CN117077013B - Sleep spindle wave detection method, electronic equipment and medium - Google Patents
Sleep spindle wave detection method, electronic equipment and medium Download PDFInfo
- Publication number
- CN117077013B CN117077013B CN202311315398.3A CN202311315398A CN117077013B CN 117077013 B CN117077013 B CN 117077013B CN 202311315398 A CN202311315398 A CN 202311315398A CN 117077013 B CN117077013 B CN 117077013B
- Authority
- CN
- China
- Prior art keywords
- eeg signal
- eeg
- spindle wave
- sleep
- sleep spindle
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 41
- 239000012634 fragment Substances 0.000 claims abstract description 19
- 238000012216 screening Methods 0.000 claims abstract description 8
- 238000000034 method Methods 0.000 claims description 51
- 238000000605 extraction Methods 0.000 claims description 43
- 230000015654 memory Effects 0.000 claims description 33
- 238000012549 training Methods 0.000 claims description 32
- 230000006870 function Effects 0.000 claims description 24
- 239000013598 vector Substances 0.000 claims description 20
- 230000008569 process Effects 0.000 claims description 16
- 238000004590 computer program Methods 0.000 claims description 14
- 238000013145 classification model Methods 0.000 claims description 12
- 238000011176 pooling Methods 0.000 claims description 7
- 238000001914 filtration Methods 0.000 claims description 5
- 230000006798 recombination Effects 0.000 claims description 5
- 238000005215 recombination Methods 0.000 claims description 5
- 238000010606 normalization Methods 0.000 claims description 4
- 238000005070 sampling Methods 0.000 claims description 3
- 230000006403 short-term memory Effects 0.000 claims description 3
- 238000002372 labelling Methods 0.000 claims 4
- 230000007423 decrease Effects 0.000 claims 2
- 230000007787 long-term memory Effects 0.000 claims 1
- 238000010586 diagram Methods 0.000 description 15
- 230000006872 improvement Effects 0.000 description 9
- 238000012545 processing Methods 0.000 description 8
- 230000000694 effects Effects 0.000 description 6
- 238000011160 research Methods 0.000 description 6
- 238000004458 analytical method Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 208000019116 sleep disease Diseases 0.000 description 4
- 230000007246 mechanism Effects 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 238000012805 post-processing Methods 0.000 description 3
- 238000011282 treatment Methods 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000003745 diagnosis Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- 208000022925 sleep disturbance Diseases 0.000 description 2
- 238000012800 visualization Methods 0.000 description 2
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 1
- 241000282412 Homo Species 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000003920 cognitive function Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 230000005055 memory storage Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 229920001296 polysiloxane Polymers 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/369—Electroencephalography [EEG]
- A61B5/372—Analysis of electroencephalograms
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
- A61B5/7267—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems involving training the classification device
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/211—Selection of the most significant subset of features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/09—Supervised learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
- G06F2218/10—Feature extraction by analysing the shape of a waveform, e.g. extracting parameters relating to peaks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
Abstract
The invention discloses a sleep spindle wave detection method, electronic equipment and medium, which comprise the following steps: intercepting an EEG signal to be detected into a plurality of EEG signal fragments with continuous time sequences and equal length, and extracting EEG characteristics; dividing the EEG signal segments into a first type of EEG signal segments and a second type of EEG signal segments according to the EEG characteristics; taking EEG signal segments with the sleep spindle wave duty ratio larger than a first threshold value as first class EEG signal segments; recombining each of the first class of EEG signal segments with its temporally adjacent two EEG signal segments, respectively; inputting the recombined EEG signal segment into a sleep spindle wave prediction model to obtain the peak position of the sleep spindle wave; when the distance between the adjacent peaks is smaller than the second threshold value, the signals between the adjacent peaks are also regarded as sleep spindle wave signals; and screening according to the length range of the sleep spindle wave to obtain the prediction result of the sleep spindle wave signal.
Description
Technical Field
The invention relates to the field of Electroencephalogram (EEG) processing, in particular to a sleep spindle wave detection method, electronic equipment and medium.
Background
Sleep spindle wave (sleep spindle) is a phenomenon of electroencephalogram discharge that is often seen during human sleep. Typically such brain waves will be observed at electrodes at frequencies of 12 to 14 hertz (Hz) for a duration of about 0.5 seconds to 2 seconds, with amplitudes typically between 200-300 microvolts. Numerous studies have demonstrated that sleep spindles exhibit their importance through effects on human learning and memory formation. The presence of sleep spindle wave density may be indicative of a signal of enhanced memory, such as an improvement in memory in terms of detail and spatial structure. Thus, sleep spindles can be used as tools for assessing cognitive functions and memory capabilities of individuals, and how to utilize sleep spindles to interfere with the learning and memory processes of humans is a problem that many researchers are struggling to solve. Furthermore, sleep spindles are also widely used in the medical field. By measuring and analyzing sleep spindles, a number of useful information can be obtained to determine whether the patient is in a normal sleep state and to determine the type of sleep disturbance. Currently, sleep disorders have become a problem afflicting many people, and detection of sleep spindles has an important role in diagnosis and treatment of sleep disorders. Many studies have shown that measurement and analysis of sleep spindles can help hospitalized patients better determine the type and extent of sleep disturbance and administer corresponding treatments as the case may be. The monitoring of sleep spindle waves can also provide important experimental data for the field of sleep research, and help scientists to further understand the nature and mechanism of human sleep deeply. Measurement and analysis of sleep spindle waves not only can provide important reference data for research and treatment of sleep, but also can be an important way for human to understand and recognize sleep mechanisms. Therefore, the development of sleep spindle wave detection research has important significance for understanding the mechanism of human memory and sleep related diseases.
The rapid and accurate detection of spindle wave signals from sleep EEG can provide powerful support for subsequent research and analysis. Conventional detection methods generally determine the occurrence time and frequency of spindle waves based on artificial visual judgment, but such methods have problems of poor subjectivity and repeatability. In order to improve the automation degree of sleep spindle wave detection, the labor cost is reduced. Researchers have focused on various automated detection methods for sleep spindle waves. Such as time-frequency analysis (TFA) methods, which convert the signal into a time-frequency diagram and automatically identify the spindle wave based on the characteristics and a priori knowledge of the signal in the time-frequency diagram. However, such a detection method based on the time-frequency characteristic of the signal often needs to determine whether the signal includes the spindle wave by presetting a threshold value, where setting the threshold value affects the accuracy of spindle wave detection and has poor universality on different data sets. Another type of spindle wave automatic detection method is a machine learning-based method, which adopts a supervised learning algorithm to train training data and automatically analyzes and determines sleep signals so as to reduce labor cost and increase diagnosis accuracy. However, the existing spindle wave automatic detection method often needs to train a constructed model based on large-scale and high-quality tag data, and low-quality tag data can greatly reduce the accuracy of model extraction, cause false detection and missing detection, and involve high cost for constructing a high-quality and large-scale tag data set. Therefore, how to use a small number of data sets to realize rapid and accurate sleep spindle wave detection is still a problem to be solved.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a sleep spindle wave detection method, electronic equipment and medium.
In a first aspect, an embodiment of the present invention provides a sleep spindle wave detection method, where the method includes:
intercepting an EEG signal to be detected into a plurality of EEG signal segments with continuous time sequences and equal length;
inputting EEG signal segments into a pre-trained feature extraction model to obtain EEG features;
inputting the EEG characteristics into a pre-trained classification model, and dividing EEG signal segments into first type EEG signal segments and second type EEG signal segments; wherein the duty cycle of sleep spindles in the first class of EEG signal segments is greater than a first threshold and the duty cycle of sleep spindles in the second class of EEG signal segments is less than the first threshold;
recombining each of the first class of EEG signal segments with its temporally adjacent two EEG signal segments, respectively;
inputting the recombined EEG signal segment into a pre-trained sleep spindle wave prediction model to obtain the peak position of the sleep spindle wave in the recombined EEG signal segment;
setting a second threshold, and when the distance between adjacent peaks is smaller than the second threshold, the signals between the adjacent peaks are also regarded as sleep spindle wave signals; and screening according to the length range of the sleep spindle wave to obtain the prediction result of the sleep spindle wave signal.
In a second aspect, an embodiment of the present invention provides a sleep spindle wave detection system, the system including:
the EEG characteristic acquisition module is used for intercepting an EEG signal to be detected into a plurality of EEG signal fragments with continuous time sequences and equal length; inputting EEG signal segments into a pre-trained feature extraction model to obtain EEG features;
an EEG signal classification module for inputting EEG features into a pre-trained classification model, dividing EEG signal segments into a first class of EEG signal segments and a second class of EEG signal segments; wherein the duty cycle of sleep spindles in the first class of EEG signal segments is greater than a first threshold and the duty cycle of sleep spindles in the second class of EEG signal segments is less than the first threshold;
an EEG signal recombination module, which is used for recombining each EEG signal segment in the first type of EEG signal segments with two EEG signal segments adjacent to the EEG signal segment in time respectively;
the sleep spindle wave peak value prediction module is used for inputting the recombined EEG signal segment into a pre-trained sleep spindle wave prediction model to obtain the peak value position of the sleep spindle wave in the recombined EEG signal segment;
the sleep spindle wave signal prediction module is used for setting a second threshold value, and when the distance between adjacent peak values is smaller than the second threshold value, the signals between the adjacent peak values are also regarded as sleep spindle wave signals; and screening according to the length range of the sleep spindle wave to obtain the prediction result of the sleep spindle wave signal.
In a third aspect, an embodiment of the present invention provides an electronic device, including a memory and a processor, the memory being coupled to the processor; the memory is used for storing program data, and the processor is used for executing the program data to realize the sleep spindle wave detection method.
In a fourth aspect, an embodiment of the present invention provides a computer readable storage medium having stored thereon a computer program, which when executed by a processor, implements the sleep spindle wave detection method described above.
Compared with the prior art, the invention has the beneficial technical effects that:
1) The invention provides a sleep spindle wave detection method, which acquires EEG characteristics through supervised comparison learning, classifies EEG signal fragments according to the EEG characteristics and recombines the EEG signals. Constructing a sleep spindle wave prediction model based on a long-short-term memory network to predict the peak position of the sleep spindle wave in the recombined EEG signal; and then preprocessing according to the distance between adjacent peaks and the length range of the sleep spindle wave to obtain the prediction result of the sleep spindle wave signal. Compared with the traditional manual algorithm, the method provided by the invention can effectively improve the spindle wave detection efficiency, reduce the labor and time cost and develop spindle wave research in the power-assisted field.
2) Compared with the traditional spindle wave detection algorithm based on the fixed characteristic threshold, the method has better applicability and consistency on different data sets, the threshold of the relevant selection characteristic is not required to be adjusted according to the different data sets, and the method can be better suitable for the spindle wave detection task of an expert on the different data sets in an actual scene.
3) Compared with the traditional machine learning method, the method only needs to train the feature extraction model, the classification model and the sleep spindle wave prediction model related in the method by using fewer EEG data samples, and finally achieves the aim of automatic spindle wave detection. Under the current environment that EEG spindle wave public data sets are rare and expert resources are precious, the method can better adapt to fewer data environments, and helps researchers to fully utilize the data resources and develop spindle wave related researches.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the description of the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort to a person skilled in the art.
FIG. 1 is a flow chart of a sleep spindle wave detection method according to an embodiment of the present invention;
fig. 2 is a schematic diagram of a sleep spindle wave detection method according to an embodiment of the present invention;
FIG. 3 is a schematic structural diagram of a feature extraction model according to an embodiment of the present invention;
FIG. 4 is a view of a TSNE visualization based on extracted EEG features provided by an embodiment of the present invention;
FIG. 5 is a schematic diagram of the effect of the sleep spindle prediction model on identifying the peak characteristics of the sleep spindle signal according to the embodiment of the present invention;
FIG. 6 is a schematic diagram showing the effect of recognizing sleep spindle wave components in EEG signal segments after post-processing according to an embodiment of the present invention;
fig. 7 is a schematic diagram of a sleep spindle wave detection system according to an embodiment of the present invention;
fig. 8 is a schematic diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
Reference will now be made in detail to exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, the same numbers in different drawings refer to the same or similar elements, unless otherwise indicated. The implementations described in the following exemplary examples do not represent all implementations consistent with the invention. Rather, they are merely examples of apparatus and methods consistent with aspects of the invention as detailed in the accompanying claims.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in this specification and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used herein refers to and encompasses any or all possible combinations of one or more of the associated listed items.
It should be understood that although the terms first, second, third, etc. may be used herein to describe various information, these information should not be limited by these terms. These terms are only used to distinguish one type of information from another. For example, first information may also be referred to as second information, and similarly, second information may also be referred to as first information, without departing from the scope of the invention. The word "if" as used herein may be interpreted as "at … …" or "at … …" or "responsive to a determination", depending on the context.
The following describes a sleep spindle wave detection method, an electronic device and a medium in detail by referring to the accompanying drawings.
As shown in fig. 1 and 2, the present invention provides a method of extracting sleep spindle waves (sleep spindle) in an EEG signal based on supervised contrast learning (Contrastive Learning) and long short term memory network (LSTM), the method comprising:
s1, intercepting an EEG signal to be detected into a plurality of EEG signal segments with the length M and continuous in time sequence.
Specifically, for an EEG signal to be detected, the EEG signal to be detected is first truncated in sequence, starting from the start point of the signal, into several segments of EEG signal of length M that are consecutive in time sequence.
S2, inputting all EEG signal segments into a pre-trained feature extraction model to obtain EEG features.
Further, the construction process of the training set of the feature extraction model includes:
and acquiring an EEG data set with expert labels, performing general filtering processing on the data, retaining sleep spindle wave components and filtering other interference signals. Illustratively, in an embodiment where the choice of using the DREAMS spindle dataset for spindle wave detection in sleep EEG signals, considering that the spindle wave signal is primarily at 11-16 Hz, the choice of using bandpass filtering with cut-off frequencies of 11 Hz and 16 Hz, respectively, to process the signal, retains the spindle wave component in the signal.
An EEG signal segment of length M (M is longer than the length of the expert annotation) is truncated from the filtered complete EEG signal, and if the truncated EEG signal segment contains more than 50% of the expert annotation segment, for example, if the expert annotation length is 100, then at least 50 of the length is truncated into the segment, then the EEG signal segment is a positive sample segment. If less than 50% of the expert annotation segments are included in the intercepted EEG signal segments, the EEG signal segment samples are considered negative samples. The final positive and negative samples were in a ratio of about 1:1.
In one embodiment using the DREAMS spindle dataset, the expert's tag information file is first parsed and its tag information read. In the DREAMS spindle data set, the label length of the spindle wave is 0.5-1 second. The sampling frequency of the data was 200Hz per second, containing 200 data points per second. Thus, 256 is selected as the intercept length. Several segments of EEG signal of length 256 are intercepted from the complete EEG signal. Randomly intercepting EEG signal fragments containing more than 50% of expert annotation content, and taking the EEG signal fragments as positive samples; randomly intercepted EEG signal segments containing less than 50% expert annotation content were taken as negative samples. The number of positive and negative samples is about 2000, and the ratio of the number of positive and negative samples is about 1:1.
Further, the feature extraction model is composed of a coding layer and a projection layer connected with the coding layer; the coding layer respectively performs dimension reduction and feature extraction on the input EEG signals through the convolution layer and the pooling layer, and the projection layer is used for projecting the dimension reduced signals into a feature space so as to calculate the distance between each sample in the feature space based on a contrast loss function during training. And training the constructed feature extraction model by adopting a label contrast learning idea.
As in an embodiment using a DREAMS spindle dataset, as shown in FIG. 3, a feature extraction model is built based on a convolutional neural network, performing a dimension reduction and feature extraction on EEG data. The structure includes an encoder module and a projection layer module. The encoder module comprises a first convolution layer, a batch normalization layer (BN layer), a first pooling layer, a second convolution layer, a second pooling layer and a third convolution layer which are sequentially connected; the encoder is connected with the projection layer, and the projection layer structure is composed of two linear layers, namely a first linear layer and a second linear layer. After the signal sample with the signal length of 256 passes through the encoder, the data characteristics are extracted through convolution kernel pooling operation, and finally, the characteristic vector with shape (2,1,18) is output. The projection layer projects the feature vectors into the same feature space, and the final output vector is the feature vector of (1, 16). The detailed parameters of each layer model and the activation functions involved are shown in table 1.
Table 1: detailed parameter table of feature extraction model
Further, the process of training the feature extraction model specifically includes:
a batch of data is randomly extracted from the positive sample data set, the negative sample data set, respectively, as training data of the feature extraction model (i.e., data of one batch), and in this example, the size of the batch is selected to be 64.
The training data is normalized, and in this example, the normalization method adopts a maximum and minimum normalization method, and the conversion formula is as follows:
y= (x-min (x))/( max (x)- min (x))
where min (x) represents the minimum value in the training data set x and max (x) represents the maximum value in the training data set x.
Inputting training data into a feature extraction model, extracting features, and calculating a feature vector z obtained after each sample passes through a projection layer i 。
And selecting a model training method of label contrast learning, and calculating a loss function loss, wherein in order to continuously reduce the value of the loss function, the distance between negative samples is required to be gradually increased in the training process, and the distance between positive samples is required to be gradually reduced. In this process, the feature extraction model continuously updates model parameters based on the back propagation manner to ensure that the feature extraction model can learn the appropriate parameters. And ending the training process until the value of the loss function loss tends to be stable, and obtaining a trained feature extraction model.
The loss function loss is expressed as follows:
wherein I is any sample in a sample set I, the sample set I is the sum of positive samples and negative samples in the current training batch, B represents the number of positive samples in a training batch, a is any negative sample therein, A represents the number of negative samples in a training batch, E is any sample in a positive sample set E in the batch, z i Representing a feature vector z obtained by a feature extraction model of a sample i e Representing the feature vector, z, of the positive sample obtained by the feature extraction model a Representing the feature vector, z, obtained by the negative sample through the feature extraction model i ·z e Representing the cosine distance, z, between the sample i and the positive sample i ·z a Representing the cosine distance between sample i and the negative sample, τ represents the hyper-parameters of the feature extraction model.
Fig. 4 shows a visualization of TSNE based on extracted features, where the dots represent positive samples and the square dots represent negative samples, and as can be seen from fig. 4, the feature extraction model can better distinguish whether or not there are spindle waves in the EEG sample based on the features.
S3, inputting all EEG features into a pre-trained classification model, and dividing the EEG features into a first type of EEG signal segment and a second type of EEG signal segment; wherein the duty cycle of sleep spindles in the first class of EEG signal segments is greater than a first threshold and the duty cycle of sleep spindles in the second class of EEG signal segments is less than a second threshold.
In this example, a classification model is constructed based on the sigmoid function, and the extracted features are classified into two categories. The structure of the classification model used is shown in table 2.
Table 2: detailed parameter table of classification model
In this example, the class model is trained using EEG features extracted by the feature extraction model, and the optimizer chooses "rmsprop" and the loss function chooses "binary cross entropy". Each training data batch was 512 batch size, training 100 rounds (epoch).
In this example, the first threshold is set to 50%, in this example whether or not positive and negative samples are distinguished using a length containing more than 50% spindless, so the classification model can divide the truncated fragments into two classes, fragments containing more than 50% spindless and fragments containing less than 50% spindless.
S4, recombining each EEG signal segment in the first type of EEG signal segments with two EEG signal segments adjacent to the EEG signal segment in time respectively.
Specifically, in this example, all EEG signal segment samples containing more than 50% spindle are retained, and each EEG signal segment is recombined with its two temporally adjacent EEG signal segments, respectively, so that each EEG signal segment can obtain two recombined segments. One recombination segment is obtained by recombining the current EEG signal segment with the previous adjacent EEG signal segment, and the other recombination segment is obtained by recombining the current EEG signal segment with the next adjacent EEG signal segment. The length of the recombined EEG signal segment is 2M, wherein the large probability contains a complete spindle interval.
In one embodiment using the DREAMS spindle dataset, each segment obtained after classification of the classification model is combined with its temporally adjacent segments, each segment resulting in two recombined segments of length 512.
S5, inputting the recombined EEG signal segment into a pre-trained sleep spindle wave prediction model to obtain the peak position of the sleep spindle wave in the recombined EEG signal segment.
In an embodiment using a DREAMS spindle data set, a reorganized sample segment is input into a pre-trained sleep spindle prediction model, the sleep spindle prediction model predicts whether each data point in the reorganized sample segment is a spindle peak value, and finally a prediction result is output to predict the position of the peak value belonging to the spindle in each reorganized sample.
The sleep spindle wave prediction model is constructed based on a Long Short Term Memory (LSTM), and is trained by using a fragment sample data set containing expert labels, so that the sleep spindle wave prediction model can learn the peak characteristics of spindle wave spindle signals so as to predict whether a signal at a certain site is the peak of the spindle wave spindle signals.
In one embodiment using the DREAMS spindle dataset, an LSTM model is constructed, the structure of which is shown in Table 3 below.
Table 3: detailed parameter table of LSTM model
The construction process of the training data set corresponding to the sleep spindle wave prediction model comprises the following steps:
2M length fragments are truncated from the complete data, which must contain expert-labeled fragments. For example, segments of data 512 in length and containing expert labels are randomly truncated from the EEG file in the DREAMS spindle dataset. Finally, the number of segments intercepted in this embodiment is 600.
The LSTM model is trained by using a training data set, a loss function is selected from 'mean_squared_error', an optimizer is selected from 'Adam', the size of the trained batch is 1, and 50 epochs are trained.
Fig. 5 shows the effect of the sleep spindle prediction model on the recognition of the peak characteristics of the spindle signal, wherein the target curve frames the range of the spindle signal and the prediction curve marks the peak interval of the spindle signal.
S6, setting a second threshold value, and when the distance between adjacent peaks is smaller than the second threshold value, the signals between the adjacent peaks are regarded as sleep spindle wave signals; and screening according to the length range of the sleep spindle wave to obtain the prediction result of the sleep spindle wave signal.
It should be noted that, based on a priori knowledge, the prediction result needs to be post-processed. Since a segment of a spindle signal may contain multiple peak characteristics, the signal between two adjacent peaks is also judged to be a spindle signal. The length of the through length of the spindle signal satisfies a certain threshold, and the spindle signal which does not satisfy the threshold length can be regarded as a non-spindle signal to be discarded.
In an embodiment using the DREAMS spindle data set, when the distance between adjacent peaks is less than 10-15 data points (the signal sampling frequency is 200Hz, and thus equal to 0.05 s-0.075 s in this embodiment), the signal region between adjacent peaks is also considered as a sleep spindle signal. Meanwhile, the length of the spindle signal is usually between 0.5 and 1.5s, so that predicted fragment signals with the total length smaller than 0.5s or larger than 1.5s are discarded, and the rest predicted fragments are the final spindle signal prediction results.
Fig. 6 shows the effect of post-processing on the identification of spindle wave spindle components in an EEG signal segment, and by comparing fig. 5 and 6, post-processing links the identified peak features into a continuous length of signal, which is then the last identified spindle wave spindle region.
As shown in fig. 7, an embodiment of the present invention further provides a sleep spindle wave detection system, which includes:
the EEG characteristic acquisition module is used for intercepting an EEG signal to be detected into a plurality of EEG signal fragments with continuous time sequences and equal length; inputting EEG signal segments into a pre-trained feature extraction model to obtain EEG features;
an EEG signal classification module for inputting EEG features into a pre-trained classification model, dividing EEG signal segments into a first class of EEG signal segments and a second class of EEG signal segments; wherein the duty cycle of sleep spindles in the first class of EEG signal segments is greater than a first threshold and the duty cycle of sleep spindles in the second class of EEG signal segments is less than the first threshold;
an EEG signal recombination module, which is used for recombining each EEG signal segment in the first type of EEG signal segments with two EEG signal segments adjacent to the EEG signal segment in time respectively;
the sleep spindle wave peak value prediction module is used for inputting the recombined EEG signal segment into a pre-trained sleep spindle wave prediction model to obtain the peak value position of the sleep spindle wave in the recombined EEG signal segment;
the sleep spindle wave signal prediction module is used for setting a second threshold value, and when the distance between adjacent peak values is smaller than the second threshold value, the signals between the adjacent peak values are also regarded as sleep spindle wave signals; and screening according to the length range of the sleep spindle wave to obtain the prediction result of the sleep spindle wave signal.
The present specification also provides a computer readable storage medium storing a computer program operable to perform the above method of data synchronization.
The present specification also provides a schematic structural diagram of the electronic device shown in fig. 8. At the hardware level, the electronic device includes a processor, an internal bus, a network interface, a memory, and a non-volatile storage, as illustrated in fig. 8, although other hardware required by other services may be included. The processor reads the corresponding computer program from the nonvolatile memory into the memory and then runs the computer program to realize the data synchronization method.
Of course, other implementations, such as logic devices or combinations of hardware and software, are not excluded from the present description, that is, the execution subject of the following processing flows is not limited to each logic unit, but may be hardware or logic devices.
In the 90 s of the 20 th century, improvements to one technology could clearly be distinguished as improvements in hardware (e.g., improvements to circuit structures such as diodes, transistors, switches, etc.) or software (improvements to the process flow). However, with the development of technology, many improvements of the current method flows can be regarded as direct improvements of hardware circuit structures. Designers almost always obtain corresponding hardware circuit structures by programming improved method flows into hardware circuits. Therefore, an improvement of a method flow cannot be said to be realized by a hardware entity module. For example, a programmable logic device (Programmable Logic Device, PLD) (e.g., a field programmable gate array (Field Programmable gate array, FPGA)) is an integrated circuit whose logic function is determined by the user programming the device. A designer programs to "integrate" a digital system onto a PLD without requiring the chip manufacturer to design and fabricate application-specific integrated circuit chips. Moreover, nowadays, instead of manually manufacturing integrated circuit chips, such programming is mostly implemented by using "logic compiler" software, which is similar to the software compiler used in program development and writing, and the original code before the compiling is also written in a specific programming language, which is called hardware description language (Hardware Description Language, HDL), but not just one of the hdds, but a plurality of kinds, such as ABEL (Advanced Boolean Expression Language), AHDL (Altera Hardware DescriptionLanguage), confluence, CUPL (Cornell University Programming Language), HDCal, JHDL (Java Hardware Description Language), lava, lola, myHDL, PALASM, RHDL (RubyHardware Description Language), etc., VHDL (Very-High-SpeedIntegrated Circuit Hardware Description Language) and Verilog are currently most commonly used. It will also be apparent to those skilled in the art that a hardware circuit implementing the logic method flow can be readily obtained by merely slightly programming the method flow into an integrated circuit using several of the hardware description languages described above.
The controller may be implemented in any suitable manner, for example, the controller may take the form of, for example, a microprocessor or processor and a computer readable medium storing computer readable program code (e.g., software or firmware) executable by the (micro) processor, logic gates, switches, application specific integrated circuits (Application Specific Integrated Circuit, ASIC), programmable logic controllers, and embedded microcontrollers, examples of which include, but are not limited to, the following microcontrollers: ARC 625D, atmel AT91SAM, microchip PIC18F26K20, and Silicone Labs C8051F320, the memory controller may also be implemented as part of the control logic of the memory. Those skilled in the art will also appreciate that, in addition to implementing the controller in a pure computer readable program code, it is well possible to implement the same functionality by logically programming the method steps such that the controller is in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers, etc. Such a controller may thus be regarded as a kind of hardware component, and means for performing various functions included therein may also be regarded as structures within the hardware component. Or even means for achieving the various functions may be regarded as either software modules implementing the methods or structures within hardware components.
The system, apparatus, module or unit set forth in the above embodiments may be implemented in particular by a computer chip or entity, or by a product having a certain function. One typical implementation is a computer. In particular, the computer may be, for example, a personal computer, a laptop computer, a cellular telephone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.
For convenience of description, the above devices are described as being functionally divided into various units, respectively. Of course, the functions of each element may be implemented in one or more software and/or hardware elements when implemented in the present specification.
It will be appreciated by those skilled in the art that embodiments of the present description may be provided as a method, system, or computer program product. Accordingly, the present specification may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present description can take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.
The present description is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the specification. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
In one typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
The memory may include volatile memory in a computer-readable medium, random Access Memory (RAM) and/or nonvolatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of computer-readable media.
Computer readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of storage media for a computer include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device. Computer-readable media, as defined herein, does not include transitory computer-readable media (transmission media), such as modulated data signals and carrier waves.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article or apparatus that comprises the element.
It will be appreciated by those skilled in the art that embodiments of the present description may be provided as a method, system, or computer program product. Accordingly, the present specification may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present description can take the form of a computer program product on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.
The description may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. The specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory storage devices.
In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments. In particular, for system embodiments, since they are substantially similar to method embodiments, the description is relatively simple, as relevant to see a section of the description of method embodiments.
The foregoing is merely exemplary of the present disclosure and is not intended to limit the disclosure. Various modifications and alterations to this specification will become apparent to those skilled in the art. Any modifications, equivalent substitutions, improvements, or the like, which are within the spirit and principles of the present description, are intended to be included within the scope of the claims of the present description.
Claims (9)
1. A method of sleep spindle wave detection, the method comprising:
intercepting an EEG signal to be detected into a plurality of EEG signal segments with continuous time sequences and equal length;
inputting EEG signal segments into a pre-trained feature extraction model, and obtaining EEG features through supervised contrast learning;
the training process of the feature extraction model comprises the following steps:
acquiring an EEG data set with expert labels, filtering the EEG data, and only retaining sleep spindle wave components;
intercepting EEG signal fragments with the length M from the filtered EEG signals, wherein the length M is greater than the length marked by an expert; setting an expert labeling length threshold, and taking the intercepted EEG signal fragments larger than the expert labeling length threshold as positive samples; taking the intercepted EEG signal segment smaller than the expert annotation length threshold as a negative sample;
extracting a batch of data from all positive samples and negative samples respectively as training data of a feature extraction model, and setting a contrast loss function to train the feature extraction model, wherein the contrast loss function gradually increases the distance between feature vectors corresponding to the negative samples and gradually decreases the distance between feature vectors corresponding to the positive samples; the expression of the contrast loss function is as follows:
Wherein I is any sample in a sample set I, the sample set I is the sum of positive samples and negative samples in the current training batch, B represents the number of positive samples in a training batch, a is any negative sample therein, A represents the number of negative samples in a training batch, E is any sample in a positive sample set E in the batch, z i Representing a feature vector z obtained by a feature extraction model of a sample i e Representing the feature vector, z, of the positive sample obtained by the feature extraction model a Representing the feature vector, z, obtained by the negative sample through the feature extraction model i ·z e Representing the cosine distance, z, between the sample i and the positive sample i ·z a Representing cosine distance between the sample i and the negative sample, and tau represents hyper-parameters of the feature extraction model;
inputting the EEG characteristics into a pre-trained classification model, and dividing EEG signal segments into first type EEG signal segments and second type EEG signal segments; wherein the duty cycle of sleep spindles in the first class of EEG signal segments is greater than a first threshold and the duty cycle of sleep spindles in the second class of EEG signal segments is less than the first threshold;
recombining each of the first class of EEG signal segments with its temporally adjacent two EEG signal segments, respectively;
Inputting the recombined EEG signal segment into a pre-trained sleep spindle wave prediction model to obtain the peak position of the sleep spindle wave in the recombined EEG signal segment;
setting a second threshold, and when the distance between adjacent peaks is smaller than the second threshold, the signals between the adjacent peaks are also regarded as sleep spindle wave signals; and screening according to the length range of the sleep spindle wave to obtain the prediction result of the sleep spindle wave signal.
2. The sleep spindle wave detection method according to claim 1, wherein the feature extraction model includes: an encoder module and a projection layer module connected thereto; the encoder module comprises a first convolution layer, a batch normalization layer, a first pooling layer, a second convolution layer, a second pooling layer and a third convolution layer which are sequentially connected; the projection layer module comprises a first linear layer and a second linear layer which are sequentially connected; the encoder module obtains a first feature vector through convolution kernel pooling operation, and the projection layer module projects the first feature vector in the same feature space to obtain EEG features.
3. The sleep spindle wave detection method according to claim 1, wherein recombining each of the EEG signal segments of the first type with its temporally adjacent two EEG signal segments, respectively, comprises:
Recombining the current EEG signal segment with the previous adjacent EEG signal segment to obtain a first recombined segment;
the current EEG signal segment is recombined with the next adjacent EEG signal segment to obtain a second recombined segment.
4. The method for detecting sleep spindles according to claim 1, wherein the sleep spindle prediction model adopts a long-term and short-term memory network.
5. The sleep spindle wave detection method according to claim 1 or 4, wherein the training process of the sleep spindle wave prediction model comprises:
intercepting a length-2M EEG signal segment from the complete EEG signal, wherein the EEG signal segment must contain expert-labeled segments;
and training the sleep spindle wave prediction model by using the EEG signal segment containing expert labels, so that the sleep spindle wave prediction model can learn the peak characteristics of spindle wave spindle signals, and the sleep spindle wave prediction model can predict whether the signals at a certain position are the peak values of the spindle wave spindle signals.
6. The sleep spindle wave detection method according to claim 1, wherein a second threshold value is set, and when a distance between adjacent peaks is smaller than the second threshold value, a signal between adjacent peaks is also regarded as a sleep spindle wave signal; screening according to the length range of the sleep spindle wave, and obtaining the prediction result of the sleep spindle wave signal comprises the following steps:
Setting a second threshold based on the signal sampling frequency, and when the distance between adjacent peaks is smaller than the second threshold, the signal between the adjacent peaks is also regarded as sleep spindle wave signal
And acquiring the length range of the sleep spindle wave, and discarding EEG prediction segment signals with the total length being greater than or less than the length range of the sleep spindle wave to obtain a prediction result of the sleep spindle wave signal.
7. A sleep spindle wave detection system, the system comprising:
the EEG characteristic acquisition module is used for intercepting an EEG signal to be detected into a plurality of EEG signal fragments with continuous time sequences and equal length; inputting EEG signal segments into a pre-trained feature extraction model, and obtaining EEG features through supervised contrast learning; the training process of the feature extraction model comprises the following steps:
acquiring an EEG data set with expert labels, filtering the EEG data, and only retaining sleep spindle wave components;
intercepting EEG signal fragments with the length M from the filtered EEG signals, wherein the length M is greater than the length marked by an expert; setting an expert labeling length threshold, and taking the intercepted EEG signal fragments larger than the expert labeling length threshold as positive samples; taking the intercepted EEG signal segment smaller than the expert annotation length threshold as a negative sample;
Extracting a batch of data from all positive samples and negative samples respectively as training data of a feature extraction model, and setting a contrast loss function to train the feature extraction model, wherein the contrast loss function gradually increases the distance between feature vectors corresponding to the negative samples and gradually decreases the distance between feature vectors corresponding to the positive samples; the expression of the contrast loss function is as follows:
wherein I is any sample in a sample set I, the sample set I is the sum of positive samples and negative samples in the current training batch, B represents the number of positive samples in a training batch, a is any negative sample therein, A represents the number of negative samples in a training batch, E is any sample in a positive sample set E in the batch, z i Representing a feature vector z obtained by a feature extraction model of a sample i e Representing the feature vector, z, of the positive sample obtained by the feature extraction model a Representing the feature vector, z, obtained by the negative sample through the feature extraction model i ·z e Representing the cosine distance, z, between the sample i and the positive sample i ·z a Representing cosine distance between the sample i and the negative sample, and tau represents hyper-parameters of the feature extraction model;
an EEG signal classification module for inputting EEG features into a pre-trained classification model, dividing EEG signal segments into a first class of EEG signal segments and a second class of EEG signal segments; wherein the duty cycle of sleep spindles in the first class of EEG signal segments is greater than a first threshold and the duty cycle of sleep spindles in the second class of EEG signal segments is less than the first threshold;
An EEG signal recombination module, which is used for recombining each EEG signal segment in the first type of EEG signal segments with two EEG signal segments adjacent to the EEG signal segment in time respectively;
the sleep spindle wave peak value prediction module is used for inputting the recombined EEG signal segment into a pre-trained sleep spindle wave prediction model to obtain the peak value position of the sleep spindle wave in the recombined EEG signal segment;
the sleep spindle wave signal prediction module is used for setting a second threshold value, and when the distance between adjacent peak values is smaller than the second threshold value, the signals between the adjacent peak values are also regarded as sleep spindle wave signals; and screening according to the length range of the sleep spindle wave to obtain the prediction result of the sleep spindle wave signal.
8. An electronic device comprising a memory and a processor, wherein the memory is coupled to the processor; wherein the memory is configured to store program data and the processor is configured to execute the program data to implement the sleep spindle wave detection method according to any one of claims 1-6.
9. A computer-readable storage medium, on which a computer program is stored, characterized in that the program, when being executed by a processor, implements the sleep spindle wave detection method according to any one of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311315398.3A CN117077013B (en) | 2023-10-12 | 2023-10-12 | Sleep spindle wave detection method, electronic equipment and medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311315398.3A CN117077013B (en) | 2023-10-12 | 2023-10-12 | Sleep spindle wave detection method, electronic equipment and medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117077013A CN117077013A (en) | 2023-11-17 |
| CN117077013B true CN117077013B (en) | 2024-03-26 |
Family
ID=88702694
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311315398.3A Active CN117077013B (en) | 2023-10-12 | 2023-10-12 | Sleep spindle wave detection method, electronic equipment and medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117077013B (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112294341A (en) * | 2020-10-29 | 2021-02-02 | 武汉大学 | Sleep electroencephalogram spindle wave identification method and system based on light convolutional neural network |
| CN113180704A (en) * | 2021-04-07 | 2021-07-30 | 北京脑陆科技有限公司 | Sleep spindle wave detection method and system based on EEG brain waves |
| WO2021248882A1 (en) * | 2020-06-08 | 2021-12-16 | 珠海格力电器股份有限公司 | Methods and apparatuses for constructing sleep state detection model and for detecting sleep state |
| CN113907709A (en) * | 2021-09-28 | 2022-01-11 | 西安电子科技大学 | Portable sleep monitoring system based on ear EEG |
| CN115137374A (en) * | 2022-06-23 | 2022-10-04 | 武汉大学 | Sleep stage oriented electroencephalogram interpretability analysis method and related equipment |
| KR20220133471A (en) * | 2021-03-25 | 2022-10-05 | (주)나연테크 | System for analysising sleep stage and method thereof |
| CN116035598A (en) * | 2023-04-03 | 2023-05-02 | 北京肿瘤医院(北京大学肿瘤医院) | Sleep spindle wave intelligent recognition method and system |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20200060566A1 (en) * | 2018-08-24 | 2020-02-27 | Newton Howard | Automated detection of brain disorders |
| US11664110B2 (en) * | 2018-10-02 | 2023-05-30 | New York University | System, method and portable devices for detection and enhancement of sleep spindles |
-
2023
- 2023-10-12 CN CN202311315398.3A patent/CN117077013B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2021248882A1 (en) * | 2020-06-08 | 2021-12-16 | 珠海格力电器股份有限公司 | Methods and apparatuses for constructing sleep state detection model and for detecting sleep state |
| CN112294341A (en) * | 2020-10-29 | 2021-02-02 | 武汉大学 | Sleep electroencephalogram spindle wave identification method and system based on light convolutional neural network |
| KR20220133471A (en) * | 2021-03-25 | 2022-10-05 | (주)나연테크 | System for analysising sleep stage and method thereof |
| CN113180704A (en) * | 2021-04-07 | 2021-07-30 | 北京脑陆科技有限公司 | Sleep spindle wave detection method and system based on EEG brain waves |
| CN113907709A (en) * | 2021-09-28 | 2022-01-11 | 西安电子科技大学 | Portable sleep monitoring system based on ear EEG |
| CN115137374A (en) * | 2022-06-23 | 2022-10-04 | 武汉大学 | Sleep stage oriented electroencephalogram interpretability analysis method and related equipment |
| CN116035598A (en) * | 2023-04-03 | 2023-05-02 | 北京肿瘤医院(北京大学肿瘤医院) | Sleep spindle wave intelligent recognition method and system |
Non-Patent Citations (2)
| Title |
|---|
| A deep learning approach for real-time detection of sleep spindles;Prathamesh M. Kulkarni等;Journal of Neural Engineering;1-53 * |
| 基于CNN-SNN的单通道睡眠脑电纺锤波检测算法研究;王枫;中国优秀硕士学位论文全文数据库医药卫生科技辑(第04期);E060-283 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117077013A (en) | 2023-11-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113095124B (en) | Face living body detection method and device and electronic equipment | |
| CN110503081B (en) | Violent behavior detection method, system, equipment and medium based on interframe difference | |
| CN110245047B (en) | Time sequence abnormality detection method, device and equipment | |
| CN116343314B (en) | Expression recognition method and device, storage medium and electronic equipment | |
| CN111134664B (en) | Epileptic discharge identification method and system based on capsule network and storage medium | |
| CN115804602A (en) | Electroencephalogram emotion signal detection method, equipment and medium based on attention mechanism and with multi-channel feature fusion | |
| Pérez-García et al. | Transfer learning of deep spatiotemporal networks to model arbitrarily long videos of seizures | |
| CN113558644B (en) | Emotion classification method, medium and equipment for 3D matrix and multidimensional convolution network | |
| CN117077013B (en) | Sleep spindle wave detection method, electronic equipment and medium | |
| CN113180696A (en) | Intracranial electroencephalogram detection method and device, electronic equipment and storage medium | |
| CN113133769A (en) | Equipment control method, device and terminal based on motor imagery electroencephalogram signals | |
| CN114386479B (en) | Medical data processing method and device, storage medium and electronic equipment | |
| CN116664956A (en) | Image recognition method and system based on multi-task automatic encoder | |
| CN116030309A (en) | Method and device for generating countermeasure sample, storage medium and electronic equipment | |
| CN116152933A (en) | Training method, device, equipment and storage medium of anomaly detection model | |
| CN114373539A (en) | Method and system for processing traditional Chinese and western medicine clinical data, storage medium and terminal | |
| Samara et al. | EEG Signal Analysis Approaches for Epileptic Seizure Event Prediction Using Deep Learning | |
| Yuan et al. | Automatic Seizure Detection Using Logarithmic Euclidean-Gaussian Mixture Models (LE-GMMs) and Improved Deep Forest Learning | |
| CN117056788B (en) | EEG signal classification method and device based on supervised comparison learning | |
| CN117357132B (en) | Task execution method and device based on multi-layer brain network node participation coefficient | |
| Liu et al. | A learnable front-end based efficient channel attention network for heart sound classification | |
| CN116167984B (en) | Method, device and equipment for fracture false positive detection for inhibiting motion artifact | |
| Hariprasath | Detection of Stress by Machine Learning in IT Industry | |
| CN116595456B (en) | Data screening and model training method and device based on brain-computer interface | |
| Korenić et al. | Review of the machine learning based methods applicable for analysis of intracellular calcium transients imaging |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |