Evaluation system for treatment effect of epileptic

Abstract

The invention relates to the technical field of medical software, and discloses a treatment effect evaluation system for epilepsy patients, which comprises a data acquisition module, a treatment module, a treatment effect evaluation module and an electroencephalogram signal analysis module, further performs prediction and automatic detection on the treatment effect through computer processing, achieves the effects of accurate evaluation and dynamic detection, and improves the treatment effect of the epilepsy patients. A'detection-intervention-evaluation 'closed loop is formed, and a complete solution is provided for individual precise treatment of epilepsy.

Description

Technical Field

[0001] The present invention relates to the technical field of medical software, and more particularly to a treatment effect evaluation system for epilepsy patients. Background Art

[0002] Epilepsy is a common chronic neurological disease that affects more than 70 million people worldwide. The treatment of epilepsy is mainly based on anti-epileptic drugs. With the continuous emergence of new anti-epileptic drugs, most people have clinical improvement. However, nearly one-third of people suffer from refractory epilepsy. These patients' epileptic seizures cannot be controlled or they cannot tolerate the side effects of drugs, leading to injuries, psychosocial dysfunction and decreased quality of life. Therefore, they need more effective alternative treatment options. In 1999, Terge et al. first used repetitive transcranial magnetic stimulation (rTMS) in patients with refractory focal epilepsy, reducing epileptic seizures by 40%. rTMS has become one of the important neuromodulatory methods. Currently, the evaluation indicators for rTMS treatment of epilepsy are mainly seizure frequency and interictal epileptiform discharges, but the clinical outcomes are not uniform.

[0003] EEG signal analysis uses computers to process EEG signals, extracting features in the time, frequency, time-frequency, spatial, and nonlinear domains to analyze more subtle EEG signal changes. Variational mode decomposition (VMD) combines time-frequency and nonlinear methods to segment the original EEG signal and then extracts fine-scale composite multiscale scatter entropy and fine-scale composite multiscale fuzzy entropy. This method classifies lesion and non-lesion signals and automatically monitors EEG signals. It accurately quantifies EEG signals across multiple dimensions, improving the accuracy of efficacy assessments for intractable epilepsy. Summary of the Invention

[0004] In order to overcome the above-mentioned defects in the prior art, the present invention provides a treatment effect evaluation system for epilepsy patients, including a data acquisition module, a treatment module, an efficacy evaluation module and an EEG signal analysis module. Through computer processing, the efficacy is further predicted and automatically detected, achieving the effects of accurate evaluation and dynamic detection, forming a "detection-intervention-evaluation" closed loop, and providing a complete solution for individualized and precise treatment of epilepsy.

[0005] The above technical objectives of the present invention are achieved through the following technical solutions: a system for evaluating the therapeutic effect of epilepsy patients, comprising a data acquisition module, a treatment module, an efficacy evaluation module and an EEG signal analysis module;

[0006] The data collection module is used to collect demographic information and medical history of the subjects. Demographic information includes age, gender, education level and occupation; medical history includes surgical history, allergy history, smoking history, family history and menstrual status of women.

[0007] The treatment module treats epilepsy by using low-frequency rTMS stimulation;

[0008] The efficacy evaluation module includes a cognitive function assessment unit and a motor function assessment unit. The cognitive function assessment unit assesses subjects with the Mini-Mental State Examination, Montreal Cognitive Assessment Scale, and a line connection test before and after rTMS treatment. The motor function assessment unit uses the Readygo motor function quantitative assessment system to record the subjects' three-dimensional spatiotemporal gait parameters, including the temporal parameters of gait speed and cadence, and the spatial parameters of stride length, stride height, and stride width.

[0009] The EEG signal analysis module evaluates the interictal EEG discharges during rTMS treatment of epilepsy and performs EEG signal analysis based on variational mode decomposition.

[0010] Furthermore, the Mini-Mental State Examination was used to evaluate global cognitive function, including orientation, immediate word memory, attention and calculation ability, word recall ability, language, and structural imitation ability.

[0011] Furthermore, the Montreal Cognitive Assessment was used to screen for cognitive impairment, including cognitive domains of attention and concentration ability, executive function, memory, language function, visual-spatial function, abstract thinking ability, calculation ability, and orientation ability.

[0012] Furthermore, the connection experiment includes connection experiment A and connection experiment B.

[0013] Furthermore, the sampling rate of the amplifier for collecting EEG data in the EEG signal analysis module is 200-1000 Hz, the high-pass filter is ≥70 Hz, the low-pass filter is ≤0.5 Hz, the notch filter is 50 Hz, and the sensitivity is 10 uV / mm.

[0014] Furthermore, the EEG signal analysis based on variational mode decomposition includes the following steps:

[0015] S1. Preprocessing; S2. Construction of variational modal model; S3. Solution of variational model; S4. Feature extraction using fuzzy entropy algorithm; S5. T-test for feature selection.

[0016] In summary, the present invention has the following beneficial effects: the present invention predicts and automatically detects the therapeutic effect through computer processing, achieves the effect of accurate evaluation and dynamic detection, forms a "detection-intervention-evaluation" closed loop, and provides a complete solution for personalized and precise treatment of epilepsy. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a graph showing the frequency of epileptic seizures at various treatment time points in an embodiment of the present invention;

[0018] Figure 2is a diagram showing the fuzzy entropy analysis results of the lucid period before and after treatment in an embodiment of the present invention;

[0019] Figure 3 This is a diagram showing the fuzzy entropy analysis results of sleep stage N1 before and after treatment in an embodiment of the present invention;

[0020] Figure 4 This is a diagram showing the fuzzy entropy analysis results of sleep stage N2 before and after treatment in an embodiment of the present invention;

[0021] Figure 5 3 is a diagram showing the fuzzy entropy analysis results of sleep stage N3 before and after treatment in an embodiment of the present invention. DETAILED DESCRIPTION

[0022] The following is combined with Figure 1-5 The present invention is described in further detail.

[0023] Example 1: A system for evaluating the therapeutic effect of epilepsy patients, comprising a data acquisition module, a treatment module, an efficacy evaluation module, and an EEG signal analysis module;

[0024] The data collection module is used to collect demographic information and medical history of the subjects. Demographic information includes age, gender, education level and occupation; medical history includes surgical history, allergy history, smoking history, family history and menstrual status of women.

[0025] The treatment module treats epilepsy by using low-frequency rTMS stimulation;

[0026] in:

[0027] Stimulation equipment: The magnetic stimulator and coil were both manufactured by Tonica Elektronik A / S of Denmark. A figure-8 coil was used, with a single-side inner diameter of 10 mm, an outer diameter of 65 mm, and a height of 18 mm. The coil was fixed tangentially to the skull surface by a robotic arm.

[0028] Preparation before stimulation: The patient sits in an armchair in a relaxed state. A single-pulse stimulation is used to measure the resting motor threshold, inducing at least 5 maximum finger movements in 10 consecutive stimulations. The stimulation site is then marked. If the resting threshold is elevated and cannot be measured due to taking 3 or more anti-epileptic drugs, 65% of the maximum output is selected to provide stimulation.

[0029] Stimulation parameters: For focal epilepsy, the epileptogenic zone or symptom-producing area was selected based on clinical manifestations during seizures, interictal epileptiform discharges on electroencephalogram (EEG), cranial magnetic resonance imaging (MRI), positron emission tomography / computed tomography (CT), magnetoencephalography (MEG), and stereotactic EEG. For multifocal or generalized epilepsy, the Cz point on the international 10-20 EEG system was used as the target. Parameters were selected in accordance with the safety recommendations of the International Federation of Clinical Electrophysiology guidelines. The stimulation frequency was 1 Hz, the intensity was 90% of the resting motor threshold, and each treatment consisted of 30 trains of 40 pulses, for a total of 1200 pulses. The interval between trains was 1 second, for a total duration of 19 minutes and 50 seconds.

[0030] Stimulation duration: once a day, continuous treatment for 15 days, rest for 10 days as a cycle, continuous treatment for 3 cycles, a total of 75 days of treatment

[0031] The efficacy evaluation module includes cognitive function evaluation unit and motor function evaluation unit:

[0032] The cognitive function assessment unit evaluated the subjects with the Mini-Mental State Examination, Montreal Cognitive Assessment, and Wiring Test before and after rTMS treatment;

[0033] The motor function assessment unit used the Readygo Motor Function Quantitative Assessment System (Zhongke Ruiyi Information Technology Co., Ltd., Beijing) to record the subjects' three-dimensional spatiotemporal gait parameters, including the temporal parameters of gait speed and cadence, and the spatial parameters of stride length, step height, and step width. Walking patterns included single-task walking and dual-task walking. Tasks included continuous calculation, language production, and hyperventilation. Single-task walking required the patient to walk three laps back and forth on a 3-meter walkway at their usual speed, for a total distance of 18 meters. Dual-task walking involved performing a cognitive task based on single-task walking. The continuous calculation task involved the subject selecting a number between 90 and 99, then continuously subtracting 7 from that number and stating the result. The language production task involved the subject selecting an animal, vegetable, or fruit category and stating as many words as possible belonging to that category. During dual-task walking, any calculation or naming errors did not require correction or discontinuation; the walking task continued. The hyperventilation task required the subjects to take deep breaths continuously for 3 minutes in a sitting position, with a respiratory rate of 20-25 times / min and a ventilation volume 5-6 times the normal level, and then immediately complete a single-task walking task.

[0034] The EEG signal analysis module evaluates the interictal EEG discharges during rTMS treatment of epilepsy and performs EEG signal analysis based on variational mode decomposition.

[0035] EEG equipment: The amplifier used to collect EEG data was from Nicolet (USA). The sampling rate was 200–1000 Hz, the high-pass filter was ≥70 Hz, the low-pass filter was ≤0.5 Hz, the notch filter was 50 Hz, and the sensitivity was 10 μV / mm.

[0036] Preparation before EEG acquisition: Antiepileptic drugs should not be discontinued, and the use of sedatives, hypnotics or stimulants should be avoided.

[0037] EEG acquisition: 21 disc electrodes were placed according to the 10-20 system, along with two electrocardiogram (ECG) electrodes in the precordial area and two myoelectric (EMG) electrodes in the left and right deltoid muscles. After acquisition began, induced experiments were performed, including an eyes-opening and closing test, a hyperventilation test, and a flash stimulation test.

[0038] Data collection time: All patients completed 14 hours of video EEG monitoring in the EEG room of the First Hospital of Shanxi Medical University before and after rTMS treatment.

[0039] Interpretation of results: Based on the recommendations of two long-term neuroelectrophysiologists, EEG images were read using standard parameters, with a time constant of 0.1 seconds and a 9-second frame. Epileptic discharges were calculated using alternative lead configurations, including ear pole reference, average reference, and bipolar reference, as needed.

[0040] The Mini-Mental State Examination was used to assess global cognitive function, including orientation, immediate word memory, attention and calculation ability, word recall ability, language, and structural imitation ability.

[0041] The Montreal Cognitive Assessment is used to screen for cognitive impairment, including the cognitive domains of attention and concentration, executive function, memory, language function, visual-spatial function, abstract thinking ability, calculation ability, and orientation.

[0042] The line connection experiment includes line connection experiment A and line connection experiment B. Line connection experiment A (TMT-A) is to connect circles with Arabic numerals in order of size and record the completion time; line connection experiment B (TMT-B) is to alternately connect white circles and black circles containing Arabic numerals and record the completion time.

[0043] The sampling rate of the amplifier that collects EEG data in the EEG signal analysis module is 200-1000 Hz, the high-pass filter is ≥70 Hz, the low-pass filter is ≤0.5 Hz, the notch filter is 50 Hz, and the sensitivity is 10 uV / mm.

[0044] EEG signal analysis based on variational mode decomposition includes the following steps:

[0045] S1. Preprocessing: Select 3 minutes of interictal EEG data during each period of wakefulness and sleep, including wakefulness, non-rapid eye movement sleep (NREM) 1, 2, and 3. First, browse the EEG signal to manually remove artifacts and cut out obvious noise segments and bad leads. Then eliminate baseline drift to prevent excessive signal deviation. Then perform a 0.5Hz high-pass filter and a 45Hz low-pass filter on the signal, and use a 50Hz notch filter to remove non-physiological artifacts such as power frequency interference. Finally, use independent component analysis (ICA) to remove electrooculogram and electrocardiogram artifacts;

[0046] S2. Construction of variational modal model. In the variational model, the intrinsic mode function (IMF) is defined as an amplitude-frequency modulation signal, and its expression is:

[0047] u k (t) = A k (t)cos(φ k (t))

[0048] Among them A k (t) is the instantaneous amplitude, φ k The derivative of (t) is the instantaneous frequency.

[0049] Assume that each mode u k All have a center frequency and a finite bandwidth. The constraints are that the estimated bandwidth of each IMF is minimum and the sum of all IMFs is equal to the input signal. The constraint model is expressed as:

[0050]

[0051]

[0052] Where K represents the number of IMFs, f is the input signal, {u k}={u1,u2,u3,...,u K} represents the K bandwidth-limited IMF components obtained by decomposition, {w k}={w1,w2,w3,…,w K} represents the center frequency of each IMF;

[0053] S3. Solve the variational model. Before signal decomposition, determine the number of decomposed modes K and the penalty factor α. A penalty factor α that is too large will cause modal overlap, while a too small value will introduce noise. A value of K that is too large or too small will affect the decomposition and, consequently, the final results. Based on literature review, this study set the penalty factor α to 2000 and the number of decomposed modes K to 5.

[0054] To solve the above constrained variational model, the constructed constrained variational model is transformed into an unconstrained variational model by introducing a quadratic penalty term and Lagrange multipliers. The expanded Lagrange expression is:

[0055]

[0056] The equation is solved in the frequency domain using the alternating direction multiplier algorithm. The main steps to solve it are as follows: Input: signal to be decomposed The number of patterns to be decomposed K; Output: K patterns obtained by decomposition.

[0057] (1) λ 1 and n are initialized.

[0058] (2) According to the following formula and λ 1 Update of , where k = 1, 2...K.

[0059]

[0060]

[0061] (3) Repeat the above steps until the stopping condition is met, that is:

[0062]

[0063] Where r is the threshold.

[0064] In summary, VMD is completed and k IMF components are obtained.

[0065] S4. Fuzzy entropy algorithm extracts features. Fuzzy entropy is used to measure the probability of a new pattern and can identify weak signals. The definition algorithm of fuzzy entropy is:

[0066] (1) The N-point sampling sequence is:

[0067] [x(1),x(2),…x(N)]

[0068] (2) Define the relative space dimension m and similarity tolerance r, and construct the image space:

[0069] U(i)=[x(i),x(i+1),...x(i+m-1)]-x0(i),i=1,2,...,N-m+1.

[0070] in

[0071] (3) Introducing fuzzy membership function:

[0072] A(y)=1,y=0

[0073]

[0074] Among them, r is the similarity tolerance.

[0075] (4) For i=1, 2, ...N-m+1, calculate:

[0076] j=1,2,...N-m+1 and j≠i

[0077] in is the maximum absolute distance between window vectors.

[0078] (5) For each i, find its average value and get:

[0079]

[0080] definition

[0081] (6) Therefore, the fuzzy entropy of the original time series is:

[0082]

[0083] S5. T-test selection of features. The purpose of feature selection is to identify and select statistically significant features with significant differences and eliminate features with no significant differences to produce positive final results. This study uses the statistical t-test method, based on the statistical significance of the distribution of features between different groups, to express the standard value of significant differences.

[0084] Example 2: Experimental Verification

[0085] General Information of Epilepsy Patients: In this example, 15 epilepsy patients were included, with an average age of 33.20 ± 16.16 years, 60% of whom were female. The age of onset of epilepsy patients ranged from 1 to 40 years old, with an average onset age of 18.73 ± 11.48 years (see Table 1).

[0086] All patients presented with focal epilepsy, with the most common seizure types being automatisms (46.6%), focal seizures progressing to bilateral tonic-clonic seizures (46.6%), autonomic seizures (26.67%), behavioral arrest seizures (13.3%), and sensory seizures (6.67%). Based on the patient's medical history, clinical manifestations, and EEG features, five patients met the clinical syndrome diagnosis: four with mesial temporal lobe epilepsy and one with lateral temporal lobe epilepsy. Comorbidities with epilepsy primarily included cognitive impairment, sleep disturbances, and mood problems.

[0087] All patients were taking two or more antiepileptic drugs, the most common of which were levetiracetam (53.3%), lamotrigine (53.3%), oxcarbazepine (40%), lacosamide (33.3%), perampanel (33.3%), sodium valproate (20%), clonazepam (20%), and phenobarbital (6.67%).

[0088] Table 1 Demographic and clinical characteristics of epilepsy patients

[0089]

[0090] Clinical efficacy of rTMS in treating epilepsy:

[0091] Comparison of seizure frequency before and after treatment: This example involved a 45-day rTMS treatment period followed by a 30-day rest observation period. Fifteen patients received the first course of treatment, 11 received the second, and eight completed all three courses. Reasons for patient dropout during treatment included: fractures preventing continued treatment at the hospital, long travel distances, lack of family support, and seeking alternative treatments at tertiary centers.

[0092] Seizure frequency was assessed before treatment, on days 15, 25, 40, 50, 65, and 75 (T0-T6). The average seizure frequency for the 15 patients was 1.75 ± 2.07 times per week in the first two months of treatment. Seizure frequency showed a downward trend throughout treatment, with the exception of a sudden increase during the second rest period (T3). Comparison of the average weekly seizure frequency at each time point with the baseline level showed no significant statistical difference (P > 0.05). Comparisons before and after each clinical treatment were also statistically significant (P > 0.05).

[0093] Clinical follow-up was conducted 6 months after the end of treatment. The seizure frequency of 2 / 4 patients who completed one course of treatment (15 days) was reduced compared with the baseline, and the effect was maintained for 1 month. 3 patients received two courses of treatment (30 days), and the average weekly seizure frequency of 1 / 3 patients decreased after treatment, and the effect lasted for 1.5 months. 8 patients continued to complete three courses (45 days) of rTMS treatment. The weekly seizure frequency of 3 / 8 patients was still lower than before treatment during follow-up 6 months after the end of treatment, but the difference was not statistically significant (P < 0.05). One patient experienced discomfort at the stimulation site, manifested as numbness and stiffness of the face on the stimulated side, which improved after the stimulation stopped and was completely relieved after 1 week. The frequency of epileptic seizures at each treatment time point is as follows: Figure 1 shown.

[0094] Comparison of the number of epileptic discharges before and after treatment: Seven patients completed 14-hour video EEG monitoring before and after treatment. The average number of epileptic discharges during the interictal period before treatment was 21.74±13.46 / hour, and after treatment it was 11.24±4.38 / hour, and the difference was statistically significant (P<0.05).

[0095] The efficacy of rTMS in treating epilepsy comorbidities:

[0096] Comparison of cognitive function before and after treatment: A total of 7 patients completed cognitive assessments using the MMSE, MoCA, and a connection test at baseline and after rTMS treatment. The mean MMSE score before treatment was 26.29±3.59, and the mean score after treatment was 28.29±1.98, with no statistically significant difference (P>0.05). The mean MoCA score before treatment was 22.86±3.81, and the mean score after treatment was 24.57±4.82, with no statistically significant difference (P>0.05). The mean time spent on the TMT-A and TMT-B before treatment was 65±13.77 and 88.67±12.43, respectively. After treatment, the mean time spent on the TMT-A and TMT-B was 61.07±19.01 and 97.67±41.49, respectively, with no statistically significant difference (P>0.05) (Table 2).

[0097] Table 2 Comparison of cognitive function scores in epilepsy patients before and after rTMS treatment

[0098]

[0099] Before treatment, the MMSE scores for time orientation, place orientation, immediate memory, attention and calculation, recall, language, and construction were 4.00±1.00, 4.43±0.79, 2.86±0.38, 3.71±1.38, 2.57±0.79, 7.71±0.75, and 1.00±0.00, respectively. After treatment, all scores improved compared to baseline, but only the scores for attention and calculation showed statistically significant differences before and after treatment (P<0.05); the other scores did not differ significantly (P>0.05) (Table 3).

[0100] Table 3 Comparison of MMSE scores in epilepsy patients before and after treatment

[0101]

[0102] Comparison of gait function before and after treatment: Single-task gait assessments were performed before treatment, on the 15th, 40th, and 65th days of treatment (T0, T1, T3, and T5) (Tables 2-4). Baseline gait speed was 0.900±11.426, which gradually increased during treatment. The mean values at T0, T3, and T5 differed by 0.120 and 0.144, respectively, and were statistically significant (P<0.05). Right stride length was 1.113±0.152 m before treatment. The mean stride length increased over time, reaching a mean of 1.214±0.111 m at T5 compared to baseline, a statistically significant difference (P<0.05). In single-task gait analysis, the average cadence on the left side was 100.934±11.426 at T0, 105.793±9.007 at T1, 106.435±8.262 at T3, and 109.043±10.935 at T5. The average values of T0, T1, T3, and T5 differed by 4.859, 5.501, and 8.109, respectively, and the differences were statistically significant (P<0.05). See Table 4 below.

[0103] Table 4 Comparison of spatiotemporal parameters at different time points in single-task gait analysis

[0104]

[0105]

[0106] During the continuous calculation task, stride length improved bilaterally before and after treatment, with statistically significant differences (P < 0.05). Gait speed gradually increased after the start of treatment, with statistically significant differences at T3 and T5 (P < 0.05). No statistically significant differences were observed in the remaining parameters, including stride width, stride height, and stride frequency (P > 0.05) (Table 5).

[0107] Table 2-5 Comparison of spatiotemporal parameters before and after rTMS treatment under continuous calculation tasks

[0108]

[0109] During the language production task, compared with pre-treatment, bilateral stride length initially increased and then decreased, with statistically significant differences between T0 and T3 (P < 0.05). Gait speed increased with increasing treatment times, with statistically significant differences between T3 and T5 (P < 0.05) (Table 6).

[0110] Table 6 Comparison of spatiotemporal parameters before and after rTMS treatment in language generation task

[0111]

[0112]

[0113] During the hyperventilation task, left stride length at baseline was 1.053 ± 0.100 m. Compared with pretreatment, there were statistically significant differences at T1, T3, and T5 (P < 0.05). Right stride length increased with treatment duration, but only the differences between T0 and T3 were statistically significant (P < 0.05). Compared with baseline T0 gait speed, there were significant differences at T3 and T5 (P < 0.05) (Table 7).

[0114] Table 7 Comparison of spatiotemporal parameters before and after rTMS treatment under hyperventilation task

[0115]

[0116] EEG signal analysis before and after treatment based on variational mode decomposition:

[0117] Feature Extraction: First, variational modal decomposition was performed on the four selected electrodes (Fp1 (frontal lobe), O2 (occipital lobe), P4 (parietal lobe), and T3 (temporal lobe) to obtain multiple modal components. Second, the low-frequency and high-frequency components were selected for fuzzy entropy feature extraction. Box plots show the distribution of fuzzy entropy features before and after rTMS treatment during wakefulness and sleep. The horizontal line in the box plot represents the median, the upper and lower lines represent the first and third quartiles, respectively, and the upper and lower bounds represent the maximum and minimum values in the data. Scattered points are called outliers. Figures 2 to 5 The fuzzy entropy feature extraction results of the low-frequency and high-frequency components of the wakefulness, NREM1, NREM2, and NREM3 phases before and after rTMS treatment, respectively. Q represents the EEG signal before treatment, and H represents the result after treatment.

[0118] By comparing the fuzzy entropy of EEG signals before and after treatment during the awake period, it was found that except for the Fp1 lead, which showed no significant difference in high-frequency and low-frequency components, all other leads showed significant differences (see Figure 2 The fuzzy entropy feature extraction before and after NREM1 treatment found that all leads showed significant differences in high-frequency components, and P4 and T3 leads showed significant differences in the fuzzy entropy of low-frequency components (see Figure 3 The fuzzy entropy of high-frequency and low-frequency components of all leads before and after treatment in NREM2 and NREM3 showed significant differences (see Figure 4 and Figure 5 ).

[0119] Feature selection: In order to further verify whether the extracted features are statistically significant, feature selection is performed through t-test to identify and select features with significant differences and statistical significance, so that they can have a positive impact on the final results and improve the quality of the final analysis results.

[0120] Before and after treatment during the awake phase, the fuzzy entropy values for the low-frequency components of leads P4 and T3 increased (P < 0.05), while the fuzzy entropy values for the high-frequency components of leads O2, P4, and T3 decreased (P < 0.05) (Tables 8 and 9). Before and after treatment during the NREM phase, the fuzzy entropy values for the low-frequency components of leads O2 and P4 decreased, while the fuzzy entropy values for the high-frequency components of leads O2 and T3 increased; these differences were statistically significant (P < 0.05) (Tables 10 and 11). In the NREM phase, except for lead T3, which showed a significant difference in the low-frequency components, the differences in the other leads were not statistically significant (P > 0.05) (Tables 12 and 13). In the NREM phase, except for lead T3, which showed a statistically significant difference in the high-frequency components, the differences in the other leads were statistically significant (P > 0.05) (Tables 14 and 15).

[0121] Table 8 Fuzzy entropy statistical analysis of low-frequency components during the waking period before and after treatment

[0122]

[0123] Table 9 Fuzzy entropy statistical analysis of high-frequency components during the waking period before and after treatment

[0124]

[0125] Table 10 Fuzzy entropy statistical analysis of low-frequency components of sleep stage N1 before and after treatment

[0126]

[0127] Table 11 Fuzzy entropy statistical analysis of high-frequency components of sleep stage N1 before and after treatment

[0128]

[0129] Table 12 Fuzzy entropy statistical analysis of low-frequency components of sleep stage N2 before and after treatment

[0130]

[0131] Table 13 Fuzzy entropy statistical analysis of high-frequency components of sleep stage N2 before and after treatment

[0132]

[0133] Table 14 Fuzzy entropy statistical analysis of low-frequency components of sleep stage N3 before and after treatment

[0134]

[0135] Table 15 Fuzzy entropy statistical analysis of high-frequency components of sleep stage N3 before and after treatment

[0136]

[0137] This specific embodiment is merely an explanation of the present invention and is not intended to limit the present invention. After reading this specification, those skilled in the art may make non-creative modifications to this embodiment as needed. However, as long as such modifications are within the scope of the claims of the present invention, they are protected by patent law.

Claims

1. A system for evaluating the therapeutic effect of epilepsy patients, characterized in that: It includes data acquisition module, treatment module, efficacy evaluation module and EEG signal analysis module; The data collection module is used to collect demographic information and medical history of the subjects. Demographic information includes age, gender, education level and occupation; medical history includes surgical history, allergy history, smoking history, family history and menstrual status of women. The treatment module treats epilepsy by using low-frequency rTMS stimulation; The efficacy evaluation module includes a cognitive function assessment unit and a motor function assessment unit. The cognitive function assessment unit assesses the subjects with the Mini-Mental State Examination, Montreal Cognitive Assessment Scale, and a tethering test before and after rTMS treatment. The motor function evaluation unit uses the Readygo motor function quantitative assessment system to record the subjects' three-dimensional spatiotemporal gait parameters, including temporal parameters of gait speed and cadence, and spatial parameters of stride length, stride height, and stride width. The EEG signal analysis module evaluates the interictal EEG discharges during rTMS treatment of epilepsy and performs EEG signal analysis based on variational mode decomposition.

2. A system for evaluating the therapeutic effect of epilepsy patients according to claim 1, characterized in that: The Mini-Mental State Examination was used to assess global cognitive function, including orientation, immediate word memory, attention and calculation ability, word recall ability, language, and structural imitation ability.

3. The system for evaluating the therapeutic effect of epilepsy patients according to claim 1, characterized in that: The Montreal Cognitive Assessment is used to screen for cognitive impairment, including the cognitive domains of attention and concentration, executive function, memory, language function, visual-spatial function, abstract thinking ability, calculation ability, and orientation.

4. The system for evaluating the therapeutic effect of epilepsy patients according to claim 1, characterized in that: The connection experiment includes connection experiment A and connection experiment B.

5. The system for evaluating the therapeutic effect of epilepsy patients according to claim 1, characterized in that: The sampling rate of the amplifier that collects EEG data in the EEG signal analysis module is 200-1000 Hz, the high-pass filter is ≥70 Hz, the low-pass filter is ≤0.5 Hz, the notch filter is 50 Hz, and the sensitivity is 10 uV / mm.

6. The system for evaluating the therapeutic effect of epilepsy patients according to claim 1, characterized in that: EEG signal analysis based on variational mode decomposition includes the following steps: S1. Preprocessing; S2. Construction of variational modal model; S3. Solution of variational model; S4. Feature extraction using fuzzy entropy algorithm; S5. T-test for feature selection.

Images