Closed-loop control electrical stimulation system for targeted facial nerve para-nucleus induced slow-wave sleep
By designing a closed-loop controlled electrical stimulation system targeting the paraneural nucleus of the facial nerve, using convolutional neural networks and fast Fourier transform to process EEG signals, the problem of difficulty in inducing and maintaining slow wave sleep in the existing technology is solved, and the rapid and Long-lasting sleep-induced effects.
Patent Information
- Application Number
- CN202411314098.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-20
- Publication Date
- 2025-01-14
AI Technical Summary
The prior art is difficult to directly induce and maintain slow wave sleep. Optogenetic regulation leads to protein inactivation, chemical genetic regulation takes effect slowly, and slow wave sleep cannot be immediately induced.
Design a closed-loop controlled electrical stimulation system targeting the paraneural nucleus of the facial nerve, including an electrical stimulation module and a sleep state recognition module, and use convolutional neural network and fast Fourier transform to process EEG signals, judge the current sleep stage, and use electrical stimulation The module outputs electrical stimulation to the paraface nerve nucleus to induce and maintain slow wave sleep.
A fast-induced and long-lasting slow-wave sleep effect is achieved, and the electrical stimulation is adjusted through closed-loop evaluation to ensure that the sleep state reaches an ideal state.
Smart Images
Figure CN119303236A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of medical devices, and in particular relates to a closed-loop controlled electrical stimulation system for inducing slow-wave sleep by targeting the parafacial nucleus. Background Art
[0002] Maintaining a certain length of slow-wave sleep at night is a necessary prerequisite for achieving optimal physiological, psychological, and cognitive functions. Studies on animals and humans have shown that slow-wave sleep can be promoted by activating the γ-aminobutyric acid neurons in the parafacial nucleus of the mammalian pons. The activated synapses of the parafacial nucleus neurons release γ-aminobutyric acid to the parabrachial nucleus neurons, which in turn project and release glutamate to the cortical projection neurons of the basal forebrain; activating the γ-aminobutyric acid neurons in the parafacial nucleus can effectively trigger slow-wave sleep and regulate cortical EEG activity. However, there is currently no technology to directly induce and maintain slow-wave sleep, making clinical translation of this target difficult.
[0003] In the prior art, optogenetic regulation is used to induce slow-wave sleep: an optical fiber is implanted unilaterally above the parafacial nucleus of VGAT-ChR2-EYFP mice (implantation coordinates: AP = -5.5mm; ML = 1.4mm; DV = -4.2mm) through stereotaxic surgery, and a stimulator (PG 4000A, Cygnus Technology, Inc., Delaware Water Gap, PA) is used to stimulate using a light pulse sequence (wavelength 473nm, frequency 40Hz). The disadvantage of optogenetic regulation is that long-term exposure to light-sensitive channel proteins will cause protein inactivation and cannot maintain a slow-wave sleep state for a long time.
[0004] Chemical genetics can also be used to induce slow wave sleep: hSyn-DIO-hM4D-mCherry-AAV virus is injected through stereotaxic surgery to selectively express hM4Di receptors in the γ-aminobutyric acid neurons of the parafacial nucleus of Vgat-IRES-cre mice (injection coordinates: AP = -5.3mm, ML = ±0.7mm, DV = -3.4mm). This receptor can bind to the drug clozapine-N-oxide to activate neurons expressing the receptor. Clozapine-N-oxide can be injected intraperitoneally into mice to specifically activate the γ-aminobutyric acid neurons of the parafacial nucleus, thereby inducing slow wave sleep. The disadvantage of chemical genetics is that it takes more than half an hour for clozapine-N-oxide to take effect and cannot immediately induce slow wave sleep. Summary of the invention
[0005] The technical purpose of the present invention is to provide a closed-loop controlled electrical stimulation system targeting the parafacial nucleus to induce slow-wave sleep, so as to solve the problem of slow-wave sleep.
[0006] To solve the above problems, the technical solution of the present invention is:
[0007] A closed-loop controlled electrical stimulation system for inducing slow-wave sleep by targeting the parafacial nucleus, comprising:
[0008] An electrical stimulation module, configured to electrically stimulate the parafacial nucleus of the target to be detected;
[0009] The sleep state recognition module is configured to collect the EEG signals of the whole brain of the target to be detected, and perform signal processing on the EEG signals to obtain the current sleep stage of the target to be detected;
[0010] Among them, the sleep state recognition module adopts a universal time series network, which includes a convolutional neural network unit, a fast Fourier transform unit, a circadian time unit and a result output unit;
[0011] The convolutional neural network unit is configured to receive the EEG signal for convolution calculation to obtain the convolution calculation result; the fast Fourier transform unit is configured to receive the EEG signal for fast Fourier calculation to obtain the fast Fourier transform result; the circadian time unit is configured to output circadian rhythm time information; the result output unit is configured to receive the convolution calculation result, the fast Fourier transform result and the circadian rhythm time information and merge them to obtain a three-dimensional vector, and judge the current sleep stage according to the three-dimensional vector.
[0012] Specifically, the electrical stimulation module includes an electrical stimulation generation unit and electrodes;
[0013] The electrical stimulation generating unit is electrically connected to the electrode and is configured to output electrical stimulation to the electrode, each electrical stimulation comprising 1000 pulses, an intensity of 20uA, and a pulse width of 5ms;
[0014] The electrode is made of two strands of nickel-titanium alloy twisted into a conductive wire with a length of 4 mm. The surface of the electrode is wrapped with an insulating layer, while the tail discharge end of the electrode is not wrapped with an insulating layer.
[0015] Among them, the convolutional neural network unit includes eight convolutional layers, which are:
[0016] In the first convolution layer, 64 convolution kernels of length 9 are used to perform convolution operation on the EEG signal with a step size of 1;
[0017] The second convolutional layer uses 64 convolution kernels of length 9 and performs convolution operation on the data output by the first convolutional layer with a step size of 2;
[0018] The third convolutional layer uses 64 convolution kernels of length 9 and performs convolution operation on the data output by the second convolutional layer with a stride of 2;
[0019] The fourth convolutional layer uses 64 convolution kernels of length 9 and performs convolution operation on the data output by the third convolutional layer with a stride of 2;
[0020] The fifth convolutional layer uses 64 convolution kernels of length 7 and performs convolution operation on the data output by the fourth convolutional layer with a step size of 1;
[0021] The sixth convolutional layer uses 64 convolution kernels of length 7 and performs convolution operation on the data output by the fifth convolutional layer with a stride of 2;
[0022] The seventh convolutional layer uses 64 convolution kernels of length 7 and performs convolution operation on the data output by the sixth convolutional layer with a stride of 2;
[0023] The eighth convolutional layer uses 64 convolution kernels of length 7 and performs convolution operation on the data output by the seventh convolutional layer with a stride of 2.
[0024] Further preferably, a rectified linear unit activation layer and a batch normalization layer are provided behind each convolution layer, the rectified linear unit activation layer is configured to realize nonlinear conversion signal, and the batch normalization layer is configured to prevent gradient disappearance.
[0025] Further preferably, a first jump connection and a second jump connection are further provided;
[0026] The first skip connection is located between the output of the first convolutional layer and the input of the fifth convolutional layer to capture coarse features;
[0027] The second skip connection is located between the output of the fifth convolutional layer and the input of the resulting output unit to capture detailed features.
[0028] The fast Fourier transform spectrum in the fast Fourier transform unit ranges from 1 to 12 Hz, and the spectrum is divided into 22 intervals, each of which is 0.5 Hz.
[0029] Specifically, the result output unit includes a fully connected network layer, a Softmax function, and an Argmax function;
[0030] The fully connected network layer is configured to receive the convolution calculation results, fast Fourier transform results and circadian rhythm time information and merge them for output;
[0031] The Softmax function is configured to receive the information output by the fully connected network layer and convert it into a three-dimensional vector, where the three-dimensional vector represents the vector of the probability distribution of the sleep stage;
[0032] The Argmax function is configured to receive a three-dimensional vector and output the sleep stage with the highest probability.
[0033] Further preferably, a main control module is further provided, and the main control module is respectively connected to the electrical stimulation module and the sleep state recognition module by signals;
[0034] The main control module is configured to receive the current sleep stage output from the sleep state recognition module, and judge the current sleep stage. If it is not the slow-wave sleep stage, the electrical stimulation module is controlled to output corresponding electrical stimulation so that the target to be detected enters the slow-wave sleep state. Otherwise, no electrical stimulation is sent and the sleep state of the target to be detected continues to be judged.
[0035] Due to the adoption of the above technical solution, the present invention has the following advantages and positive effects compared with the prior art:
[0036] The present invention can judge the sleep state and is used for closed-loop evaluation of the effect of the last electrical stimulation of the parafacial nerve nucleus on inducing slow-wave sleep. When the sleep state is judged to be awake or rapid eye movement sleep, it means that electrical stimulation of the parafacial nerve nucleus is needed at this time, that is, to achieve the ideal induction of slow-wave sleep. It can be judged whether to perform the next electrical stimulation based on the effect of the last electrical stimulation. After multiple closed-loop cycles, a fast-induced and lasting slow-wave sleep effect can be obtained. BRIEF DESCRIPTION OF THE DRAWINGS
[0037] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the following detailed description of the preferred embodiment.The drawings are only for the purpose of illustrating the preferred embodiments and are not to be construed as limiting the invention.
[0038] Figure 1 is a structural block diagram of a sleep state recognition module of the present invention;
[0039] Figure 2 A signal flow chart of a closed-loop controlled electrical stimulation system for inducing slow-wave sleep by targeting the parafacial nucleus of the present invention;
[0040] Figure 3 The present invention is a structural block diagram of a closed-loop controlled electrical stimulation system for inducing slow-wave sleep by targeting the parafacial nucleus. DETAILED DESCRIPTION
[0041] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the specific implementation methods of the present invention will be described below with reference to the accompanying drawings. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings and other implementation methods can be obtained based on these drawings without creative work.
[0042] In order to simplify the drawings, only the parts related to the present invention are schematically shown in each figure, and they do not represent the actual structure of the product. In addition, in order to simplify the drawings and facilitate understanding, in some figures, only one of the parts with the same structure or function is schematically drawn or marked. In this article, "one" not only means "only one", but also means "more than one".
[0043] The following is a further detailed description of a closed-loop controlled electrical stimulation system for inducing slow-wave sleep by targeting the parafacial nucleus of the present invention in conjunction with the accompanying drawings and specific embodiments. The advantages and features of the present invention will become more apparent from the following description and claims.
[0044] Example
[0045] See also Figures 1 to 3 , this embodiment is a closed-loop controlled electrical stimulation system for inducing slow-wave sleep by targeting the parafacial nerve nucleus, mainly including three modules: one is an electrical stimulation module, which is used to electrically stimulate the parafacial nerve nucleus of the target to be detected. The second is a sleep state recognition module, which is used to collect the EEG signals of the whole brain of the target to be detected, and perform signal processing on the EEG signals to obtain the sleep stage of the current target to be detected. The third is a main control module, which is respectively connected to the above-mentioned electrical stimulation module and the sleep state recognition module signal, and can receive the current sleep stage of the current target to be detected from the output of the sleep state recognition module, and judge the current sleep stage. If it is not a slow-wave sleep stage (the sleep stage includes wakefulness, slow-wave sleep and rapid eye movement sleep), the electrical stimulation module is controlled to output the corresponding electrical stimulation, so that the target to be detected enters a slow-wave sleep state, otherwise, no electrical stimulation is sent and the sleep state of the target to be detected continues to be judged.
[0046] Specifically, the electrical stimulation module includes an electrical stimulation generation unit and an electrode. The electrical stimulation generation unit is electrically connected to the electrode, and can output electrical stimulation to the electrode, which is used to electrically stimulate the electrode at the position of the paranucleus of the facial nerve to induce and maintain slow wave sleep. Each electrical stimulation includes 1000 pulses, an intensity of 20uA, and a pulse width of 5ms. In order to make the electrode more effectively stimulate the target point during implantation, the electrode in this embodiment uses two strands of nickel-titanium alloy twisted into a conductive wire with a length of 4mm, and then uses a coating process to spray the insulating material onto the surface of the conductive wire where the insulating layer is required to form an insulating layer, so that the electrode can only discharge at the tail discharge end. For a specific implementation example, the experimenter implanted an electrode near the paranucleus of the facial nerve of the mouse (target coordinates: AP = -5.3mm, ML = ±0.7mm, DV = -3.4mm), and used the pulses of the electrode for electrical stimulation, and slow wave sleep was induced by electrical stimulation.
[0047] Then, the sleep state recognition module is described in detail. The sleep state recognition module can judge the sleep state of the mouse (target to be detected) from the whole-brain EEG signal. The whole-brain EEG signal can be collected and recorded by the skull pin electrodes fixed near the anterior and posterior fontanelles. In this embodiment, the convolutional neural network is grouped into a universal time series network, and the universal time series network specifically includes the following structures: a convolutional neural network unit, a fast Fourier transform unit, a circadian time unit, and a result output unit. In order to make the universal time series network suitable for closed feedback loop applications, the EEG signal time window is set to 10 seconds. The EEG signal during this period is used as the input vector. Assuming that z is the input vector, it is input into the d-dimensional convolution kernel κ. Then the output of the τth input vector after the convolution layer network is a vector that satisfies the following formula:
[0048] See also Figure 1 and Figure 2 , the convolutional neural network unit is essentially a plurality of convolutional layers connected in sequence, which are used to perform convolution calculations on the received EEG signals to obtain convolution calculation results. Specifically: the first convolutional layer uses 64 convolution kernels of length 9 and a step size of 1 to perform convolution operations on the EEG signals. The second convolutional layer uses 64 convolution kernels of length 9 and a step size of 2 to perform convolution operations on the data (128 sampling points) output by the first convolutional layer. The third convolutional layer uses 64 convolution kernels of length 9 and a step size of 2 to perform convolution operations on the data (128 sampling points) output by the second convolutional layer. The fourth convolutional layer uses 64 convolution kernels of length 9 and a step size of 2 to perform convolution operations on the data (128 sampling points) output by the third convolutional layer. The fifth convolutional layer uses 64 convolution kernels of length 7 and a step size of 1 to perform convolution operations on the data (128 sampling points) output by the fourth convolutional layer. The sixth convolutional layer uses 64 convolution kernels of length 7 and performs convolution operation on the data (128 sampling points) output by the fifth convolutional layer with a stride of 2. The seventh convolutional layer uses 64 convolution kernels of length 7 and performs convolution operation on the data (128 sampling points) output by the sixth convolutional layer with a stride of 2. The eighth convolutional layer uses 64 convolution kernels of length 7 and performs convolution operation on the data (128 sampling points) output by the seventh convolutional layer with a stride of 2.
[0049] Furthermore, each convolutional layer is followed by a rectified linear unit activation layer (ReLU) and a batch normalization layer (BN). The rectified linear unit activation layer is used to realize nonlinear conversion signals, and the batch normalization layer is used to prevent the gradient vanishing problem that occurs during deep learning. In addition, the above neural network is also provided with a first jump connection and a second jump connection. The first jump connection is located between the output end of the first convolutional layer and the input end of the fifth convolutional layer to capture rough features. The second jump connection is located between the output end of the fifth convolutional layer and the input end of the result output unit to capture detailed features.
[0050] See also Figure 2 , the fast Fourier transform unit is used to receive EEG signals for fast Fourier calculation and obtain fast Fourier transform results. Specifically, its fast Fourier transform spectrum range is 1-12 Hz, covering the key characteristic oscillation activities of the delta wave (1-4 Hz) and theta wave (6-9 Hz) frequency bands. The spectrum is divided into 22 intervals, each interval is 0.5 Hz. The circadian time unit is used to output circadian rhythm time information.
[0051] See also Figure 2 , the result output unit is used to receive the convolution calculation results, the fast Fourier transform results and the circadian rhythm time information and merge them to obtain a three-dimensional vector, and judge the current sleep stage according to the three-dimensional vector. Specifically, the result output unit includes a fully connected network layer, a Softmax function and an Argmax function. The fully connected network layer is used to receive the convolution calculation results, the fast Fourier transform results and the circadian rhythm time information and merge them for output. Since the task is multi-class classification, the Softmax function is used to receive the information output by the fully connected network layer and convert it into a three-dimensional vector. The three-dimensional vector represents the vector of the probability distribution of the sleep stage. Finally, the Argmax function is used to receive the three-dimensional vector and judge the sleep stage with the highest output probability.
[0052] When closed-loop deep brain stimulation based on a general time series network is used, the principle is as follows: the parafacial nucleus is electrically stimulated and the whole-brain EEG signal is obtained at the same time. The whole-brain EEG signal is processed, and the output of the convolutional neural network, the fast Fourier transform results of the EEG signal and the circadian rhythm time are combined and converted into a three-dimensional vector through a fully connected neural network. The vector corresponds to the probability of each sleep stage. The stage with the highest probability is taken as the judgment of the sleep stage. If it is not a slow-wave sleep stage, the stimulation device sends an electric pulse to the parafacial nucleus and continues to judge the sleep state. Otherwise, no electric pulse is sent and the sleep state is continued to be judged, realizing closed-loop control.
[0053] The embodiments of the present invention are described in detail above in conjunction with the accompanying drawings, but the present invention is not limited to the above embodiments. Even if various changes are made to the present invention, if these changes fall within the scope of the claims of the present invention and their equivalents, they still fall within the protection scope of the present invention.
Claims
1. A closed-loop controlled electrical stimulation system targeting the parafacial nucleus to induce slow-wave sleep, characterized in that: include: An electrical stimulation module, configured to electrically stimulate the parafacial nucleus of the target to be detected; The sleep state recognition module is configured to collect the EEG signals of the whole brain of the target to be detected, and perform signal processing on the EEG signals to obtain the current sleep stage of the target to be detected; The sleep state recognition module adopts a universal time series network, which includes a convolutional neural network unit, a fast Fourier transform unit, a circadian time unit and a result output unit; The convolutional neural network unit is configured to receive the EEG signal for convolution calculation to obtain the convolution calculation result; the fast Fourier transform unit is configured to receive the EEG signal for fast Fourier calculation to obtain the fast Fourier transform result; the circadian time unit is configured to output circadian rhythm time information; the result output unit is configured to receive the convolution calculation result, the fast Fourier transform result and the circadian rhythm time information and merge them to obtain a three-dimensional vector, and judge the current sleep stage according to the three-dimensional vector.
2. The closed-loop controlled electrical stimulation system for inducing slow-wave sleep by targeting the parafacial nucleus according to claim 1, characterized in that: The electrical stimulation module includes an electrical stimulation generation unit and electrodes; The electrical stimulation generating unit is electrically connected to the electrode and is configured to output electrical stimulation to the electrode, each electrical stimulation comprising 1000 pulses, with an intensity of 20uA and a pulse width of 5ms; The electrode is made of two strands of nickel-titanium alloy twisted into a conductive wire with a length of 4 mm. The surface of the electrode is wrapped with an insulating layer, while the tail discharge end of the electrode is not wrapped with the insulating layer.
3. The closed-loop controlled electrical stimulation system for inducing slow-wave sleep by targeting the parafacial nucleus according to claim 1, characterized in that: The convolutional neural network unit includes eight convolutional layers, which are: In the first convolution layer, 64 convolution kernels of length 9 are used to perform convolution operation on the EEG signal with a step size of 1; The second convolutional layer uses 64 convolution kernels of length 9 and performs a convolution operation on the data output by the first convolutional layer with a step size of 2; The third convolutional layer uses 64 convolution kernels of length 9 and performs a convolution operation on the data output by the second convolutional layer with a step size of 2; The fourth convolutional layer uses 64 convolution kernels with a length of 9 and performs a convolution operation on the data output by the third convolutional layer with a step size of 2; A fifth convolutional layer, using 64 convolution kernels of length 7 and performing a convolution operation on the data output by the fourth convolutional layer with a step size of 1; A sixth convolutional layer, using 64 convolution kernels of length 7 and performing a convolution operation on the data output by the fifth convolutional layer with a step size of 2; The seventh convolutional layer uses 64 convolution kernels with a length of 7 and performs a convolution operation on the data output by the sixth convolutional layer with a step size of 2; The eighth convolutional layer uses 64 convolution kernels with a length of 7 and performs a convolution operation on the data output by the seventh convolutional layer with a step size of 2.
4. The closed-loop controlled electrical stimulation system for inducing slow-wave sleep by targeting the parafacial nucleus according to claim 3, characterized in that: A rectified linear unit activation layer and a batch normalization layer are provided behind each convolution layer, wherein the rectified linear unit activation layer is configured to implement a nonlinear conversion signal, and the batch normalization layer is configured to prevent gradient disappearance.
5. The closed-loop controlled electrical stimulation system for inducing slow-wave sleep by targeting the parafacial nucleus according to claim 3 or 4, characterized in that: A first hop connection and a second hop connection are also provided; The first jump connection is located between the output end of the first convolutional layer and the input end of the fifth convolutional layer to capture coarse features; The second jump connection is located between the output end of the fifth convolutional layer and the input end of the result output unit to capture detailed features.
6. The closed-loop controlled electrical stimulation system for inducing slow-wave sleep by targeting the parafacial nucleus according to claim 1, characterized in that: The fast Fourier transform spectrum in the fast Fourier transform unit has a range of 1-12 Hz, and the spectrum is divided into 22 intervals, each interval being 0.5 Hz.
7. The closed-loop controlled electrical stimulation system for inducing slow-wave sleep by targeting the parafacial nucleus according to claim 1, characterized in that: The result output unit includes a fully connected network layer, a Softmax function and an Argmax function; The fully connected network layer is configured to receive the convolution calculation result, the fast Fourier transform result and the circadian rhythm time information and output them after merging; The Softmax function is configured to receive the information output by the fully connected network layer and convert it into a three-dimensional vector, where the three-dimensional vector represents a vector of probability distribution of sleep stages; The Argmax function is configured to receive a three-dimensional vector and determine the sleep stage with the highest output probability.
8. The closed-loop controlled electrical stimulation system for inducing slow-wave sleep by targeting the parafacial nucleus according to claim 2, characterized in that: A main control module is also provided, and the main control module is respectively connected to the electrical stimulation module and the sleep state recognition module by signals; The main control module is configured to receive the current sleep stage output from the sleep state recognition module, and judge the current sleep stage. If it is not the slow-wave sleep stage, the electrical stimulation module is controlled to output corresponding electrical stimulation so that the target to be detected enters the slow-wave sleep state. Otherwise, no electrical stimulation is sent and the sleep state of the target to be detected continues to be judged, thereby realizing closed-loop control electrical control.