CN117174328A - Detrusor overactivity judgment model and establishing method, system and device thereof - Google Patents
Detrusor overactivity judgment model and establishing method, system and device thereof Download PDFInfo
- Publication number
- CN117174328A CN117174328A CN202310981011.1A CN202310981011A CN117174328A CN 117174328 A CN117174328 A CN 117174328A CN 202310981011 A CN202310981011 A CN 202310981011A CN 117174328 A CN117174328 A CN 117174328A
- Authority
- CN
- China
- Prior art keywords
- detrusor overactivity
- dimensional
- sampling
- sample
- bladder pressure
- 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.)
- Pending
Links
- 206010037211 Psychomotor hyperactivity Diseases 0.000 title claims abstract description 57
- 238000000034 method Methods 0.000 title claims abstract description 36
- 230000003187 abdominal effect Effects 0.000 claims abstract description 49
- 238000005070 sampling Methods 0.000 claims abstract description 43
- 239000000523 sample Substances 0.000 claims abstract description 32
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 31
- 238000012549 training Methods 0.000 claims abstract description 19
- 238000007781 pre-processing Methods 0.000 claims abstract description 10
- 230000009467 reduction Effects 0.000 claims abstract description 10
- 239000013068 control sample Substances 0.000 claims abstract description 7
- 238000002203 pretreatment Methods 0.000 claims abstract description 3
- 238000012545 processing Methods 0.000 claims description 11
- 238000013528 artificial neural network Methods 0.000 claims description 6
- 238000011176 pooling Methods 0.000 claims description 6
- 230000000694 effects Effects 0.000 claims description 5
- 230000008569 process Effects 0.000 claims description 5
- 238000004590 computer program Methods 0.000 claims description 2
- 238000004195 computer-aided diagnosis Methods 0.000 abstract description 2
- 208000024891 symptom Diseases 0.000 description 13
- 238000012360 testing method Methods 0.000 description 9
- 238000003745 diagnosis Methods 0.000 description 4
- 230000003202 urodynamic effect Effects 0.000 description 3
- 238000010276 construction Methods 0.000 description 2
- 230000000875 corresponding effect Effects 0.000 description 2
- 238000002790 cross-validation Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 206010011803 Cystocele Diseases 0.000 description 1
- 206010046543 Urinary incontinence Diseases 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000003759 clinical diagnosis Methods 0.000 description 1
- 230000008602 contraction Effects 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000002401 inhibitory effect Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000027939 micturition Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000003205 muscle Anatomy 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 230000002269 spontaneous effect Effects 0.000 description 1
Landscapes
- Measuring And Recording Apparatus For Diagnosis (AREA)
Abstract
The invention belongs to the field of computer aided diagnosis, and particularly discloses a detrusor overactivity judgment model, and a method, a system and a device for establishing the detrusor overactivity judgment model, wherein the detrusor overactivity judgment model comprises the following steps: obtaining a plurality of samples, each sample comprising a set of bladder pressure curves and abdominal pressure curves; dividing the plurality of samples into a detrusor overactivity sample and a control sample; preprocessing each sample to obtain a training set; the pretreatment method comprises the following steps: noise reduction is firstly carried out on each sample, then the bladder pressure curve and the abdominal pressure curve are sampled, bladder pressure signals and abdominal pressure signals obtained by sampling are combined in a row, and normal distribution standardization treatment is carried out on the bladder pressure signals and the abdominal pressure signals; and training the one-dimensional two-channel convolutional neural network through a training set, wherein the trained one-dimensional two-channel convolutional neural network is the detrusor overactivity judgment model. The invention can realize more accurate and effective judgment of detrusor overactivity.
Description
Technical Field
The invention belongs to the field of computer aided diagnosis, and in particular relates to a detrusor overactivity judgment model, and a building method, a system and a device thereof.
Background
Detrusor Overactivity (DO) is an important association with bladder activity and urinary incontinence in subjects, and is one of the important criteria for assessing bladder status in subjects in urodynamic examination (UDS).
Since detrusor pressure is difficult to measure directly in clinic, detrusor pressure is generally calculated indirectly by measuring the bladder pressure and abdominal pressure of a subject. At present, clinicians mainly judge detrusor overactivity by observing and analyzing an examination curve when a subject runs into UDS, which has high requirements on clinical experience and professional knowledge of the clinicians. On the other hand, since there is no internationally uniform definition of detrusor overactivity, the results are also highly susceptible to subjective factors.
Disclosure of Invention
In response to the above-identified deficiencies of the art or improvements, the present invention provides a detrusor overactivity determination model, and a method, system and apparatus for establishing the same, for more accurate and efficient determination of detrusor overactivity.
To achieve the above object, according to a first aspect of the present invention, there is provided a method for establishing a detrusor overactivity determination model, comprising the steps of:
obtaining a plurality of samples, each sample comprising a set of bladder pressure curves and abdominal pressure curves; dividing the plurality of samples into a detrusor overactivity sample and a control sample;
preprocessing each sample to obtain a training set; the pretreatment method comprises the following steps: noise reduction is firstly carried out on each sample, then the bladder pressure curve and the abdominal pressure curve are sampled, bladder pressure signals and abdominal pressure signals obtained by sampling are combined in a row, and normal distribution standardization treatment is carried out on the bladder pressure signals and the abdominal pressure signals;
and training the one-dimensional two-channel convolutional neural network through a training set, wherein the trained one-dimensional two-channel convolutional neural network is the detrusor overactivity judgment model.
As a further preferred option, the bladder pressure curve and the abdominal pressure curve are each noise reduced by a sliding average algorithm.
As a further preferred option, the detrusor overactivity sample is pre-treated by:
adopting a peak value sampling method, adopting the peak value sampling method, taking a time point corresponding to a peak of the bladder pressure curve as a sampling midpoint, and taking a first preset time before the sampling midpoint to a second preset time after the sampling midpoint as a sampling time range; and acquiring a bladder pressure curve and an abdominal pressure curve within a sampling time range to obtain a bladder pressure signal and an abdominal pressure signal.
As a further preferred option, the control sample is pre-treated:
and (3) adopting an overlap sampling method, and synchronously sampling on the bladder pressure curve and the abdominal pressure curve by utilizing a sliding sampling window to obtain a bladder pressure signal and an abdominal pressure signal.
As a further preferred aspect, the one-dimensional two-channel convolutional neural network includes two one-dimensional two-channel CNN groups and one counter-propagating neural network BPNN connected in sequence, each preprocessed sample is subjected to feature extraction through the one-dimensional two-channel CNN groups, and then the extracted features are subjected to type recognition through the counter-propagating neural network BPNN; each one-dimensional two-channel CNN layer group comprises a plurality of one-dimensional two-channel convolution kernels, a convolution layer and a pooling layer, wherein the size of the one-dimensional two-channel convolution kernels in the first one-dimensional two-channel CNN layer group is larger than that of the one-dimensional two-channel convolution kernels in the second one-dimensional two-channel CNN layer group.
According to a second aspect of the present invention, there is provided a detrusor overactivity determination model constructed using the above-described detrusor overactivity determination model construction method.
According to a third aspect of the present invention there is provided detrusor overactivity determination apparatus comprising a data processing module and a category determination module, wherein:
the data processing module is used for preprocessing the acquired target object data;
the category judgment module is used for inputting the preprocessed target object data into the detrusor overactivity judgment model to obtain the detrusor activity category of the target object.
As a further preferable aspect, in the data processing module, the method for preprocessing the target object data includes:
and carrying out moving average noise reduction on the bladder pressure curve and the abdominal pressure curve in the target object data, then adopting a peak value sampling method to sample, combining the bladder pressure signal and the abdominal pressure signal which are obtained by sampling into a row, and carrying out normal distribution standardization processing on the bladder pressure signal and the abdominal pressure signal.
According to a fourth aspect of the present invention, there is provided a system for establishing a detrusor overactivity determination model comprising a processor for performing the above-described detrusor overactivity determination model establishing method.
According to a fifth aspect of the present invention, there is provided a computer readable storage medium having stored thereon a computer program which, when executed by a processor, implements the above-described method of establishing a detrusor overactivity determination model.
In general, compared with the prior art, the above technical solution conceived by the present invention mainly has the following technical advantages:
1. detrusor overactivity is a spontaneous or induced non-inhibitory contraction of the detrusor muscle during the bladder filling phase, but current methods are primarily judged by observing amplitude changes near the peak on the detrusor overactivity curve, as well as the waveform itself. However, in the actual examination process, the abdominal appearance is reduced due to the possible change of the body position of the target object, so that the accuracy of judgment is affected. The model construction method provided by the invention comprehensively considers two original curves of bladder pressure and abdominal pressure, combines the two curves together to construct a two-channel one-dimensional signal, trains the one-dimensional two-channel convolutional neural network, and can correlate data of the two curves in the same time period in the characteristic extraction process, so that the detrusor pressure rise caused by the abdominal pressure drop can be identified, and the detrusor overactivity can be more accurately and effectively judged.
2. Noise is generated in the bladder pressure and abdominal pressure signal acquisition process, and because the frequency of the noise is lower than that of the acquired signal, the random fluctuation caused by the noise in the signal can be effectively and rapidly eliminated by adopting a sliding average algorithm.
3. According to the invention, different sampling methods are adopted for different types of samples, and a peak sampling method is adopted for DO samples, so that each acquired DO symptom sample can cover a complete DO event period; the control samples are subjected to overlapping sampling in a sliding window mode, so that the diversity of the non-DO samples can be improved, and the training precision of the follow-up model can be improved.
Drawings
FIG. 1 is a flow chart of a method of establishing a detrusor overactivity determination model in accordance with an embodiment of the present invention;
FIG. 2 is a schematic diagram of a one-dimensional two-channel convolutional neural network model according to an embodiment of the present invention;
FIG. 3 is a ROC graph of five training results of a one-dimensional two-channel convolutional neural network model according to an embodiment of the present invention;
fig. 4 is a graph comparing the output result of the model of the embodiment of the present invention with the real label.
Detailed Description
The present invention will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention. In addition, the technical features of the embodiments of the present invention described below may be combined with each other as long as they do not collide with each other.
The invention mainly aims at the bladder filling period of urodynamic examination, and provides a method for establishing a detrusor overactivity judgment model based on a one-dimensional double-channel convolutional neural network. Clinically, detrusor pressure is calculated indirectly by measuring bladder pressure and abdominal pressure, since detrusor pressure is difficult to measure directly. The physician will carefully compare the original curves of bladder pressure and abdominal pressure to prevent leakage of critical symptoms when actually judging whether detrusor overactivity is occurring. Based on the above situation, the one-dimensional two-channel sample is constructed by using the original bladder pressure signal and the abdominal pressure signal with DO event to train the convolutional neural network model, so that the integrity of the original information is ensured as much as possible, the trained model can assist in identifying whether DO occurs in the newly input curve, the reliability of checking data acquisition can be improved, and the workload of doctors is reduced.
The method for establishing detrusor overactivity judgment model provided by the embodiment of the invention, as shown in fig. 1, comprises the following steps:
s1, acquiring a bladder pressure curve and an abdominal pressure curve of a plurality of subjects in a urodynamic examination process, wherein a group of bladder pressure curves and abdominal pressure curves of one subject are used as one sample; dividing the sample into a detrusor overactivity sample and a control sample of detrusor not overactivity according to the condition of the subject;
s2, preprocessing each sample, wherein one part of the preprocessed samples is used as a training set, and the other part of the preprocessed samples is used as a testing set;
further, each sample is preprocessed, specifically including:
s21, respectively carrying out data noise reduction on the bladder pressure curve and the abdominal pressure curve by using a sliding average algorithm;
specifically, through F t ＝(A t-1 +A t-2 +A t-3 +...+A t-n ) Noise reduction is performed on n, wherein F t Represents the predicted value for the next time, n represents the sliding window length of the sliding average, A t-1 To A t-n The actual values at times t-1 to t-n are indicated, respectively.
During the bladder pressure and abdominal pressure signal acquisition, noise is generated in the acquired data due to the adjustment of the subject's posture and errors in the physician's operation, thereby confusing the symptom information contained in the signal itself. Denoising the samples is therefore an indispensable key step before model training and testing. Because the noise is lower in frequency than the acquired signal, the random fluctuations in the signal due to the noise can be effectively and quickly eliminated by using a moving average algorithm. Typically, the noise induced random fluctuations will last only 0.3s, the present embodiment sets the sliding window length in the sliding average algorithm to preferably 10.
S22, sampling a bladder pressure curve and an abdominal pressure curve after noise reduction:
for a detrusor overactivity sample, taking a peak of a bladder pressure curve as a sampling midpoint, and extracting the bladder pressure curve and the abdominal pressure curve in a range from a first preset time before the sampling midpoint to a second preset time after the sampling midpoint to obtain a bladder pressure signal and an abdominal pressure signal; typically the first preset time is equal to the second preset time.
And (3) sampling the control sample on the bladder pressure curve and the abdominal pressure curve synchronously by adopting overlapped sampling and utilizing a sliding sampling window to obtain a bladder pressure signal and an abdominal pressure signal.
S23, arranging each sampled sample into a row according to the sequence of the bladder pressure signal and the abdominal pressure signal, and carrying out normal distribution standardization processing on the data.
S3, training and testing the one-dimensional double-channel convolutional neural network through a training set and a testing set to obtain a detrusor overactivity judgment model; detrusor activity classification for the target subject can be achieved by the detrusor overactivity determination model.
Further, the one-dimensional two-channel convolutional neural network comprises two one-dimensional two-channel CNN layer groups and a back propagation neural network BPNN which are sequentially connected. Each one-dimensional double-channel CNN layer group comprises a plurality of one-dimensional double-channel convolution kernels, a convolution layer and a pooling layer, the features of each preprocessed sample are extracted through the convolution layer and the pooling layer, the one-dimensional double-channel convolution kernels are designed for one-dimensional double-channel signals, data between two channels can be correlated at the same time, the two channels are not mutually independent, and the extracted features can be determined together based on original signals of the two channels; the extracted features are then type-identified by a Softmax classifier at the end of the back propagation neural network BPNN. In addition, the convolution kernel of the first CNN layer is relatively large in size, so that obvious features in the signal can be extracted; the convolution kernel size of the second CNN layer is relatively small, and fine features in the salient features can be further extracted.
Specifically, the structure and parameters of the one-dimensional two-channel convolutional neural network are shown in fig. 2 and table 1, wherein 300×2 represents a signal of 2 channels of the input signal, the signal length of each channel is 300×2×60, the number of convolution kernels is 60, each convolution kernel has 2 channels, and the convolution kernel length of each channel is 3.
TABLE 1 one-dimensional two-channel convolutional neural network structure parameter table
| Network layer | Parameter name | Parameter size |
| Input device | Node count | 300×2 |
| Convolutional layer C1 | Convolution kernel | 3×2×60 |
| Pooling layer S2 | Pool window | 3×2 |
| Convolutional layer C3 | Convolution kernel | 3×2×30 |
| Pooling layer S4 | Pool window | 3×2 |
| Full connection layer FC6 | Node count | 128 |
| Output layer | Node count | 2 |
The invention also provides a detrusor overactivity judging device, which comprises a data processing module and a category judging module, wherein:
the data processing module is used for preprocessing the acquired target object data;
the category judgment module is used for inputting the preprocessed target object data into the detrusor overactivity judgment model to obtain the detrusor activity category of the target object.
Further, during preprocessing, the bladder pressure curve and the abdominal pressure curve in the target object data are subjected to moving average noise reduction, and then sampled by adopting a peak sampling method, specifically, all peak points in the checked data are found, the peak points are taken forward and backward by taking the peak points as the centers, a sample with the data length of 300 (the time span is 30 seconds) is formed, and the samples are put into the detrusor overactivity judging model to identify detrusor overactivity events.
The following are specific examples:
samples of 44 subjects data for model training were partitioned, with 18 subjects presenting with DO symptoms. 146 DO symptom samples were collected by manual identification of the bladder pressure and abdominal pressure curves of 18 subjects. In order to ensure that each DO symptom sample is collected to cover a complete DO event cycle, each sample covers a period of 30 seconds, and the maximum peak in the cystocele signal corresponding to the DO symptom sample is taken as the sampling midpoint.
And for the data of the healthy control without DO symptoms, overlapping sampling is carried out by adopting a sliding window mode, and the mode can improve the diversity of non-DO samples and is beneficial to improving the training precision of the follow-up diagnosis model. The bladder pressure and abdominal pressure signals were sampled simultaneously through a sliding window, the width of the sampling window was set to 300, and the step size was set to 50. By sampling data from healthy controls without DO symptoms, a total of 1863 samples with non-DO symptoms were obtained.
In order to reduce random errors in the model training process, the performance of the model is evaluated by adopting five-fold cross validation, and the evaluation mode selects the area under the curve (AUC) of a subject work characteristic curve (ROC), wherein the larger the AUC value is, the better the comprehensive performance of the model is. Fig. 3 is the ROC, AUC, mean AUC and variance of five-fold cross-validation.
Meanwhile, data of 48 subjects were additionally selected for the diagnostic performance of the test model, wherein 32 subjects were diagnosed as having no DO symptoms by a doctor and 16 subjects were diagnosed as having DO symptoms.
For non-DO samples in the test data, non-overlapping sampling is performed in a sliding window mode, namely the sampling window width is set to 300, the step length is also set to 300, and finally 427 non-DO event samples are obtained; for DO samples in the test data, manually labeling by a doctor to select 118 DO event samples altogether;
carrying out sliding average noise reduction, normalization and splicing on the test data in the same preprocessing mode of the training data to form a one-dimensional double-channel signal; and inputting the one-dimensional two-channel signals into a trained DO diagnosis model, and comparing the output results of the model with real labels respectively to obtain a confusion matrix shown in figure 4. It can be seen that the sensitivity of the test results reaches 99.2% and the specificity reaches 98.5%.
In actual clinical diagnosis, the physician needs to comprehensively consider information such as medical history of the subject, as well as information provided in the examination curve. For example, a subject's UDS curve shows DO symptoms, but the subject's daily urination is not affected, which may be related to the subject's noisy examination environment and uncontrolled neural activity in the hospital, in which case the physician may not download DO diagnostic results in the diagnostic book in order to avoid unnecessary psychological burden on the subject. The invention only gives whether DO exists objectively according to the UDS curve through a diagnosis algorithm, and the main purpose is to assist a doctor to quickly acquire the objective result of whether DO exists or not, and the final diagnosis book still needs the doctor to comprehensively judge according to actual conditions.
It will be readily appreciated by those skilled in the art that the foregoing description is merely a preferred embodiment of the invention and is not intended to limit the invention, but any modifications, equivalents, improvements or alternatives falling within the spirit and principles of the invention are intended to be included within the scope of the invention.
Claims (10)
1. A method for establishing a detrusor overactivity determination model, comprising the steps of:
obtaining a plurality of samples, each sample comprising a set of bladder pressure curves and abdominal pressure curves; dividing the plurality of samples into a detrusor overactivity sample and a control sample;
preprocessing each sample to obtain a training set; the pretreatment method comprises the following steps: noise reduction is firstly carried out on each sample, then the bladder pressure curve and the abdominal pressure curve are sampled, bladder pressure signals and abdominal pressure signals obtained by sampling are combined in a row, and normal distribution standardization treatment is carried out on the bladder pressure signals and the abdominal pressure signals;
and training the one-dimensional two-channel convolutional neural network through a training set, wherein the trained one-dimensional two-channel convolutional neural network is the detrusor overactivity judgment model.
2. A method of constructing a detrusor overactivity assessment model as claimed in claim 1, wherein the bladder pressure curve and the abdominal pressure curve are each denoised by a sliding average algorithm.
3. A method of constructing a detrusor overactivity assessment model as claimed in claim 1, wherein, when the detrusor overactivity sample is pre-treated:
adopting a peak value sampling method, taking a time point corresponding to a peak of the bladder pressure curve as a sampling midpoint, and taking a first preset time before the sampling midpoint to a second preset time after the sampling midpoint as a sampling time range; and acquiring a bladder pressure curve and an abdominal pressure curve within a sampling time range to obtain a bladder pressure signal and an abdominal pressure signal.
4. A method of constructing a detrusor overactivity assessment model as claimed in claim 1, wherein, when the control sample is pre-treated:
and (3) adopting an overlap sampling method, and synchronously sampling on the bladder pressure curve and the abdominal pressure curve by utilizing a sliding sampling window to obtain a bladder pressure signal and an abdominal pressure signal.
5. A method of constructing a detrusor overactivity assessment model according to any one of claims 1 to 4, wherein the one-dimensional two-channel convolutional neural network comprises two one-dimensional two-channel CNN sets of layers and one back propagation neural network BPNN connected in sequence, each pre-processed sample being feature extracted by the one-dimensional two-channel CNN sets of layers, and the extracted features being then type-identified by the back propagation neural network BPNN; each one-dimensional two-channel CNN layer group comprises a plurality of one-dimensional two-channel convolution kernels, a convolution layer and a pooling layer, wherein the size of the one-dimensional two-channel convolution kernels in the first one-dimensional two-channel CNN layer group is larger than that of the one-dimensional two-channel convolution kernels in the second one-dimensional two-channel CNN layer group.
6. A detrusor overactivity determination model constructed using a method of constructing a detrusor overactivity determination model as claimed in any one of claims 1 to 5.
7. A detrusor overactivity determination apparatus comprising a data processing module and a category determination module, wherein:
the data processing module is used for preprocessing the acquired target object data;
the category judgment module is used for inputting the preprocessed target object data into the detrusor overactivity judgment model according to claim 6 to obtain the detrusor activity category of the target object.
8. The detrusor overactivity determination apparatus of claim 7, wherein the data processing module pre-processes the target object data by:
and carrying out moving average noise reduction on the bladder pressure curve and the abdominal pressure curve in the target object data, then adopting a peak value sampling method to sample, combining the bladder pressure signal and the abdominal pressure signal which are obtained by sampling into a row, and carrying out normal distribution standardization processing on the bladder pressure signal and the abdominal pressure signal.
9. A system for modeling detrusor overactivity determination, comprising a processor for performing a method of modeling detrusor overactivity determination as claimed in any one of claims 1 to 5.
10. A computer readable storage medium, having stored thereon a computer program, which when executed by a processor, implements a method of establishing a detrusor overactivity assessment model as claimed in any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310981011.1A CN117174328A (en) | 2023-08-04 | 2023-08-04 | Detrusor overactivity judgment model and establishing method, system and device thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310981011.1A CN117174328A (en) | 2023-08-04 | 2023-08-04 | Detrusor overactivity judgment model and establishing method, system and device thereof |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117174328A true CN117174328A (en) | 2023-12-05 |
Family
ID=88928963
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310981011.1A Pending CN117174328A (en) | 2023-08-04 | 2023-08-04 | Detrusor overactivity judgment model and establishing method, system and device thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117174328A (en) |
-
2023
- 2023-08-04 CN CN202310981011.1A patent/CN117174328A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108186011B (en) | Atrial fibrillation detection method, atrial fibrillation detection device and readable storage medium | |
| CN107837082A (en) | Electrocardiogram automatic analysis method and device based on artificial intelligence self study | |
| CN111127467B (en) | Image quantization method, computer device, and storage medium | |
| CN111739647B (en) | Magnetic resonance image-based bipolar disorder suicide risk prediction method and related device | |
| CN110148466A (en) | A kind of heart impact signal atrial fibrillation computer aided diagnosing method based on transfer learning | |
| CN112806977A (en) | Physiological parameter measuring method based on multi-scale fusion network | |
| JP2004503316A (en) | Method and apparatus for providing an indication of autonomic nervous system function | |
| CN113729648B (en) | Wearable pulse-taking bracelet system based on multiple pulse sensors | |
| Buzoianu et al. | Adjusting for verification bias in diagnostic test evaluation: a Bayesian approach | |
| CN117017310A (en) | Acoustic-electric dual-mode congenital heart disease prediction device based on knowledge distillation | |
| Zhou et al. | A pilot study: detrusor overactivity diagnosis method based on deep learning | |
| CN117174328A (en) | Detrusor overactivity judgment model and establishing method, system and device thereof | |
| CN116864104A (en) | Chronic thromboembolic pulmonary artery high-pressure risk classification system based on artificial intelligence | |
| CN112842355B (en) | Electrocardiosignal heart beat detection and identification method based on deep learning target detection | |
| CN113066549B (en) | Clinical effectiveness evaluation method and system of medical instrument based on artificial intelligence | |
| CN110811673A (en) | Heart sound analysis system based on probabilistic neural network model | |
| CN114550896A (en) | Dizziness patient emergency pre-examination triage decision-making method, device and model based on artificial neural network | |
| CN113436724A (en) | Model and system based on scale and brain function index and application thereof | |
| CN113729708A (en) | Lie evaluation method based on eye movement technology | |
| Smith et al. | Post hoc interpretation of urodynamic evaluation is qualitatively different than interpretation at the time of urodynamic study | |
| CN115227266B (en) | Myoelectric signal classification method, computer equipment and readable storage medium | |
| US20230041528A1 (en) | Detrusor pressure estimation from single channel bladder pressure recordings | |
| CN116738352B (en) | Method and device for classifying abnormal rod cells of retinal vascular occlusion disease | |
| CN117877748B (en) | Multi-parameter heart-lung function test equipment and data processing method thereof | |
| CN210354670U (en) | Physiological monitoring and analysis system based on hybrid sensing |
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 |