An iNPH prediction method based on visual oddball paradigm

Abstract

The present invention relates to the technical field of medical data processing, and discloses an iNPH prediction method based on a visual oddball paradigm, comprising: performing the same visual oddball paradigm experiment on the target population before and after lumbar puncture to obtain electroencephalogram signal data; preprocessing the electroencephalogram signal data to obtain event-related potential features, which are P300 amplitude features; and training an iNPH prediction model based on the event-related potential features. The present invention can obtain more objective electroencephalogram signal data through a visual oddball paradigm experiment, analyze the characteristic changes of the P300 amplitude features of the electroencephalogram signal data, and quantify the degree of improvement of cognitive function. The iNPH prediction model obtained through training can be used to assist doctors in quickly and accurately diagnosing iNPH, so that patients can receive timely and effective treatment, which can reduce the medical burden and bring considerable medical value.

Description

Technical Field

[0001] The present invention relates to the technical field of medical data processing, and in particular to an iNPH prediction method based on a visual oddball paradigm. Background Art

[0002] Idiopathic Normal Pressure Hydrocephalus (iNPH) refers to a syndrome characterized by enlarged ventricles and normal cerebrospinal fluid pressure on imaging, with typical manifestations of gait instability, dementia and urinary incontinence. Excluding secondary factors such as bleeding, trauma, meningitis and intracranial mass, it is more common in the elderly and seriously affects the normal life of patients. iNPH patients can improve the above symptoms through cerebrospinal fluid shunt surgery, thereby reducing the burden on society and family. However, the three main symptoms of iNPH are also common in other elderly diseases such as Alzheimer's Disease (AD), which often causes family members and even doctors to mistakenly understand the above manifestations of iNPH as the result of aging of patients, thereby increasing the difficulty of diagnosis and treatment of iNPH patients. Lumbar Tap Test (LTT) is a common method to distinguish iNPH from other diseases in clinical practice. Before and after LTT, doctors observed the improvement in cognitive function and gait of suspected iNPH patients. When the cognitive function, gait and urinary dysfunction of suspected iNPH patients showed significant improvement after LTT, the patient was clinically diagnosed with iNPH.

[0003] The "Chinese Guidelines for Clinical Management of Idiopathic Normal Pressure Hydrocephalus 2022" points out that the commonly used methods for evaluating cognitive function in clinical practice generally use the Mini-Mental State Examination (MMSE) scale and the Montreal Cognitive Assessment (MoCA) scale. However, the evaluation results based on the scale are easily affected by the patient's subjective factors, especially the patient's education level, personal life experience, and learning effects, resulting in low accuracy of the evaluation results.

[0004] Therefore, there is an urgent need for a method that can assist doctors in accurately diagnosing iNPH. Summary of the invention

[0005] The present invention provides an iNPH prediction method based on the visual oddball paradigm, so as to solve the defect that the prior art cannot assist doctors in accurately diagnosing iNPH.

[0006] The present invention provides a method for constructing an iNPH prediction model based on a visual oddball paradigm, comprising:

[0007] Before the target population undergoes lumbar puncture and drainage, a visual oddball paradigm experiment is conducted on the target population to obtain the target population's standard stimulation EEG signal data before drainage and the target stimulation EEG signal data before drainage;

[0008] After the target population underwent lumbar puncture and drainage, the same visual oddball paradigm experiment was conducted on the target population to obtain the target population's standard stimulation EEG signal data after drainage and the target stimulation EEG signal data after drainage;

[0009] Preprocessing the EEG signal data of standard stimulation before fluid discharge, the EEG signal data of target stimulation before fluid discharge, the EEG signal data of standard stimulation after fluid discharge, and the EEG signal data of target stimulation after fluid discharge, to obtain event-related potential features of standard stimulation before fluid discharge, event-related potential features of target stimulation before fluid discharge, event-related potential features of standard stimulation after fluid discharge, and event-related potential features of target stimulation after fluid discharge, wherein the event-related potential features are P300 amplitude features;

[0010] The iNPH prediction model is trained based on the event-related potential characteristics of standard stimulation before fluid discharge, the event-related potential characteristics of target stimulation before fluid discharge, the event-related potential characteristics of standard stimulation after fluid discharge, and the event-related potential characteristics of target stimulation after fluid discharge;

[0011] Among them, the target population is iNPH patients.

[0012] According to a method for constructing an iNPH prediction model based on a visual oddball paradigm provided by the present invention, the visual oddball paradigm experiment is performed on a target population, comprising:

[0013] Configure EEG caps for the target population;

[0014] Several images are presented to the target group at preset time intervals. In the visual oddball paradigm experiment, the stimulus with a high probability of image appearance is defined as the standard stimulus, and the stimulus with a low probability of image appearance is defined as the target stimulus. The target group is required to remember the number of times the target stimulus appears.

[0015] While presenting images to the target population, the EEG signal data of the target population is obtained through the EEG cap.

[0016] According to a method for constructing an iNPH prediction model based on a visual oddball paradigm provided by the present invention, presenting a plurality of images to a target population at a preset time interval comprises:

[0017] Presenting several rounds of image stimulation to the target population, each round of image stimulation includes a first pattern image stimulation and a second pattern image stimulation, wherein the probability of the first pattern image appearing is 80%, the first pattern image stimulation is a standard stimulation, the probability of the second pattern image appearing is 20%, the second pattern image stimulation is a target stimulation, and there is a preset rest time between each round of image stimulation;

[0018] The first pattern image or the second pattern image is presented to the target population each time according to a preset maintenance time, and there is a preset time interval between each pattern image stimulation.

[0019] According to a method for constructing an iNPH prediction model based on a visual oddball paradigm provided by the present invention, configuring an EEG cap for a target population is specifically as follows:

[0020] An EEG cap was configured for the target population. The cap used standard Ag / AgCl electrodes and the electrodes were placed according to the international 10-20 system. The number of channels was 32 leads. The parameters of the EEG cap were set to a 1000 Hz sampling rate and a 0.1-200 Hz band-pass filter. The acquisition process used the top of the target population's head as a reference, the forehead was grounded, and the impedance between the scalp and the electrodes was kept below 10 KΩ.

[0021] According to a method for constructing an iNPH prediction model based on the visual oddball paradigm provided by the present invention, the preprocessing includes: filtering out invalid data, filtering, downsampling, data segmentation and baseline correction, wherein:

[0022] Filtering out invalid data specifically includes: checking bad data in the EEG signal data and filtering it out;

[0023] The filtering is specifically as follows: using a filter with a filtering range of 0.5 to 30 Hz to filter the EEG signal data;

[0024] Downsampling is specifically as follows: downsampling the EEG signal data to 200 Hz;

[0025] The data segmentation is specifically as follows: defining the start time of each image stimulation as the zero time, and segmenting and intercepting the EEG signal data according to the first preset time window;

[0026] The baseline correction is specifically as follows: using the EEG signal data segment of the second preset time window as a baseline to perform baseline correction on the EEG signal data obtained from each image stimulation.

[0027] According to a method for constructing an iNPH prediction model based on a visual oddball paradigm provided by the present invention, the pre-discharge standard stimulation EEG signal data, the pre-discharge target stimulation EEG signal data, the post-discharge standard stimulation EEG signal data, and the post-discharge target stimulation EEG signal data are pre-processed to obtain event-related potential features of the pre-discharge standard stimulation, event-related potential features of the pre-discharge target stimulation, event-related potential features of the post-discharge standard stimulation, and event-related potential features of the post-discharge target stimulation, including:

[0028] According to the preprocessed EEG signal data, the event-related potential characteristics are obtained by superimposing and averaging the EEG signal data of each trial on each electrode.

[0029] According to a method for constructing an iNPH prediction model based on a visual oddball paradigm provided by the present invention, the event-related potential characteristics of healthy people and iNPH patients are expressed in different forms, wherein the event-related potential characteristics of iNPH patients are expressed in the following forms: standard stimulus event-related potential characteristics before fluid discharge, target stimulus event-related potential characteristics before fluid discharge, standard stimulus event-related potential characteristics after fluid discharge, and target stimulus event-related potential characteristics after fluid discharge, reflecting that the cognitive function of iNPH patients is improved, but has not reached the lowest cognitive function level of healthy people;

[0030] Among them, the iNPH prediction model is trained based on the standard stimulation event-related potential characteristics before fluid discharge, the target stimulation event-related potential characteristics before fluid discharge, the standard stimulation event-related potential characteristics after fluid discharge, and the target stimulation event-related potential characteristics after fluid discharge, specifically:

[0031] The iNPH prediction model was trained based on the change characteristics of the event-related potential characteristics of standard stimulation before fluid discharge, the event-related potential characteristics of target stimulation before fluid discharge, the event-related potential characteristics of standard stimulation after fluid discharge, and the event-related potential characteristics of target stimulation after fluid discharge.

[0032] The present invention also provides a device for constructing an iNPH prediction model based on a visual oddball paradigm, comprising:

[0033] A data receiving module, wherein the data acquisition module is configured to: receive standard stimulation EEG signal data before fluid drainage and target stimulation EEG signal data before fluid drainage obtained by the target population undergoing a visual oddball paradigm experiment before fluid drainage, and acquire standard stimulation EEG signal data after fluid drainage and target stimulation EEG signal data after fluid drainage obtained by the target population undergoing the same visual oddball paradigm experiment after fluid drainage;

[0034] A data processing module, wherein the data processing module is configured to: pre-process the standard stimulation EEG signal data before fluid discharge, the target stimulation EEG signal data before fluid discharge, the standard stimulation EEG signal data after fluid discharge, and the target stimulation EEG signal data after fluid discharge received by the data acquisition module to obtain event-related potential features of the standard stimulation before fluid discharge, event-related potential features of the target stimulation before fluid discharge, event-related potential features of the standard stimulation after fluid discharge, and event-related potential features of the target stimulation after fluid discharge, wherein the event-related potential features are P300 amplitude features;

[0035] A model training module, wherein the model training module is configured to: train and obtain an iNPH prediction model according to the standard stimulation event-related potential characteristics before fluid discharge, the target stimulation event-related potential characteristics before fluid discharge, the standard stimulation event-related potential characteristics after fluid discharge, and the target stimulation event-related potential characteristics after fluid discharge obtained by the data processing module;

[0036] Among them, the target population is iNPH patients.

[0037] The present invention also provides an iNPH prediction device based on a visual oddball paradigm, comprising:

[0038] A data acquisition module, wherein the data acquisition module is configured to: acquire standard stimulation EEG signal data before fluid drainage and target stimulation EEG signal data before fluid drainage obtained by the test subject undergoing a visual oddball paradigm experiment before fluid drainage, and acquire standard stimulation EEG signal data after fluid drainage and target stimulation EEG signal data after fluid drainage obtained by the test subject undergoing the same visual oddball paradigm experiment after fluid drainage;

[0039] An iNPH prediction module, wherein the iNPH prediction module is configured to: input the standard stimulation EEG signal data before fluid discharge, the target stimulation EEG signal data before fluid discharge, the standard stimulation EEG signal data after fluid discharge, and the target stimulation EEG signal data after fluid discharge obtained by the data acquisition module into an iNPH prediction model obtained by any of the above-mentioned methods for constructing an iNPH prediction model based on a visual oddball paradigm, to obtain an iNPH prediction result;

[0040] The test subjects are suspected iNPH patients, iNPH patients to be excluded, or patients with the triad of gait disorder, urinary incontinence and cognitive impairment.

[0041] The present invention also provides a brain-computer interface system for predicting iNPH, comprising:

[0042] An EEG cap, which is worn on the head of the tester to obtain the EEG signal of the tester;

[0043] An EEG acquisition device, configured to be connected to the electrodes of the EEG cap to acquire the EEG signal of the tester obtained by the EEG cap;

[0044] A computer, comprising a communication module, a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer is configured to communicate with the EEG acquisition device to acquire the EEG signal of the tester, and the processor is configured to implement the following steps when executing the computer program:

[0045] Obtaining the standard stimulation EEG signal data before lumbar puncture and the target stimulation EEG signal data before lumbar puncture obtained by the test subject undergoing a visual oddball paradigm experiment before lumbar puncture, and obtaining the standard stimulation EEG signal data after lumbar puncture and the target stimulation EEG signal data after lumbar puncture obtained by the test subject undergoing the same visual oddball paradigm experiment;

[0046] Preprocessing the standard stimulation EEG signal data before fluid discharge, the target stimulation EEG signal data before fluid discharge, the standard stimulation EEG signal data after fluid discharge, and the target stimulation EEG signal data after fluid discharge;

[0047] Input the pre-processed standard stimulation EEG signal data before fluid discharge, target stimulation EEG signal data before fluid discharge, standard stimulation EEG signal data after fluid discharge, and target stimulation EEG signal data after fluid discharge into the iNPH prediction model obtained by the construction method of the iNPH prediction model based on the visual oddball paradigm described in any one of the above items to obtain a prediction result of whether the test subject is an iNPH patient;

[0048] The test subjects are suspected iNPH patients, iNPH patients to be excluded, or patients with the triad of gait disorder, urinary incontinence and cognitive impairment.

[0049] The present invention provides an iNPH prediction method based on a visual oddball paradigm, designs a visual oddball paradigm experiment that is not easily affected by subjective factors such as the cultural level and life experience of patients, and performs a visual oddball paradigm experiment on a target population before and after lumbar puncture to obtain more objective electroencephalogram signal data. By analyzing the characteristic changes of event-related potential characteristics (P300 amplitude characteristics) of the electroencephalogram signal data, the degree of improvement in cognitive function can be quantified. The iNPH prediction model trained accordingly can be used to assist doctors in quickly and accurately diagnosing iNPH, so that patients can receive effective treatment in a timely manner, which can reduce the medical burden and bring considerable medical value. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0051] Figure 1 A schematic flow chart of a method for constructing an iNPH prediction model based on a visual oddball paradigm provided by the present invention.

[0052] Figure 2 Shown is the 10-20 international standard electrode distribution.

[0053] Figure 3 An example of a Visual Oddball paradigm experiment is shown.

[0054] Figure 4 Shows the time domain waveform of healthy people in the visual oddball paradigm experiment.

[0055] Figure 5(a)-(c) shows the time domain waveforms of the iNPH population in the visual oddball paradigm experiment.

[0056] Figure 6 A schematic structural diagram of a device for constructing an iNPH prediction model based on a visual oddball paradigm provided by the present invention.

[0057] Figure 7 This is a schematic structural diagram of an electronic device provided by the present invention.

[0058] in, Figure 4-Figure 5(c) In the table, time / ms indicates time / millisecond, amplitude indicates amplitude, Fz, Cz, and Pz indicate midline electrodes, standard indicates standard stimulation, target indicates target stimulation, TTbefore indicates before lumbar puncture, and TT24 indicates 24 hours after lumbar puncture. DETAILED DESCRIPTION

[0059] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings in the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments, and they should not be understood as limitations on the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. In the description of the present invention, it should be understood that the terms used are only for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0060] Combine the following Figure 1-Figure 7 The present invention describes an iNPH prediction method based on the visual oddball paradigm.

[0061] Figure 1 Schematic diagram of the process of constructing the iNPH prediction model based on the visual oddball paradigm provided by the present invention. Figure 1 The present invention provides a method for constructing an iNPH prediction model based on a visual oddball paradigm, which may include:

[0062] Step S110, before the target population undergoes lumbar puncture and drainage, a visual oddball paradigm experiment is performed on the target population to obtain standard stimulation EEG signal data and target stimulation EEG signal data before drainage of the target population;

[0063] Step S120, after the target population undergoes lumbar puncture and fluid drainage, the same visual oddball paradigm experiment is performed on the target population to obtain standard stimulation EEG signal data after fluid drainage and target stimulation EEG signal data after fluid drainage of the target population;

[0064] Step S130, pre-processing the EEG signal data of standard stimulation before fluid discharge, the EEG signal data of target stimulation before fluid discharge, the EEG signal data of standard stimulation after fluid discharge, and the EEG signal data of target stimulation after fluid discharge, to obtain event-related potential features of standard stimulation before fluid discharge, event-related potential features of target stimulation before fluid discharge, event-related potential features of standard stimulation after fluid discharge, and event-related potential features of target stimulation after fluid discharge (Event-Related Potentials, ERP features), wherein the event-related potential feature is the P300 amplitude feature, which refers to the amplitude feature around 300 milliseconds after the stimulation occurs;

[0065] Step S140, training an iNPH prediction model based on the event-related potential characteristics of standard stimulation before fluid discharge, the event-related potential characteristics of target stimulation before fluid discharge, the event-related potential characteristics of standard stimulation after fluid discharge, and the event-related potential characteristics of target stimulation after fluid discharge.

[0066] It should be noted that the target population is iNPH patients.

[0067] It should be noted that the oddball paradigm refers to the random presentation of two stimuli acting on the same sensory channel (such as vision) in the experiment, in which the stimulus with a high probability of occurrence is called the standard stimulus, and the stimulus with a low probability of occurrence is called the target stimulus. The oddball paradigm experiment is specifically designed to induce the P300 amplitude characteristics.

[0068] It should be noted that the brain computer interface (BCI) is a technology for direct communication between the human brain and output devices. Electroencephalography (EEG) is a commonly used method of BCI due to its advantages of being non-invasive, easy to use, and low cost. This application implements BCI technology based on EEG.

[0069] In one embodiment, the performing of a visual oddball paradigm experiment on a target population includes:

[0070] Configure EEG caps for the target population;

[0071] Several images are presented to the target group at preset time intervals. In the visual oddball paradigm experiment, the stimulus with a high probability of image appearance is defined as the standard stimulus, and the stimulus with a low probability of image appearance is defined as the target stimulus. The target group is required to remember the number of times the target stimulus appears and finally provide feedback on the number of times.

[0072] While presenting images to the target population, the EEG signal data of the target population is obtained through the EEG cap.

[0073] Specifically, refer to Figure 2 , the target population was equipped with an EEG cap, which used standard Ag / AgCl electrodes, and placed electrodes according to the international 10-20 system, with 32 channels. The parameters of the EEG cap were set to 1000Hz sampling rate and 0.1-200Hz bandpass filter, and a 50Hz notch filter was used to filter out power frequency interference. The acquisition process used the top of the target population's head as a reference, the forehead was grounded, and the impedance between the scalp and the electrode was kept below 10KΩ. During the experiment, the target population was required to remain as still as possible, avoid random eye movements and subtle movements unrelated to the task, to ensure the reliability of the collected EEG signal data.

[0074] Specifically, the EEG signal data can be obtained by electrically connecting the SynAmps2 EEG amplifier to the EEG cap.

[0075] In one embodiment, presenting a plurality of images to a target group at a preset time interval includes:

[0076] Presenting several rounds of image stimulation to the target population, each round of image stimulation includes a first pattern image stimulation and a second pattern image stimulation, wherein the probability of the first pattern image appearing is 80%, the first pattern image stimulation is a standard stimulation, the probability of the second pattern image appearing is 20%, the second pattern image stimulation is a target stimulation, and there is a preset rest time between each round of image stimulation;

[0077] The first pattern image or the second pattern image is presented to the target population each time according to a preset maintenance time, and there is a preset time interval between each pattern image stimulation.

[0078] Specifically, in this embodiment, the same visual oddball paradigm experiment is conducted on the target population once at the same time before and after lumbar puncture (for example, 24 hours before and after lumbar puncture), each time including two rounds of image stimulation, i.e., two blocks. Between the two blocks, a rest period of 5 to 10 minutes (preset rest time) can be allowed depending on the state of the target population. During the experiment, images can be presented to the target population through a display, and the resolution of the display can be 1920×1080, with a refresh rate of 60 Hz.

[0079] The visual oddball paradigm used in this embodiment is as follows Figure 3 As shown, the circular pattern image is used as the first pattern image, and the five-pointed star pattern image is used as the second pattern image. Each block contains 150 randomly presented trials, of which 120 trials (80%) are standard stimuli and 30 trials (20%) are target stimuli. The preset maintenance time of each trial is 500ms, and the preset time interval is 1500±100ms. Before the experiment begins, the target population is required to memorize the number of times the target stimulus appears and finally feedback the number of times information to maintain their focus on the experiment. The experimental results show that due to the low level of cognitive function of iNPH, the accuracy of iNPH feedback on the number of times the target stimulus appears is much lower than that of healthy people.

[0080] In one embodiment, the preprocessing includes: filtering out invalid data, filtering, downsampling, data segmentation, and baseline correction.

[0081] Specifically, the EEGLAB open source toolkit can be used to pre-process the standard stimulation EEG signal data before fluid discharge, the target stimulation EEG signal data before fluid discharge, the standard stimulation EEG signal data after fluid discharge, and the target stimulation EEG signal data after fluid discharge to ensure the accuracy of the data.

[0082] Specifically, the invalid data can be filtered out by importing the EEG signal data into the EEGLAB toolkit for data preview. Information such as the number of leads, total length of the data, number of epochs of the data, number of events included, and data sampling rate of the EEG signal data can be observed. The data can be preliminarily understood and checked through data preview, so as to dig out bad lead data or other invalid data (that is, information that is obviously inconsistent with the experimental design) in the EEG signal data and filter them out.

[0083] Filtering is specifically as follows: EEG signals are weak and easily interfered by various noises, including physiological noises such as heartbeat and electromyography, as well as external factors such as acquisition equipment. Since this application mainly studies the P300 amplitude characteristics, this embodiment uses a third-order Butterworth filter with a filtering range of 0.5 to 30 Hz to filter the EEG signal data.

[0084] Downsampling is specifically as follows: commonly used EEG signal data are mainly concentrated within 100 Hz. To improve the operation efficiency, under the premise of ensuring that the signal is not distorted, this embodiment downsamples the EEG signal data to 200 Hz.

[0085] The data segmentation is as follows: the collected EEG signal data includes the entire experimental process, while the actual application uses the response signal generated after the paradigm stimulation. Therefore, the start time of each image stimulation is defined as time zero, and the EEG signal data is segmented and intercepted according to the first preset time window (in this embodiment, [-0.2, 0.8] seconds).

[0086] The baseline correction is specifically as follows: the EEG signal data segments of the second preset time window (in this embodiment, [-0.2, 0] seconds) are used as the baseline to perform baseline correction on the EEG signal data obtained from each image stimulation (the EEG signal data segments segmented according to the first preset time window) to prevent the impact of data drift.

[0087] In one embodiment, the pre-discharging standard stimulation EEG signal data, the pre-discharging target stimulation EEG signal data, the post-discharging standard stimulation EEG signal data, and the post-discharging target stimulation EEG signal data are pre-processed to obtain the pre-discharging standard stimulation event-related potential feature, the pre-discharging target stimulation event-related potential feature, the post-discharging standard stimulation event-related potential feature, and the post-discharging target stimulation event-related potential feature, including:

[0088] According to the preprocessed EEG signal data, the event-related potential characteristics are obtained by superimposing and averaging the EEG signal data of each trial on each electrode.

[0089] The purpose of superimposing and averaging EEG signal data is to reduce noise so that it is easier to observe brain activity related to events. The characteristics of ERP are time-locked and phase-locked, that is, the latency is constant and the waveform is constant. The collected single-trial EEG signal data contains brain activity related to the event (ie, ERP), as well as other activities unrelated to the event (for example, skin potential, electromyographic artifacts, eye artifacts, and induced noise, etc.). These signals unrelated to the event are randomly changing, with highs and lows. Therefore, by superimposing and averaging multiple segmented EEG signal data, the random irrelevant positive and negative signals can be offset each other, and finally the accurate P300 amplitude characteristics can be extracted to form an ERP waveform (also known as a time domain waveform).

[0090] In one embodiment, the event-related potential characteristics of healthy people and iNPH patients are expressed differently, wherein:

[0091] The manifestations of the event-related potential characteristics of the healthy population are specifically as follows: the event-related potential characteristics of standard stimulation before fluid discharge, the event-related potential characteristics of target stimulation before fluid discharge, the event-related potential characteristics of standard stimulation after fluid discharge, and the event-related potential characteristics of target stimulation after fluid discharge, which reflect that the cognitive function of the healthy population has not changed;

[0092] The event-related potential characteristics of iNPH patients are specifically manifested in the following forms: standard stimulation event-related potential characteristics before fluid discharge, target stimulation event-related potential characteristics before fluid discharge, standard stimulation event-related potential characteristics after fluid discharge, and target stimulation event-related potential characteristics after fluid discharge. This reflects that the cognitive function of iNPH patients has improved, but has not reached the lowest cognitive function level of healthy people.

[0093] Among them, the iNPH prediction model is trained based on the standard stimulation event-related potential characteristics before fluid discharge, the target stimulation event-related potential characteristics before fluid discharge, the standard stimulation event-related potential characteristics after fluid discharge, and the target stimulation event-related potential characteristics after fluid discharge, specifically:

[0094] The iNPH prediction model was trained based on the change characteristics of the event-related potential characteristics of standard stimulation before fluid discharge, the event-related potential characteristics of target stimulation before fluid discharge, the event-related potential characteristics of standard stimulation after fluid discharge, and the event-related potential characteristics of target stimulation after fluid discharge.

[0095] The main scalp distribution of P300 amplitude characteristics was at the midline electrodes (Fz, Cz, and Pz), and its amplitude generally increased from the frontal to the parietal electrodes.

[0096] See also Figure 4 , Figure 4 The results of the visual oddball paradigm experiment of healthy people are shown. Target (solid line) is the ERP waveform under target stimulation, and standard (dashed line) is the ERP waveform under standard stimulation. Compared with standard stimulation, target stimulation can induce a larger P300 amplitude. This is a change in the event-related potential characteristics of healthy people. It does not require lumbar puncture and drainage. The visual oddball paradigm experiment can be performed on healthy people at the same time before and after lumbar puncture and drainage of iNPH patients. It will be found that the results of the two visual oddball paradigm experiments in healthy people are consistent, and there is no change in cognitive function. Even if there may be some differences, they are negligible changes.

[0097] See Figure 5, which shows the results of the visual oddball paradigm experiment of the iNPH population. Figures 5(a) and 5(b) are the time domain waveforms of the visual oddball paradigm experiment before and after lumbar puncture, respectively. Target (solid line) is the ERP waveform under target stimulation, and standard (dashed line) is the ERP waveform under standard stimulation. Compared with standard stimulation, target stimulation induces a larger P300 amplitude, but the P300 amplitude is still lower than that of healthy people, and the difference between target stimulation and standard stimulation is also small. Figure 5(c) is a comparison of the time domain waveforms of the target stimulation in the visual oddball paradigm experiment before and after lumbar puncture. Target (solid line) is the ERP waveform of the target stimulation after lumbar puncture, and standard (dashed line) is the ERP waveform of the target stimulation before lumbar puncture. It can be seen that the P300 amplitude after lumbar puncture is larger than the P300 amplitude before lumbar puncture. From this, it can be concluded that the P300 amplitude characteristics of iNPH patients changed significantly before and after lumbar puncture, but the effect did not reach the lowest cognitive function level of healthy people (that is, even if iNPH patients underwent lumbar puncture and fluid drainage to improve their cognitive function, their cognitive function level was still worse than that of the people with the lowest cognitive function level in the healthy population). This shows that lumbar puncture and fluid drainage improved the cognitive function level of iNPH patients, but it still did not reach the lowest cognitive function level of healthy people.

[0098] Therefore, the changes in P300 amplitude characteristics can be analyzed and learned based on the EEG signal data of healthy people and iNPH under the visual oddball paradigm stimulation, and the iNPH prediction model can be trained to assist doctors in diagnosing suspected iNPH patients. It should be noted that the iNPH prediction model can adopt any model framework that meets the requirements in the existing technology, such as a binary classification model.

[0099] When assisting doctors in diagnosing suspected iNPH patients, first, before the tester undergoes lumbar puncture and fluid drainage, a visual oddball paradigm experiment is performed on the tester to obtain the tester's standard stimulation EEG signal data before drainage and the tester's target stimulation EEG signal data before drainage; then, after the tester undergoes lumbar puncture and fluid drainage, the same visual oddball paradigm experiment is performed on the tester to obtain the tester's standard stimulation EEG signal data after drainage and the tester's target stimulation EEG signal data after drainage; then, the standard stimulation EEG signal data before drainage, the target stimulation EEG signal data before drainage, the standard stimulation EEG signal data after drainage and the target stimulation EEG signal data after drainage are preprocessed to obtain the standard stimulation event phase before drainage. off potential characteristics, event-related potential characteristics of target stimulation before fluid discharge, event-related potential characteristics of standard stimulation after fluid discharge and event-related potential characteristics of target stimulation after fluid discharge, among which the event-related potential characteristics are P300 amplitude characteristics; finally, according to the event-related potential characteristics of standard stimulation before fluid discharge, event-related potential characteristics of target stimulation before fluid discharge, event-related potential characteristics of standard stimulation after fluid discharge and event-related potential characteristics of target stimulation after fluid discharge, iNPH prediction is performed by the iNPH prediction model obtained by the above-mentioned method of constructing the iNPH prediction model based on the visual oddball paradigm. Doctors can further determine the diagnosis results according to the model prediction results and the actual situation to improve the accuracy of the diagnosis results.

[0100] The present invention provides an iNPH prediction method based on a visual oddball paradigm, designs a visual oddball paradigm experiment that is not easily affected by subjective factors such as the cultural level and life experience of patients, and the target population passively performs the visual oddball paradigm experiment before and after lumbar puncture and drainage, effectively inducing ERP features, and being able to obtain more objective electroencephalogram signal data. By analyzing the characteristic changes of event-related potential features (P300 amplitude features) of the electroencephalogram signal data, the degree of improvement in cognitive function can be quantified, and the iNPH prediction model obtained by training can be used to assist doctors in quickly and accurately diagnosing iNPH, so that patients can receive effective treatment in a timely manner, and the effectiveness of shunt surgery in improving iNPH symptoms can be determined, which can reduce the medical burden and bring considerable medical value.

[0101] The following is a description of a construction device for an iNPH prediction model based on a visual oddball paradigm provided by the present invention. The construction device for an iNPH prediction model based on a visual oddball paradigm described below and the construction method for an iNPH prediction model based on a visual oddball paradigm described above can be referred to each other.

[0102] Reference Figure 6 The present invention provides a device for constructing an iNPH prediction model based on a visual oddball paradigm, which may include:

[0103] A data receiving module, wherein the data acquisition module is configured to: receive standard stimulation EEG signal data before fluid drainage and target stimulation EEG signal data before fluid drainage obtained by the target population undergoing a visual oddball paradigm experiment before fluid drainage, and acquire standard stimulation EEG signal data after fluid drainage and target stimulation EEG signal data after fluid drainage obtained by the target population undergoing the same visual oddball paradigm experiment after fluid drainage;

[0104] A data processing module, wherein the data processing module is configured to: pre-process the standard stimulation EEG signal data before fluid discharge, the target stimulation EEG signal data before fluid discharge, the standard stimulation EEG signal data after fluid discharge, and the target stimulation EEG signal data after fluid discharge received by the data acquisition module to obtain event-related potential features of the standard stimulation before fluid discharge, event-related potential features of the target stimulation before fluid discharge, event-related potential features of the standard stimulation after fluid discharge, and event-related potential features of the target stimulation after fluid discharge, wherein the event-related potential features are P300 amplitude features;

[0105] A model training module, wherein the model training module is configured to: train and obtain an iNPH prediction model according to the standard stimulation event-related potential characteristics before fluid discharge, the target stimulation event-related potential characteristics before fluid discharge, the standard stimulation event-related potential characteristics after fluid discharge, and the target stimulation event-related potential characteristics after fluid discharge obtained by the data processing module;

[0106] Among them, the target population is iNPH patients.

[0107] The present invention also provides an iNPH prediction device based on a visual oddball paradigm, comprising:

[0108] A data acquisition module, wherein the data acquisition module is configured to: acquire standard stimulation EEG signal data before fluid drainage and target stimulation EEG signal data before fluid drainage obtained by the test subject undergoing a visual oddball paradigm experiment before fluid drainage, and acquire standard stimulation EEG signal data after fluid drainage and target stimulation EEG signal data after fluid drainage obtained by the test subject undergoing the same visual oddball paradigm experiment after fluid drainage;

[0109] An iNPH prediction module, wherein the iNPH prediction module is configured to: input the standard stimulation EEG signal data before fluid discharge, the target stimulation EEG signal data before fluid discharge, the standard stimulation EEG signal data after fluid discharge and the target stimulation EEG signal data after fluid discharge obtained by the data acquisition module into the iNPH prediction model obtained by the construction method of the iNPH prediction model based on the visual oddball paradigm described in any one of the above items, to obtain an iNPH prediction result.

[0110] Figure 7An example of a physical structure diagram of an electronic device is shown in FIG. Figure 7 As shown, the electronic device may include: a processor 810, a communication interface 820, a memory 830 and a communication bus 840, wherein the processor 810, the communication interface 820 and the memory 830 communicate with each other through the communication bus 840. The processor 810 may call the logic instructions in the memory 830 to perform the following steps:

[0111] Obtaining the standard stimulation EEG signal data before lumbar puncture and the target stimulation EEG signal data before lumbar puncture obtained by the test subject undergoing a visual oddball paradigm experiment before lumbar puncture, and obtaining the standard stimulation EEG signal data after lumbar puncture and the target stimulation EEG signal data after lumbar puncture obtained by the test subject undergoing the same visual oddball paradigm experiment;

[0112] Preprocessing the standard stimulation EEG signal data before fluid discharge, the target stimulation EEG signal data before fluid discharge, the standard stimulation EEG signal data after fluid discharge, and the target stimulation EEG signal data after fluid discharge;

[0113] Input the pre-processed standard stimulation EEG signal data before fluid discharge, target stimulation EEG signal data before fluid discharge, standard stimulation EEG signal data after fluid discharge, and target stimulation EEG signal data after fluid discharge into the iNPH prediction model obtained by the construction method of the iNPH prediction model based on the visual oddball paradigm described in any one of the above items to obtain a prediction result of whether the test subject is an iNPH patient;

[0114] The test subjects are suspected iNPH patients, iNPH patients to be excluded, or patients with the triad of gait disorder, urinary incontinence and cognitive impairment.

[0115] In addition, the logic instructions in the above-mentioned memory 830 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0116] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0117] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0118] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for constructing an iNPH prediction model based on the visual oddball paradigm, characterized in that: include: Before the target population undergoes lumbar puncture and drainage, a visual oddball paradigm experiment is conducted on the target population to obtain the target stimulation EEG signal data of the target population before drainage; After the target population underwent lumbar puncture and drainage, the same visual oddball paradigm experiment was conducted on the target population to obtain the target stimulation EEG signal data of the target population after drainage; Preprocessing the EEG signal data of the target stimulation before and after the fluid discharge to obtain the event-related potential features of the target stimulation before and after the fluid discharge, wherein the event-related potential features are P300 amplitude features; According to the event-related potential characteristics of target stimulation before and after fluid release, the iNPH prediction model is trained; Among them, the target population is iNPH patients; Among them, the target stimulation event-related potential characteristics of the target population are specifically manifested as follows: the target stimulation event-related potential characteristics after fluid release reflect that the cognitive function of iNPH patients has improved, but has not reached the lowest cognitive function level of healthy people; The iNPH prediction model is trained based on the event-related potential characteristics of the target stimulation before and after the discharge, specifically: According to the changing characteristics of the event-related potential characteristics of the target stimulation before and after the fluid discharge, the iNPH prediction model was trained.

2. The method for constructing an iNPH prediction model based on the visual oddball paradigm according to claim 1, characterized in that: The visual oddball paradigm experiment conducted on the target population includes: Configure EEG caps for the target population; Several images are presented to the target group at preset time intervals. In the visual oddball paradigm experiment, the stimulus with a high probability of image appearance is defined as the standard stimulus, and the stimulus with a low probability of image appearance is defined as the target stimulus. The target group is required to remember the number of times the target stimulus appears. While presenting images to the target population, the EEG signal data of the target population is obtained through the EEG cap.

3. The method for constructing an iNPH prediction model based on the visual oddball paradigm according to claim 2, characterized in that: The presenting of a plurality of images to the target group at a preset time interval includes: Presenting several rounds of image stimulation to the target population, each round of image stimulation includes a first pattern image stimulation and a second pattern image stimulation, wherein the probability of the first pattern image appearing is 80%, the first pattern image stimulation is a standard stimulation, the probability of the second pattern image appearing is 20%, the second pattern image stimulation is a target stimulation, and there is a preset rest time between each round of image stimulation; The first pattern image or the second pattern image is presented to the target population each time according to a preset maintenance time, and there is a preset time interval between each pattern image stimulation.

4. The method for constructing an iNPH prediction model based on the visual oddball paradigm according to claim 3, characterized in that: The EEG cap is configured for the target population, specifically: An EEG cap was configured for the target population. The cap used standard Ag / AgCl electrodes and the electrodes were placed according to the international 10-20 system. The number of channels was 32 leads. The parameters of the EEG cap were set to a 1000 Hz sampling rate and a 0.1-200 Hz band-pass filter. The acquisition process used the top of the target population's head as a reference, the forehead was grounded, and the impedance between the scalp and the electrodes was kept below 10 KΩ.

5. The method for constructing an iNPH prediction model based on the visual oddball paradigm according to claim 4, characterized in that: The preprocessing includes: filtering out invalid data, filtering, downsampling, data segmentation and baseline correction, wherein: Filtering out invalid data specifically includes: checking bad data in the EEG signal data and filtering it out; The filtering is specifically as follows: using a filter with a filtering range of 0.5 to 30 Hz to filter the EEG signal data; Downsampling is specifically as follows: downsampling the EEG signal data to 200 Hz; The data segmentation is specifically as follows: defining the start time of each image stimulation as the zero time, and segmenting and intercepting the EEG signal data according to the first preset time window; The baseline correction is specifically as follows: using the EEG signal data segment of the second preset time window as a baseline to perform baseline correction on the EEG signal data obtained from each image stimulation.

6. The method for constructing an iNPH prediction model based on the visual oddball paradigm according to claim 5, characterized in that: The pre-fluid discharge target stimulation EEG signal data and the post-fluid discharge target stimulation EEG signal data are pre-processed to obtain the pre-fluid discharge target stimulation event-related potential feature and the post-fluid discharge target stimulation event-related potential feature, including: According to the preprocessed EEG signal data, the event-related potential characteristics are obtained by superimposing and averaging the EEG signal data of each trial on each electrode.

7. A device for constructing an iNPH prediction model based on a visual oddball paradigm, characterized in that: include: A data receiving module, wherein the data receiving module is configured to: receive the target stimulation EEG signal data before lumbar puncture and fluid drainage obtained by the target population undergoing a visual oddball paradigm experiment, and obtain the target stimulation EEG signal data after lumbar puncture and fluid drainage obtained by the target population undergoing the same visual oddball paradigm experiment; A data processing module, wherein the data processing module is configured to: pre-process the EEG signal data of the target stimulation before and after the fluid discharge received by the data receiving module to obtain event-related potential features of the target stimulation before and after the fluid discharge, wherein the event-related potential features are P300 amplitude features; A model training module, wherein the model training module is configured to: train and obtain an iNPH prediction model according to the event-related potential characteristics of the target stimulation before and after the fluid discharge obtained by the data processing module; Among them, the target population is iNPH patients; The target population's target stimulation event-related potential characteristics were specifically expressed as follows: the target stimulation event-related potential characteristics after fluid release reflected that the cognitive function of iNPH patients was improved, but did not reach the lowest cognitive function level of healthy people; Among them, according to the event-related potential characteristics of the target stimulation before and after the discharge, the iNPH prediction model is trained, specifically: According to the changing characteristics of the event-related potential characteristics of the target stimulation before and after the fluid discharge, the iNPH prediction model was trained.

8. An iNPH prediction device based on the visual oddball paradigm, characterized in that: include: A data acquisition module, wherein the data acquisition module is configured to: acquire the EEG signal data of the target stimulation before the lumbar puncture and fluid drainage obtained by the test subject undergoing a visual oddball paradigm experiment before the lumbar puncture and fluid drainage, and acquire the EEG signal data of the target stimulation after the lumbar puncture and fluid drainage obtained by the test subject undergoing the same visual oddball paradigm experiment; An iNPH prediction module, wherein the iNPH prediction module is configured to: input the target stimulation EEG signal data before and after the fluid release obtained by the data acquisition module into the iNPH prediction model obtained by the method for constructing an iNPH prediction model based on the visual oddball paradigm according to any one of claims 1 to 6, to obtain an iNPH prediction result.

9. A brain-computer interface system for predicting iNPH, characterized in that: include: An EEG cap, which is worn on the head of the tester to obtain the EEG signal of the tester; An EEG acquisition device, configured to be connected to the electrodes of the EEG cap to acquire the EEG signal of the tester obtained by the EEG cap; A computer, comprising a communication module, a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer is configured to communicate with the EEG acquisition device to acquire the EEG signal of the tester, and the processor is configured to implement the following steps when executing the computer program: Obtaining the EEG signal data of the target stimulation before the lumbar puncture and the EEG signal data of the target stimulation after ... Preprocessing the target stimulation EEG signal data before and after the fluid discharge; The pre-processed target stimulation EEG signal data before and after fluid discharge are input into the iNPH prediction model obtained by the method for constructing an iNPH prediction model based on the visual oddball paradigm according to any one of claims 1 to 6 to obtain a prediction result of whether the test subject is an iNPH patient.

Images