Neural network modeling method for studying sleep disorder of Parkinson's disease

Abstract

The invention belongs to the technical field of Parkinson's disease sleep disorder research, and particularly relates to a neural network modeling method for researching Parkinson's disease sleep disorder. Firstly, nucleuses forming the neural network are determined, the nucleuses comprise a cortex, striatum, a globular pallidum outer side, a globular pallidum inner side, a subthalamic nucleus, a thalamus and a foot bridge nucleus, the cortex comprises rCTX and iCTX neurons, and the striatum comprises D1 middle spinous neurons and D2 middle spinous neurons; then, mathematical modeling is carried out on neurons of various nuclei, and membrane potential equations of various neurons are obtained; and finally, performing synaptic connection on neurons, adopting an Alpha synaptic model for synaptic connection between neurons rCTX and D2, between rCTX and iCTX, between iCTX and rCTX, between D1 and GPi, between D2 and GPe, between D1 and D1, between GPe and GPi, between GPe and GPe, between STN and GPi, between GPi and Th, between GPi and PPN, and between Th and rCTX, and adopting a double-index synaptic model for synaptic connection between neurons GPe and STN, between rCTX and STN, between PPN and STN, between STN and GPe, and between STN and PPN. According to the method, the foot bridge nucleus is introduced on the basis of the basal nucleus-thalamus-cortex neural network, the association between the foot bridge nucleus and the sleep disorder is established, and the blank of research on the sleep disorder of the Parkinson's disease is filled.

Description

Technical Field

[0001] The invention belongs to the technical field of Parkinson's disease sleep disorder research, and specifically is a neural network modeling method for studying Parkinson's disease sleep disorder. Background Art

[0002] Parkinson's disease is a common neurodegenerative disease in the elderly, which not only causes movement disorders, but also is accompanied by non-motor symptoms. Among them, sleep disorders are extremely common in Parkinson's patients. More than 80% of Parkinson's patients have sleep problems such as insomnia and rapid eye movement behavior disorder, which increase the physical and mental burden of patients, affect cognition and daily life ability, and hinder recovery from the disease. The neuronal activity of the human body during normal sleep follows a specific rhythm and frequency, but Parkinson's patients have an imbalance of brain neurotransmitters, which affects the neuronal activity during sleep, causing abnormalities in related neural circuits, resulting in an imbalance in the frequency of neuronal activity during sleep and causing sleep disorders.

[0003] The existing human brain neural network used to study Parkinson's disease mainly adopts the basal ganglia-thalamus-cortex neural network. When simulating the complex electrical activity of neurons in Parkinson's disease patients, the occurrence mechanism of sleep disorders in Parkinson's disease patients is not fully considered, and the complex role of neuronal electrical activity when Parkinson's disease sleep disorders occur cannot be accurately described. Therefore, this application proposes a neural network modeling method for studying Parkinson's disease sleep disorders. Summary of the invention

[0004] In view of the deficiencies in the prior art, the technical problem to be solved by the present invention is to provide a neural network modeling method for studying sleep disorders in Parkinson's disease.

[0005] The present invention solves the technical problem by adopting the following technical solution:

[0006] A neural network modeling method for studying sleep disorders in Parkinson's disease, characterized in that the method comprises the following steps:

[0007] Step 1: Identify the nuclei that make up the neural network, including the cortex, striatum, globus pallidus lateralis GPe, globus pallidus internus GPi, subthalamic nucleus STN, thalamic Th and pedunculopontine nucleus PPN. Each nucleus is composed of multiple neurons; the cortex includes rCTX and iCTX neurons, and the striatum includes D1 medium spiny neurons and D2 medium spiny neurons;

[0008] Step 2: Mathematically model the neurons of various nuclei to obtain the membrane potential equations of various neurons;

[0009] Step 3: Make synaptic connections between neurons;

[0010] The synaptic current Iij from neuron i to neuron j is defined as:

[0011] Iij = gijS(Esyn - Vi) (11)

[0012] Among them, g ij represents the synaptic coupling strength of neuron i to neuron j, S represents the synaptic exponent, E syn represents the back electromotive force, V i represents the membrane potential of neuron i;

[0013] If the synaptic connections between rCTX neurons and D2 medium spiny neurons, rCTX neurons and iCTX neurons, iCTX neurons and rCTX neurons, D1 medium spiny neurons and GPi neurons, D2 medium spiny neurons and GPe neurons, D1 medium spiny neurons and D1 medium spiny neurons, GPe neurons and GPi neurons, GPe neurons and GPe neurons, STN neurons and GPi neurons, GPi neurons and Th neurons, GPi neurons and PPN neurons, Th neurons and rCTX neurons adopt the Alpha synaptic model, then the synaptic exponent is calculated by the following formula:

[0014]

[0015] Among them, represents the maximum synaptic conductance, t d represents the synaptic delay time between neurons, and τ represents the time constant.

[0016] If the synaptic connections between GPe neurons and STN neurons, rCTX neurons and STN neurons, PPN neurons and STN neurons, STN neurons and GPe neurons, STN neurons and PPN neurons adopt the double-exponential synaptic model, then the synaptic exponent is calculated by the following formula:

[0017]

[0018] Among them, f represents the double-exponential function, τ d represents the decay time, τ r represents the rise time, t p represents the time parameter of synaptic transmission.

[0019] Furthermore, the membrane potential equation of the rCTX neuron is:

[0020]

[0021] Among them, V rCTX represents the membrane potential of the rCTX neuron, u rCTX represents the membrane potential recovery variable of the rCTX neuron, I ie represents the inhibitory current of the iCTX neuron to the rCTX neuron, I th-rCTXrepresents the excitatory current from Th neuron to rCTX neuron, a rCTX b represents the time scale of the membrane potential recovery variable of rCTX neurons, rCTX It indicates the sensitivity of the rCTX neuron membrane potential recovery variable to the membrane voltage downthreshold oscillation;

[0022] The membrane potential equation of the iCTX neuron is:

[0023]

[0024] Among them, V iCTX represents the membrane potential of iCTX neurons, u iCTX represents the membrane potential recovery variable of iCTX neurons, I ei represents the inhibitory current of rCTX neurons on iCTX neurons, I th-iCTX represents the excitatory current from Th neuron to iCTX neuron, a iCTX b represents the time scale of the recovery variable of the membrane potential of iCTX neurons, iCTX It indicates the sensitivity of the iCTX neuron membrane potential recovery variable to the membrane voltage downthreshold oscillation;

[0025] The membrane potential equations of D1 and D2 medium spiny neurons are:

[0026]

[0027] Among them, C m represents the lipid capacitance, V Str represents the membrane potential of D1 or D2 medium spiny neurons, I Na represents the sodium ion current, I K represents potassium ion current, I l Represents leakage current, I m represents the output potassium current regulated by acetylcholine, I gaba represents the inhibitory synaptic current from the same neuron, I costr represents the excitatory synaptic current from cortex to striatum;

[0028] The membrane potential equation of STN neurons is:

[0029]

[0030] Among them, V STN I represents the neuronal membrane potential of the subthalamic nucleus STN. a represents the a-type potassium current, I L represents L-type calcium current, I T represents T-type calcium current, I CaK represents the potassium current that depends on calcium ions, I gesnRepresents the inhibitory current of GPe neurons on STN neurons, I cosn Represents the excitatory current from CTX neurons, I pnsn Represents the excitatory current of PPN neurons on STN neurons;

[0031] The membrane potential equation of GPe neurons is:

[0032]

[0033] Among them, V GPe Represents the membrane potential of GPe neurons, I Ca Represents the high-threshold calcium ion current, I ahp Represents the hyperpolarizing current, I snge Represents the excitatory current of STN neurons on GPe neurons, I gege Represents the inhibitory current of GPe neurons on GPe neurons, I strge Represents the inhibitory current of Str neurons on GPe neurons, I appgpe Represents the synaptic current of other nuclei on GPe neurons;

[0034] The membrane potential equation of GPi neurons is:

[0035]

[0036] Among them, V GPi Represents the membrane potential of GPi neurons, I sngi Represents the excitatory current of STN neurons on GPi neurons, I gegi Represents the inhibitory current of GPe neurons on GPi neurons, I strgi Represents the inhibitory current of Str neurons on GPe neurons; I appgpi Represents the synaptic current of other nuclei on GPi neurons;

[0037] The membrane potential equation of Th neurons is:

[0038]

[0039] Among them, V Th Represents the membrane potential of Th neurons, I gith Represents the inhibitory current of GPi neurons on Th neurons, I appth Represents the synaptic current of other nuclei on Th neurons;

[0040] The membrane potential equation of PPN neurons is:

[0041]

[0042] Among them, VPPN represents the membrane potential of the PPN neuron, I hyp represents the activation current, I Na,p represents the persistent sodium current, I K,p represents the persistent potassium current, I gipn represents the inhibitory current of the GPi neuron on the PPN neuron, I snpn represents the excitatory current of the STN neuron on the PPN neuron.

[0043] Furthermore, when V rCTX ≥ 30 mV, the membrane potential and the membrane potential recovery variable of the rCTX neuron are reset according to the following formula;

[0044]

[0045] In the formula, c rCTX represents the rCTX neuron membrane voltage after being reset exceeding the threshold, d rCTX represents the rCTX neuron membrane potential recovery variable after reset;

[0046] When V iCTX ≥ 30 mV, the membrane potential and the membrane potential recovery variable of the iCTX neuron are reset according to the following formula;

[0047]

[0048] In the formula, c iCTX represents the iCTX neuron membrane voltage after being reset exceeding the threshold, d iCTX represents the iCTX neuron membrane potential recovery variable after reset.

[0049] Furthermore, the synaptic coupling strength of the GPi neuron on the PNN neuron is 0.003, the synaptic coupling strength of the STN neuron on the PPN neuron is 0.08, and the synaptic coupling strength of the PPN neuron on the STN neuron is 0.3.

[0050] Compared with the prior art, the beneficial effects of the present invention are:

[0051] 1. The present invention introduces the pedunculopontine nucleus (PPN) on the basis of the basal ganglia-thalamus-cortex neural network, and innovatively constructs a four-dimensional neural network model of the basal ganglia-thalamus-cortex-pedunculopontine nucleus including the pedunculopontine nucleus, breaking through the limitation of the three-node model of the basal ganglia-thalamus-cortex in traditional Parkinson's disease research. Under the pathological state of Parkinson's disease, the power oscillation of the pedunculopontine nucleus neural cluster in the low-frequency band decreases compared with the healthy state, while the power oscillation in the high-frequency band increases. This frequency-domain characteristic reflects that the brain is in an abnormally excited state. Dynamic analysis shows the pathological mechanism of reduced slow-wave sleep and difficulty in maintaining sleep in Parkinson's disease patients. For the first time, a quantitative association between the oscillation characteristics of the pedunculopontine nucleus and sleep disorders is established at the level of computational neuroscience, filling the gap in multi-scale modeling technology for the mechanism of sleep disorders in Parkinson's disease, providing new ideas and methods for constructing a high-precision neural network model, and having important academic and technical value.

[0052] 2. The synaptic connection between neurons has an important information transmission function. The Alpha synaptic model is often used to simulate the activity of neuron populations that require rapid response due to its simplicity and high efficiency in quickly processing and transmitting signals, which can significantly improve the computational efficiency. The double-exponential synaptic model takes into account the complex dynamic characteristics of ion channels and is widely used in constructing high-precision and complex neural networks. It can make the constructed neural network closer to the behavior of real biological neural networks and plays a key role in studying the complex cognitive functions of the brain. By adopting a hybrid synaptic modeling method, the Alpha synaptic model and the double-exponential model are organically combined, which not only ensures the operation efficiency but also significantly improves the biological simulation degree of the oscillation characteristics of neuron populations, and can accurately simulate the complex behavior of the neural network of sleep disorders in Parkinson's disease. Brief Description of the Drawings

[0053] Figure 1 is the structural diagram of the basal ganglia-thalamus-cortex-pedunculopontine nucleus neural network of the present invention;

[0054] Figure 2 is the average field potential power spectral density map of the pedunculopontine nucleus in healthy and Parkinson's disease states;

[0055] Figure 3 is the spatio-temporal discharge power spectral density map of the pedunculopontine nucleus in healthy and Parkinson's disease states;

[0056] Figure 4 is the time-frequency diagram of the pedunculopontine nucleus in the healthy state;

[0057] Figure 5 is the time-frequency diagram of the pedunculopontine nucleus in the Parkinson's disease state. Detailed Embodiments

[0058] The following provides specific embodiments in conjunction with the accompanying drawings. The specific embodiments are only used to introduce the technical solutions of the present invention in detail, and do not limit the protection scope of this application.

[0059] The present invention provides a neural network modeling method for studying sleep disorders in Parkinson's disease (hereinafter referred to as the method, see Figures 1 to 5 ), which includes the following steps:

[0060] Step 1: Determine the nuclei that make up the basal ganglia-thalamus-cortex-pedunculopontine nucleus neural network;

[0061] The neurophysiological characteristics of sleep disorders in Parkinson's disease can be summarized as abnormal discharges and pathological synchronization behaviors of the basal ganglia-thalamus-cortex-pedunculopontine nucleus neural network. As Figure 1 shown, the nuclei include the cortex (Cortex, CTX), striatum (Striatum, Str), external globus pallidus (Globus Pallidus externa, GPe), internal globus pallidus (Globus Pallidus interna, GPi), subthalamic nucleus (Subthalamic Nucleus, STN), thalamus (Thalamus, Th), and pedunculopontine nucleus (Pedunculopontine Nucleus, PPN); among them, the cortex includes excitatory regular spiking neurons (Regular Spiking Cortex, rCTX) and inhibitory fast-spiking interneurons (Fast-Spiking Inhibitory Interneurons Cortex, iCTX), and the striatum (STr) includes D1 medium spiny neurons and D2 medium spiny neurons, and each type of nucleus consists of 10 neurons.

[0062] Step 2: Perform mathematical modeling on the neurons of various nuclei to obtain the membrane potential equations of various types of neurons;

[0063] In CTX, the membrane potential equation of rCTX neurons is:

[0064]

[0065] When V rCTX ≥ 30 mV, the membrane potential and membrane potential recovery variable of rCTX neurons are reset according to the following formula;

[0066]

[0067] Among them, V rCTX represents the membrane potential of rCTX neurons, u rCTX represents the membrane potential recovery variable of rCTX neurons, and I ieRepresents the inhibitory current of iCTX neurons on rCTX neurons, I th-rCTX Represents the excitatory current from the thalamus Th to rCTX neurons, a rCTX Represents the time scale of the membrane potential recovery variable of rCTX neurons, b rCTX Represents the sensitivity of the membrane potential recovery variable of rCTX neurons to subthreshold oscillations of the membrane voltage, c rCTX Represents the membrane voltage of rCTX neurons reset after exceeding the threshold, d rCTX Represents the membrane potential recovery variable of rCTX neurons after reset.

[0068] The membrane potential equation of iCTX neurons is:

[0069]

[0070] When V iCTX ≥ 30 mV, the membrane potential and membrane potential recovery variable of iCTX neurons are reset according to the following formula;

[0071]

[0072] Among them, V iCTX Represents the membrane potential of iCTX neurons, u iCTX Represents the membrane potential recovery variable of iCTX neurons, I ei Represents the inhibitory current of rCTX neurons on iCTX neurons, I th-iCTX Represents the excitatory current from the thalamus Th to iCTX neurons, a iCTX Represents the time scale of the membrane potential recovery variable of iCTX neurons, b iCTX Represents the sensitivity of the membrane potential recovery variable of iCTX neurons to subthreshold oscillations of the membrane voltage, c iCTX Represents the membrane voltage of iCTX neurons reset after exceeding the threshold, d iCTX Represents the membrane potential recovery variable of iCTX neurons after reset.

[0073] In the striatum Str, the membrane potential equations of medium spiny neurons in D1 and D2 are:

[0074]

[0075] Among them, C m Represents the lipid capacitance, V Str Represents the membrane potential of medium spiny neurons in D1 or D2 in the striatum Str, I Na Represents the sodium ion current, I K Represents the potassium ion current, I l Represents the leak current, I m Represents the output potassium current regulated by acetylcholine, Igaba represents the inhibitory synaptic current from neurons of the same type, I costr represents the excitatory synaptic current from the cortex CTX to the striatum Str.

[0076] The membrane potential equation of STN neurons is:

[0077]

[0078] where, V STN represents the membrane potential of neurons in the subthalamic nucleus STN, I a represents the a-type potassium current, I L represents the L-type calcium current, I T represents the T-type calcium current, I CaK represents the potassium current that varies depending on calcium ions, I gesn represents the inhibitory current of GPe neurons on STN neurons, I cosn represents the excitatory current from CTX neurons, I pnsn represents the excitatory current of PPN neurons on STN neurons.

[0079] The membrane potential equation of GPe neurons is:

[0080]

[0081] where, V GPe represents the membrane potential of GPe neurons, I Ca represents the high-threshold calcium current, I ahp represents the hyperpolarizing current, I snge represents the excitatory current of STN neurons on GPe neurons, I gege represents the inhibitory current of GPe neurons on GPe neurons, I strge represents the inhibitory current of Str neurons on GPe neurons, I appgpe represents the synaptic current of other nuclei on GPe neurons.

[0082] The membrane potential equation of GPi neurons is:

[0083]

[0084] where, V GPi represents the membrane potential of GPi neurons, I sngi represents the excitatory current of STN neurons on GPi neurons, I gegi represents the inhibitory current of GPe neurons on GPi neurons, I strgi represents the inhibitory current of Str neurons on GPe neurons; I appgpiRepresents the synaptic current of other nuclei on GPi neurons.

[0085] The membrane potential equation of Th neurons is:

[0086]

[0087] Among them, V Th Represents the membrane potential of Th neurons, and I gith Represents the inhibitory current of GPi neurons on Th neurons, and I appth Represents the synaptic current of other nuclei on Th neurons.

[0088] The membrane potential equation of PPN neurons is:

[0089]

[0090] Among them, V PPN Represents the membrane potential of PPN neurons, and I hyp Represents the activation current, and I Na,p Represents the persistent sodium current, and I K,p Represents the persistent potassium current, and I gipn Represents the inhibitory current of GPi neurons on PPN neurons, and I snpn Represents the excitatory current of STN neurons on PPN neurons.

[0091] Step 3: Perform synaptic connections on the neurons;

[0092] In the present invention, chemical synapse and electrical synapse models are selected to connect the neurons. Then, the synaptic current Iij of neuron i on neuron j is defined as:

[0093] Iij = gijS(Esyn - Vi) (11)

[0094] Among them, g ij Represents the synaptic coupling strength of neuron i on neuron j, S represents the synaptic index that satisfies the kinetic equation, and E syn Represents the back electromotive force, and V i Represents the membrane potential of neuron i.

[0095] The Alpha synapse model is suitable for describing synaptic responses with relatively simple temporal dynamics, mainly showing unimodal changes, which is conducive to rapid modeling in large-scale neural networks and obtaining the general dynamic characteristics of neural networks. Therefore, the synaptic connections between rCTX neurons and D2 medium spiny neurons, rCTX neurons and iCTX neurons, iCTX neurons and rCTX neurons, D1 medium spiny neurons and GPi neurons, D2 medium spiny neurons and GPe neurons, D1 medium spiny neurons and D1 medium spiny neurons, GPe neurons and GPi neurons, GPe neurons and GPe neurons, STN neurons and GPi neurons, GPi neurons and Th neurons, GPi neurons and PPN neurons, and Th neurons and rCTX neurons all adopt the Alpha synapse model, and the synaptic index S that satisfies the kinetic equation is expressed as:

[0096]

[0097] Among them, represents the maximum synaptic conductance, t d represents the synaptic delay time between neurons, and τ represents the time constant.

[0098] The double-exponential synapse model is suitable for more precise and complex temporal dynamic synaptic responses. When there are multiple neurotransmitter receptors or synapses with complex modulation mechanisms, the double-exponential synapse model can more accurately describe their dynamic behaviors during information transmission. Therefore, the synaptic connections between GPe neurons and STN neurons, rCTX neurons and STN neurons, PPN neurons and STN neurons, STN neurons and GPe neurons, and STN neurons and PPN neurons adopt the double-exponential synapse model, and the synaptic index S that satisfies the kinetic equation is expressed as:

[0099]

[0100] Among them, f represents the double-exponential function, t p represents the time parameter of synaptic transmission, τ r represents the rise time, τ d represents the decay time.

[0101] To better reflect the neural network of Parkinson's disease sleep disorders, the values of some parameters of the neural network are shown in Table 1;

[0102] Table 1 Values of Some Parameters of the Neural Network

[0103]

[0104]

[0105] Among them, g gpippnRepresents the synaptic coupling strength of GPi neurons to PNN neurons, g stnppn Represents the synaptic coupling strength of STN neurons to PPN neurons, g ppnstn Represents the synaptic coupling strength of PPN neurons to STN neurons, t dCTX-Str Represents the synaptic delay time between CTX neurons and Str neurons, t dCTX-STN Represents the synaptic delay time between CTX neurons and STN neurons, t dStr-GPi Represents the synaptic delay time between Str neurons and GPi neurons, t dStr-GPe Represents the synaptic delay time between Str neurons and GPe neurons, t dSTN-GPi Represents the synaptic delay time between STN neurons and GPi neurons, t dSTN-GPe Represents the synaptic delay time between STN neurons and GPe neurons, t dGPe-STN Represents the synaptic delay time between GPe neurons and STN neurons, t dGPi-GPi Represents the synaptic delay time between GPi neurons and GPi neurons, t dGPi-Th Represents the synaptic delay time between GPi neurons and Th neurons, t dTh-CTX Represents the synaptic delay time between Th neurons and CTX neurons.

[0106] According to the complex connection mode of neurons in the real brain, a random connection method is adopted in this model to simulate the electrical activities between neurons. Parkinson's disease is caused by the reduction of striatal dopamine neurons. By changing the synaptic coupling strength in the neural network to meet the simulation requirements of Parkinson's disease, including the synaptic coupling strength g costr of CTX neurons to Str neurons, the synaptic coupling strength g m of D1 medium spiny neurons in the striatum to D2 medium spiny neurons, and the synaptic coupling strength g gege of GPe neurons to GPe neurons. The values of these parameters in the healthy state and the Parkinson's disease state are shown in Table 2.

[0107] Table 2 Values of some parameters in the healthy state and the Parkinson's disease state

[0108]

[0109] To verify the effectiveness of this modeling method, the constructed neural network is simulated, and the simulation results of the pedunculopontine nucleus in the healthy and Parkinson's disease states are output, including the mean field potential power spectral density map and the spatio-temporal discharge power spectral density map of the pedunculopontine nucleus, as shown in Figure 2 and 3 ; Figure 4 and 5Time-frequency diagrams of the pedunculopontine nucleus in the healthy and Parkinson's disease states, respectively. It can be seen from the simulation results that, compared with the healthy state, the power oscillation of the pedunculopontine nucleus in the Parkinson's disease state weakens in the low-frequency band (0-10 Hz) and strengthens in the high-frequency band (13-35 Hz), indicating that the brain is in an abnormally excited state, which is consistent with the physiological phenomenon that patients with Parkinson's disease have difficulty falling asleep and maintaining deep sleep, verifying the effectiveness of the neural network constructed by the present invention in studying sleep disorders in Parkinson's disease.

[0110] Matters not described in the present invention are applicable to the prior art.

Claims

1. A neural network modeling method for studying sleep disorders in Parkinson's disease, characterized in that: The method comprises the following steps: Step 1: Identify the nuclei that make up the neural network, including the cortex, striatum, globus pallidus lateralis GPe, globus pallidus internus GPi, subthalamic nucleus STN, thalamic Th and pedunculopontine nucleus PPN. Each nucleus is composed of multiple neurons; the cortex includes rCTX and iCTX neurons, and the striatum includes D1 medium spiny neurons and D2 medium spiny neurons; Step 2: Mathematically model the neurons of various nuclei to obtain the membrane potential equations of various neurons; Step 3: Make synaptic connections between neurons; The synaptic current Iij from neuron i to neuron j is defined as: Iij=gijS(Esyn-Vi) (11) Among them, g ij represents the synaptic coupling strength of neuron i to neuron j, S represents the synaptic index, and E syn Represents the back electromotive force, V i represents the membrane potential of neuron i; The synaptic connections between rCTX neurons and D2 medium spiny neurons, rCTX neurons and iCTX neurons, iCTX neurons and rCTX neurons, D1 medium spiny neurons and GPi neurons, D2 medium spiny neurons and GPe neurons, D1 medium spiny neurons and D1 medium spiny neurons, GPe neurons and GPi neurons, GPe neurons and GPe neurons, STN neurons and GPi neurons, GPi neurons and Th neurons, GPi neurons and PPN neurons, and Th neurons and rCTX neurons adopt the Alpha synaptic model, and the synaptic index is calculated by the following formula: in, represents the maximum synaptic conductance, t d represents the synaptic delay time between neurons, and τ represents the time constant. The synaptic connections between GPe neurons and STN neurons, rCTX neurons and STN neurons, PPN neurons and STN neurons, STN neurons and GPe neurons, and STN neurons and PPN neurons adopt a double exponential synaptic model, and the synaptic index is calculated by the following formula: Where f represents the double exponential function, τ d represents the decay time, τ r represents the rise time, t p Represents the timing parameter of synaptic transmission.

2. The neural network modeling method for studying sleep disorders in Parkinson's disease according to claim 1, characterized in that: The membrane potential equation of the rCTX neuron is: Among them, V rCTX represents the membrane potential of rCTX neurons, u rCTX represents the membrane potential recovery variable of rCTX neurons, I ie represents the inhibitory current of iCTX neurons on rCTX neurons, I th-rCTX represents the excitatory current from Th neuron to rCTX neuron, a rCTX b represents the time scale of the membrane potential recovery variable of rCTX neurons, rCTX It indicates the sensitivity of the rCTX neuron membrane potential recovery variable to the membrane voltage downthreshold oscillation; The membrane potential equation of the iCTX neuron is: Among them, V iCTX represents the membrane potential of iCTX neurons, u iCTX represents the membrane potential recovery variable of iCTX neurons, I ei represents the inhibitory current of rCTX neurons on iCTX neurons, I th-iCTX represents the excitatory current from Th neuron to iCTX neuron, a iCTX b represents the time scale of the recovery variable of the membrane potential of iCTX neurons, iCTX It indicates the sensitivity of the iCTX neuron membrane potential recovery variable to the membrane voltage downthreshold oscillation; The membrane potential equations of D1 and D2 medium spiny neurons are: Among them, C m represents the lipid capacitance, V Str represents the membrane potential of D1 or D2 medium spiny neurons, I Na represents the sodium ion current, I K represents potassium ion current, I l Represents leakage current, I m represents the output potassium current regulated by acetylcholine, I gaba represents the inhibitory synaptic current from the same neuron, I costr represents the excitatory synaptic current from cortex to striatum; The membrane potential equation of STN neurons is: Among them, V STN I represents the neuronal membrane potential of the subthalamic nucleus STN. a represents the a-type potassium current, I L represents L-type calcium current, I T represents T-type calcium current, I CaK represents the potassium current that depends on calcium ions, I gesn represents the inhibitory current of GPe neurons on STN neurons, I cosn represents the excitatory current from CTX neurons, I pnsn represents the excitatory current of PPN neurons to STN neurons; The membrane potential equation of the GPe neuron is: Among them, V GPe represents the membrane potential of GPe neuron, I Ca Represents the high threshold calcium ion current, I ahp represents the hyperpolarization current, I snge represents the excitatory current of STN neurons to GPe neurons, I gege represents the inhibitory current of GPe neuron to GPe neuron, I strge represents the inhibitory current of Str neurons on GPe neurons, I appgpe represents the synaptic currents of other nuclei to GPe neurons; The membrane potential equation of the GPi neuron is: Among them, V GPi represents the membrane potential of GPi neurons, I sngi represents the excitatory current of STN neurons to GPi neurons, I gegi represents the inhibitory current of GPe neurons on GPi neurons, I strgi represents the inhibitory current of Str neurons on GPe neurons; I appgpi represents the synaptic currents of other nuclei to GPi neurons; The membrane potential equation of Th neuron is: Among them, V Th represents the membrane potential of Th neurons, I gith represents the inhibitory current of GPi neurons on Th neurons, I appth represents the synaptic currents of other nuclei to Th neurons; The membrane potential equation of the PPN neuron is: Among them, V PPN represents the membrane potential of PPN neurons, I hyp Represents the activation current, I Na,p represents the continuous sodium current, I K,p represents the continuous potassium current, I gipn represents the inhibitory current of GPi neurons on PPN neurons, I snpn Represents the excitatory current of STN neurons to PPN neurons.

3. The neural network modeling method for studying sleep disorders in Parkinson's disease according to claim 1 or 2, characterized in that: When V rCTX When ≥30 mV, the membrane potential and membrane potential recovery variables of the rCTX neuron were reset according to the following formula; In the formula, c rCTX represents the membrane voltage of rCTX neurons after exceeding the threshold and being reset, d rCTX represents the rCTX neuron membrane potential recovery variable after reset; When V iCTX When ≥30 mV, the membrane potential and membrane potential recovery variables of the iCTX neuron were reset according to the following formula; In the formula, c iCTX represents the iCTX neuron membrane voltage after exceeding the threshold and being reset, d iCTX Represents the variable of membrane potential recovery of iCTX neurons after reset.

4. The neural network modeling method for studying sleep disorders in Parkinson's disease according to claim 1, characterized in that: The synaptic coupling strength of GPi neurons to PNN neurons is 0.003, the synaptic coupling strength of STN neurons to PPN neurons is 0.08, and the synaptic coupling strength of PPN neurons to STN neurons is 0.3.

Images