Closed-loop neural regulation system based on drug and movement status of Parkinson's disease patients

Abstract

The present invention discloses a closed-loop neural regulation system based on the medication and movement state of Parkinson's patients, including: a parameter setting module for determining the low beta frequency band, the high beta frequency band, the upper threshold and the lower threshold; a signal acquisition module for collecting and preprocessing the local field potential signal of the STN of Parkinson's patients; a feature calculation module for calculating the low beta frequency band energy and the high beta frequency band energy through short-time Fourier transform, and calculating the average value of the ratio of the two; a judgment output module for comparing the average value with the upper threshold and the lower threshold; if it is greater than the upper threshold, it is judged that the patient is in the drug failure-movement state and outputs high-intensity stimulation; if it is greater than the lower threshold and less than the upper threshold, it is judged that the patient is in the drug failure-resting state and outputs medium-intensity stimulation; if it is less than the lower threshold, it is judged that the patient is in the drug effective state and outputs low-intensity stimulation. The present invention can achieve precise closed-loop deep brain stimulation based on medication and movement state.

Description

Technical Field

[0001] The present invention belongs to the field of digital medical instruments, and in particular relates to a closed-loop neural regulation system based on medication and movement status of Parkinson's disease patients. Background Art

[0002] Deep brain stimulation (DBS) has been widely used in the clinical treatment of Parkinson's disease. By precisely delivering electrical stimulation to deep brain nuclei, usually the subthalamic nucleus (STN) or the internal globus pallidus (GPi), the function of the corresponding circuits can be restored to normal.

[0003] The conventional open-loop DBS system, which provides continuous, fixed-parameter stimulation, has been shown to significantly alleviate motor symptoms in Parkinson's patients. However, clinical symptoms vary among Parkinson's patients, and individual symptoms fluctuate over varying timescales. Therefore, open-loop DBS, where stimulation parameters cannot be dynamically adjusted based on symptom changes, can potentially produce numerous side effects, such as depression, movement disorders, and dysarthria.

[0004] Using a closed-loop deep brain stimulator can dynamically adjust stimulation parameters based on changes in Parkinson's disease patients' symptoms, achieving closed-loop adaptive DBS (aDBS). Chinese patent publication CN114462455A discloses a method for calculating evaluation indicators for the effectiveness of closed-loop DBS stimulation in Parkinson's disease based on a computational model, including: S1: Signal acquisition: Acquiring LFP signals based on a computational model; S2: Signal preprocessing: Filtering and downsampling the LFP signals; S3: Time domain segmentation: Separating beta bursts from the preprocessed LFP signals; S4: Statistical analysis: Performing statistical analysis based on the beta burst durations obtained in step S3 and performing a binary classification into long and short oscillations; S5: Quantification: Quantifying the binary classification results obtained in step S4 to obtain an evaluation indicator for the effectiveness of closed-loop DBS stimulation.

[0005] In order to make closed-loop DBS therapy more effective, finding biomarkers that are quantifiable, easy to measure, and can accurately reflect the severity of Parkinson's symptoms is a top priority.

[0006] Numerous studies have shown that abnormally increased beta (12-35Hz) activity in the STN nuclei of the basal ganglia is closely associated with the severity of motor symptoms in Parkinson's disease patients. Existing research uses real-time acquisition and analysis of pathological beta oscillations in the STN local field potential (LFP) signal to use the energy in the beta frequency band as a feedback signal to regulate the output of the stimulation current, thereby achieving closed-loop DBS control.

[0007] Based on this, some recent studies have set the closed-loop DBS induction frequency band to the frequency band with the highest concentration of beta energy in patients based on the EEG characteristics of Parkinson's disease patients. However, these methods that only use a single beta frequency band to control stimulation output still have the following two problems:

[0008] First, the energy of the selected beta frequency band is often affected by non-pathological factors, such as daily movement and circadian rhythms, causing its energy to vary or its peak value to shift. Second, the energy of a single beta frequency band often cannot accurately reflect the patient's symptoms, dopaminergic medications, and exercise. This means that it is impossible to distinguish between pathological and autonomic STN-beta activity, resulting in inaccurate closed-loop DBS control. Summary of the Invention

[0009] The present invention provides a closed-loop neural regulation system based on the medication and exercise status of Parkinson's patients, which can accurately identify the patients' different medication and exercise conditions, and accurately implement adaptive closed-loop deep brain electrical stimulation based on dopaminergic drugs and walking movement status.

[0010] A closed-loop neural regulation system based on the medication and movement status of Parkinson's disease patients, including:

[0011] a parameter setting module for determining a low beta frequency band, a high beta frequency band, an upper threshold, and a lower threshold; wherein the upper threshold is used to distinguish between motion states, and the lower threshold is used to distinguish between drug states;

[0012] The signal acquisition module is used to collect the local field potential signal of STN of Parkinson's disease patients in real time, and input the signal into the feature calculation module after signal preprocessing;

[0013] Feature calculation module, used to calculate the low beta band energy several times per second through short-time Fourier transform and high beta band energy , and calculate the low beta band energy every few seconds and high beta band energy Average value of the ratio ;

[0014] The judgment output module is used to calculate the average value obtained by the feature calculation module Compare with the upper and lower thresholds; if If the value is greater than the upper threshold, the patient is judged to be in the drug failure-exercise state, and a high-intensity stimulus is output; if If the value is greater than the lower threshold and less than the upper threshold, the patient is judged to be in the drug failure-resting state. At this time, a medium-intensity stimulus is output; if If the value is less than the lower threshold, it is judged that the patient is in a drug-effective state, and at this time, a low-intensity stimulation is output.

[0015] Furthermore, the specific working process of the parameter setting module is as follows:

[0016] In the three states of drug failure-rest, drug failure-exercise, and drug effectiveness, several segments of bilateral STN local field potential signals of Parkinson's disease patients were collected as baselines;

[0017] Compare the power spectral density between the drug-ineffective and drug-effective states and select the low beta frequency band of drug inhibition; compare the power spectral density between the resting and exercise states and select the high beta frequency band of exercise inhibition;

[0018] The upper threshold for distinguishing the exercise state and the lower threshold for distinguishing the drug state were set according to the baseline data.

[0019] Furthermore, in the parameter setting module, the low beta frequency band, the high beta frequency band, the upper threshold and the lower threshold are all determined individually according to different patients and different sides of the brain.

[0020] Preferably, the sampling rate of the signal acquisition module is 256 Hz, and signal preprocessing of 1-100 Hz bandpass filtering is performed simultaneously.

[0021] Furthermore, in the feature calculation module, the low beta band energy The calculation formula is:

[0022] ;

[0023] Where, Indicates the The low beta band energy corresponding to the window, and is the range of the low beta band, For signal, is the total length of the signal, is the window width, is the window index, For window shifting, is the window function, represents the index of the frequency component, Indicates the index of the signal point, A unit that represents a plural number.

[0024] Furthermore, in the feature calculation module, the high beta band energy The calculation formula is:

[0025] ;

[0026] Where, For the The high beta band energy corresponding to the window, and It is the range of high beta frequency band.

[0027] Furthermore, in the feature calculation module, the low beta band energy is calculated and high beta band energy Average value of the ratio , the formula is:

[0028] ;

[0029] Where, is the window index, is the number of windows.

[0030] Preferably, in the feature calculation module, the low beta band energy is calculated 10 times per second and high beta band energy ; Calculate low beta band energy every 10 seconds and high beta band energy Average value of the ratio .

[0031] Compared with the prior art, the present invention has the following beneficial effects:

[0032] 1. Based on the observation in Parkinson's patients that dopaminergic drugs can significantly suppress low-beta frequency band energy in the STN, while walking exercise can significantly suppress high-beta frequency band energy, this invention defines the ratio of low-beta to high-beta energy in the STN (LHR), thereby accurately identifying different patient medication and exercise conditions. This solves the problem of current closed-loop DBS using a single beta frequency band as a feedback metric, which cannot distinguish between pathological and autonomic STN-beta activity, resulting in inaccurate regulation.

[0033] 2. The present invention's online signal processing and feature calculation method for closed-loop deep-brain stimulators performs a simple short-time Fourier transform several times per second to calculate the energy in a specific frequency band. The average LHR is calculated every several seconds and compared with a threshold. This method strikes a balance between computational accuracy and latency, enabling direct application to the limited performance of implantable pulse stimulators in closed-loop deep-brain stimulator systems. Furthermore, this calculation method addresses the problem of spontaneous fluctuations in STN activity on shorter timescales. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 These are the time-frequency diagrams and power spectral density of the left STN local field potential signal of Parkinson's disease patient A before and after taking dopaminergic drugs in an embodiment of the present invention.

[0035] Figure 2 1 is a time-frequency diagram and power spectral density of the left STN local field potential signal before and after the dopaminergic drug failure in Parkinson's disease patient B in an embodiment of the present invention.

[0036] Figure 3 1 is a time-frequency diagram and power spectral density of the left STN local field potential signal of Parkinson's disease patient A in resting and moving states in an embodiment of the present invention.

[0037] Figure 4 1 is a time-frequency diagram and power spectral density of the left STN local field potential signal of Parkinson's disease patient B during alternating rest and movement in an embodiment of the present invention.

[0038] Figure 5 : The power spectral density of the left STN local field potential signal of Parkinson's disease patients A and B in the embodiment of the present invention under the drug "off"-resting, drug "off"-exercise, and drug "on" states respectively.

[0039] Figure 6 1 shows the time-frequency diagram of the left STN local field potential signal of Parkinson's disease patient A in an embodiment of the present invention, the changes in the low and high beta energy ratio LHR value, and the changes in the closed-loop stimulation amplitude controlled by it.

[0040] Figure 7 This is a schematic diagram of a closed-loop neural regulation system based on the medication and movement status of Parkinson's disease patients in an embodiment of the present invention. DETAILED DESCRIPTION

[0041] The present invention will be described in further detail below with reference to the accompanying drawings and examples. It should be noted that the following examples are intended to facilitate understanding of the present invention and do not have any limiting effect on the present invention.

[0042] The present invention provides a closed-loop neuromodulation system based on medication and movement status of Parkinson's disease patients, using a closed-loop neuromodulation method and a closed-loop deep brain stimulator device to achieve aDBS. The closed-loop neuromodulation method meets the following requirements:

[0043] First, the EEG features selected by the algorithm should be able to accurately identify different dopaminergic drugs and movement states in Parkinson's patients, so that the corresponding current intensity can be output according to the current state. Second, the algorithm's computational complexity should be moderate to avoid overly complex feature calculation methods and excessively large amounts of data for a single calculation, which may exceed the performance of the closed-loop deep brain stimulator device or increase the delay of the closed-loop system.

[0044] Offline analysis of STN LFP signals in resting Parkinson's patients before and after medication revealed significant pathological low-beta (12-20 Hz) energy during the "off" (ineffective) state, when symptoms are more severe. The peak value of this frequency band varies between patients and on different sides of the brain. Approximately 0.5-1 hour after dopaminergic medication administration, the drug gradually takes effect, accompanied by improvement in symptoms. This is the "on" (effective) state, where a significant decrease in pathological low-beta energy can be detected. After 2-4 hours, the drug concentration decreases, causing the drug to lose its effectiveness. EEG results in a return of pathological low-beta energy to pre-drug levels, accompanied by the reappearance of symptoms.

[0045] During this process, varying levels of STN high-beta (20-35Hz) activity persisted, and the frequency range of this energy band varied between different hemispheres of the brain in different patients. Similarly, when the drug was "off," the patient's voluntary walking exercise significantly suppressed STN high-beta activity and attenuated the energy in the high-beta frequency band. Furthermore, voluntary walking exercise also shifted the peak of the previously highly active low-beta frequency band to the right, to a higher frequency, while maintaining essentially the same total energy. This suggests that dopaminergic drugs and walking exercise modulate STN activity by attenuating low-beta and high-beta energy, respectively.

[0046] Therefore, the present invention calculates the ratio of low-beta to high-beta energy (LHR) in specific frequency bands within the left and right STNs of Parkinson's patients to differentiate between different medication and exercise states. Specifically, the "off" state (exercise state) corresponds to higher low-beta energy and lower high-beta energy, resulting in the highest LHR value; the "off" state (resting state) corresponds to higher low-beta and high-beta energy, resulting in a moderate LHR value; and the "on" state (drug state) corresponds to lower low-beta energy and higher high-beta energy, resulting in the lowest LHR value. Using the LHR value as a feedback signal, aDBS based on dopaminergic medication and exercise state can be implemented.

[0047] Abnormal enhancement of beta (12-35 Hz) activity in the STN nuclei of the basal ganglia is closely related to the severity of motor symptoms in patients with Parkinson's disease. Figure 1 The time spectrum and power spectral density (PSD) of the left STN local field potential signal of Parkinson's disease patient A before and after taking dopaminergic drugs. Figure 1 As shown, the top graph is a time spectrum plot, with time (in minutes) on the horizontal axis and frequency (in hertz) on the vertical axis. The colors in the graph represent the power spectral density value (in decibels) of a specific frequency component at a given moment. Colors closer to red indicate higher energy, while colors closer to blue indicate lower energy. Power spectral density analysis was performed on the 5-minute signal before and after medication administration (marked in white), resulting in the power spectral density plot shown at the bottom, with frequency on the horizontal axis and power spectral density value on the vertical axis. It can be seen that before medication, when the patient's symptoms were more severe, there was significant pathological low-beta energy with a peak frequency of 13 Hz, as well as some high-beta energy. After the medication took effect, this pathological low-beta energy was rapidly suppressed, leaving only some high-beta energy. At this point, the patient's symptoms were alleviated and he felt comfortable.

[0048] Figure 2 These are the time-frequency plots and power spectral density of the local field potential signal in the left STN of Parkinson's disease patient B before and after the dopaminergic drug expires. The top plot shows the time-frequency plot, and the power spectral density plot below is obtained by analyzing the 5-minute signal before and after the drug expires (white highlights). This analysis reveals that, unlike Parkinson's disease patient A, patient B exhibits more high-beta energy in his EEG when the drug is effective. However, after the drug expires, the pathological low-beta energy, peaking at 18 Hz, increases significantly, while the high-beta energy returns to a lower level, and the patient's symptoms reappear. In summary, dopaminergic drugs can significantly inhibit pathological low-beta activity in the STN.

[0049] Currently, most studies, based on the above-mentioned EEG characteristics, set the closed-loop DBS induction frequency band to the frequency band where the pathological low-beta energy of Parkinson's patients is most concentrated. However, using only a single beta band energy to control the stimulation output still has the following problems: First, the energy of the selected beta band is often affected by non-pathological factors, such as daily movement and circadian rhythm, resulting in its energy changes or peak shifts; second, the energy of a single beta band often cannot accurately reflect the patient's symptoms, dopaminergic drugs, and movement at the same time, that is, it is impossible to distinguish between the pathological activity and autonomic activity of STN-beta, resulting in inaccurate closed-loop DBS regulation.

[0050] According to the latest research and observations of Parkinson's disease patients, it was found that under the daily programmed open-loop stimulation parameters, the fluctuation of dopaminergic drugs (such as Figure 1 and Figure 2 As shown in the figure, it can lead to negative symptoms such as dyskinesia, so the closed-loop stimulation dynamically adjusted according to the changes in pathological low beta energy is necessary. In addition, more importantly, it was found that the current size of daily programming cannot meet the needs of Parkinson's patients for autonomous movement, especially walking movement. The results of clinical tests show that, based on the drug failure state and daily programming parameters, increasing the stimulation intensity within a certain range can improve the patient's motor performance to a certain extent, especially further alleviating their bradykinesia symptoms. Therefore, it is equally important to dynamically adjust the stimulation parameters according to the patient's movement state.

[0051] Figure 3 Figure 3. Time-frequency plots and power spectral density of the left STN local field potential signal in Patient A with Parkinson's disease, both at rest and during exercise, after drug inactivation. The top plot shows the time-frequency plot, while the bottom plot shows the power spectral density of the resting and walking signals, which are marked as intercepted. As can be seen, when the patient's medication is "off," after transitioning from rest to walking, the energy in the high-beta band (20-35Hz) is significantly suppressed, while the peak frequency of pathological low-beta energy shifts rightward from 13Hz to 14Hz. Furthermore, there is a significant increase in low-frequency energy in the 2-10Hz range.

[0052] Taking Parkinson's disease patient B as an example, Figure 4 The time-frequency plots and power spectral density of the left STN local field potential signal during alternating rest and exercise in this patient are shown. The top plot shows the time-frequency plot, and the bottom plot shows the power spectral density of the resting and walking signals, respectively. It can be seen that in the patient's "off" drug state, after transitioning from rest to walking, the energy in the high-beta frequency band (24-35Hz) is significantly suppressed, while the peak frequency of pathological low-beta energy shifts rightward from 18Hz to 20Hz.

[0053] In summary, dopaminergic drugs significantly suppress the energy of the low-beta frequency band in the STN, while walking exercise significantly suppresses the energy of the high-beta frequency band. Therefore, we defined the ratio of low-beta to high-beta energy in the STN as the LHR. The change in this value serves as the feedback control signal for closed-loop DBS, allowing appropriate stimulation to be delivered based on accurate identification of the patient's medication and exercise status.

[0054] like Figure 7 As shown, a closed-loop neural control system based on the medication and movement status of Parkinson's disease patients includes:

[0055] a parameter setting module for determining a low beta frequency band, a high beta frequency band, an upper threshold, and a lower threshold; wherein the upper threshold is used to distinguish between motion states, and the lower threshold is used to distinguish between drug states;

[0056] The signal acquisition module is used to collect the local field potential signal of STN of Parkinson's disease patients in real time, and input the signal into the feature calculation module after signal preprocessing;

[0057] Feature calculation module, used to calculate the low beta band energy several times per second through short-time Fourier transform and high beta band energy , and calculate the low beta band energy every few seconds and high beta band energy Average value of the ratio ;

[0058] The judgment output module is used to calculate the average value obtained by the feature calculation module Compare with the upper and lower thresholds; if If the value is greater than the upper threshold, the patient is judged to be in the drug failure-exercise state, and a high-intensity stimulus is output; if If the value is greater than the lower threshold and less than the upper threshold, the patient is judged to be in the drug failure-resting state. At this time, a medium-intensity stimulus is output; if If the value is less than the lower threshold, it is judged that the patient is in a drug-effective state, and at this time, a low-intensity stimulation is output.

[0059] Specifically, the closed-loop neural regulation system of the present invention has the following specific algorithm flow:

[0060] S1. Under three states: drug "off" (ineffective) - rest, drug "off" (ineffective) - exercise, and drug "on" (effective), several segments of local field potential signals from the bilateral STN of Parkinson's disease patients were collected as baselines.

[0061] S2, offline calculation of the average power spectrum density of each of the above states, Figure 5The power spectral density of the left STN local field potential signal for Parkinson's disease patients A and B during the drug "off" resting state, drug "off" exercise state, and drug "on" state, respectively. Comparison of the power spectral density between the drug "off" and "on" states revealed significant drug suppression in the low-beta frequency band (11-17 Hz for the left patient A and 14-22 Hz for the right patient B). Comparison of the power spectral density between the resting and exercise states revealed significant suppression in the high-beta frequency band (18-34 Hz ​​for the left patient A and 24-34 Hz ​​for the right patient B).

[0062] S3. Use the ratio of low-beta to high-beta frequency band energy (LHR) as an EEG marker for identifying medication and exercise status in Parkinson's patients. For Parkinson's patient A, based on the corresponding LHR baseline values ​​for the three states, set the upper threshold for distinguishing exercise status, TH_h, to 5.0, and the lower threshold for distinguishing medication status, TH_l, to 1.0.

[0063] S4. Real-time acquisition of local field potential signals from the STN of Parkinson's disease patients with a sampling rate of 256 Hz, and simultaneous signal preprocessing with 1-100 Hz bandpass filtering. Figure 6 The top image shows the time-frequency plot of the left STN-LFP acquired in real time from patient A, showing the process of drug onset and the transition between resting and walking. The walking time period is marked in blue, the drug-onset time period in orange, and the remaining unmarked time period is the resting time period before the drug takes effect. This signal segment is used as an example to further illustrate the detection of changes in the low-to-high beta power ratio (LHR) and the corresponding control of closed-loop stimulation intensity.

[0064] S5, through short-time Fourier transform, online calculation 10 times per second (set the sliding window width to 1 second and the coverage rate to 90%) low beta band energy and high beta band energy , the calculation formula is:

[0065] ;

[0066] ;

[0067] in, and is the range of the low beta band, and is the range of the high beta band, For signal, is the total length of the signal, is the window width, is the window index, For window shifting, is the window function, represents the index of the frequency component, represents the index of the signal point, represents the unit of complex numbers. Divide the low beta energy by the high beta energy to obtain the LHR value, and its variation over time is as shown by Figure 6 the black curve in the middle.

[0068] S6. Calculate the average value of the LHR, which is the ratio of the high and low beta band energies within 10 seconds. The calculation formula is:

[0069] ;

[0070] where is the number of windows. The variation of the average LHR value is as shown by Figure 6 the blue line segment in the middle.

[0071] S7. Compare the average LHR value with the upper and lower thresholds: If LHR > TH_h, it means the patient is in the "off"-medication - movement state, and at this time, a higher-intensity stimulation (2.4 mA) is output to relieve the movement symptoms to the greatest extent; if TH_h > LHR > TH_l, it means the patient is in the "off"-medication - rest state, and at this time, a moderate daily programmed intensity stimulation (2.2 mA) is output to meet the need for basic symptom relief; if LHR < TH_l, it means the patient is in the "on"-medication state, and at this time, a lower-intensity stimulation (2.0 mA) is output to avoid over-stimulation symptoms such as dyskinesia and dizziness. Specifically, as shown by Figure 6 the current change curve at the bottom.

[0072] S8. Analyze the operation results of the personalized aDBS algorithm based on the medication and movement states, that is Figure 6 the time-frequency diagram of the local field potential signal of the left STN of Parkinson's disease patient A, the variation of the LHR value representing the proportion of low and high beta energies, and the variation of the closed-loop stimulation amplitude controlled thereby. It can be seen that before the onset of drug efficacy, the LHR value is always greater than the lower threshold, and during the vast majority of walking movements, the LHR value exceeds the upper threshold, and at this time, the output current is adjusted up to 2.4 mA, the maximum gear; while during rest, the LHR value often lies between the upper and lower thresholds, and at this time, the output current is adjusted to 2.2 mA, the middle gear of the daily program; after the onset of drug efficacy, the LHR value is basically lower than the lower threshold, and at this time, the output current is adjusted down to 2.0 mA, the lowest gear.

[0073] Therefore, the present invention can distinguish the electroencephalogram signal characteristics of Parkinson's disease patients in different medication and movement states, use the LHR of the low and high beta energy ratio as the feedback control signal of closed-loop DBS, and achieve adaptive on-demand stimulation in the three states of the patient's "off"-medication - movement, "off"-medication - rest, and "on"-medication.

[0074] The embodiments described above provide a detailed description of the technical solutions and beneficial effects of the present invention. It should be understood that the above are only specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, supplements and equivalent substitutions made within the scope of the principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A closed-loop neural regulation system based on the medication and movement status of Parkinson's disease patients, characterized by: include: The parameter setting module is used to determine the low beta frequency band, the high beta frequency band, the upper threshold, and the lower threshold. The upper threshold is used to distinguish between exercise states, and the lower threshold is used to distinguish between drug states. The specific working process of the parameter setting module is as follows: In the three states of drug failure-rest, drug failure-exercise, and drug effectiveness, several segments of bilateral STN local field potential signals of Parkinson's disease patients were collected as baselines; Compare the power spectral density between the drug-ineffective and drug-effective states and select the low beta frequency band of drug inhibition; compare the power spectral density between the resting and exercise states and select the high beta frequency band of exercise inhibition; The upper threshold for distinguishing the exercise state and the lower threshold for distinguishing the drug state are set according to the baseline data; In the parameter setting module, the low beta frequency band, high beta frequency band, upper threshold, and lower threshold are determined individually for different patients and different sides of the brain; The signal acquisition module is used to collect the local field potential signal of STN of Parkinson's disease patients in real time, and input the signal into the feature calculation module after signal preprocessing; Feature calculation module, used to calculate the low beta band energy several times per second through short-time Fourier transform and high beta band energy , and calculate the low beta band energy every few seconds and high beta band energy Average value of the ratio ; The judgment output module is used to calculate the average value obtained by the feature calculation module Compare with the upper and lower thresholds; if If the value is greater than the upper threshold, the patient is judged to be in the drug failure-exercise state, and a high-intensity stimulus is output; if If the value is greater than the lower threshold and less than the upper threshold, the patient is judged to be in the drug failure-resting state. At this time, a medium-intensity stimulus is output; if If the value is less than the lower threshold, it is judged that the patient is in a drug-effective state, and at this time, a low-intensity stimulation is output.

2. The closed-loop neural regulation system based on Parkinson's disease patient medication and movement status according to claim 1, characterized in that: The sampling rate of the signal acquisition module is 256 Hz, and the signal preprocessing is performed with 1-100 Hz band-pass filtering.

3. The closed-loop neural regulation system based on Parkinson's disease patient medication and movement status according to claim 1, characterized in that: In the feature calculation module, the low beta band energy The calculation formula is: ； Where, Indicates the The low beta band energy corresponding to the window, and is the range of the low beta band, For signal, is the total length of the signal, is the window width, is the window index, For window shifting, is the window function, represents the index of the frequency component, Indicates the index of the signal point, A unit that represents a plural number.

4. The closed-loop neural regulation system based on Parkinson's disease patient medication and movement status according to claim 3, characterized in that: In the feature calculation module, high beta band energy The calculation formula is: ； Where, For the The high beta band energy corresponding to the window, and It is the range of high beta frequency band.

5. The closed-loop neural regulation system based on Parkinson's disease patient medication and movement status according to claim 4, characterized in that: In the feature calculation module, calculate the low beta band energy and high beta band energy Average value of the ratio , the formula is: ； Where, is the window index, is the number of windows.

6. The closed-loop neural regulation system based on Parkinson's disease patient medication and movement status according to claim 5, characterized in that: In the feature calculation module, the low beta band energy is calculated 10 times per second and high beta band energy ; Calculate low beta band energy every 10 seconds and high beta band energy Average value of the ratio .

Images