca2+ activated K+ channel • Test with a more realistic neural network • Variable inhibition and excitation strength • Test with other conditions that cause seizure • Example abnormal Ca2+ concentration • Glia’s failure to regulate K+ ions

Thank You