CN114159071A - Parkinson prediction intelligent method and system based on electrocardiogram image - Google Patents
Parkinson prediction intelligent method and system based on electrocardiogram image Download PDFInfo
- Publication number
- CN114159071A CN114159071A CN202111580521.5A CN202111580521A CN114159071A CN 114159071 A CN114159071 A CN 114159071A CN 202111580521 A CN202111580521 A CN 202111580521A CN 114159071 A CN114159071 A CN 114159071A
- Authority
- CN
- China
- Prior art keywords
- parkinson
- neural network
- convolutional neural
- data
- image
- 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
Images
Classifications
-
- 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
- A61B5/318—Heart-related electrical modalities, e.g. electrocardiography [ECG]
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/40—Detecting, measuring or recording for evaluating the nervous system
- A61B5/4076—Diagnosing or monitoring particular conditions of the nervous system
- A61B5/4082—Diagnosing or monitoring movement diseases, e.g. Parkinson, Huntington or Tourette
-
- 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
- A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
-
- 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
- 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
- G06N3/045—Combinations of networks
-
- 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
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H30/00—ICT specially adapted for the handling or processing of medical images
- G16H30/20—ICT specially adapted for the handling or processing of medical images for handling medical images, e.g. DICOM, HL7 or PACS
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
Abstract
The invention provides an electrocardio-image-based intelligent method and system for Parkinson prediction. The method mainly comprises the following steps: firstly, acquiring an electrocardiogram image and related data thereof; secondly, screening and preprocessing the acquired image; thirdly, dividing the preprocessed image into a training set, a testing set and a verification set, and inputting the constructed convolutional neural network for training and verification; fourthly, optimizing parameters of the trained convolutional neural network, predicting the optimized convolutional neural network, and detecting the accuracy of the convolutional neural network; and finally, predicting the Parkinson by using the detected convolutional neural network model. The invention adopts artificial intelligence to interpret the electrocardiogram, can quickly and accurately extract the hidden information of the electrocardiogram exceeding the eyesight limit of human doctors and classify the hidden information, quickly screens the Parkinson's disease in a noninvasive manner, improves the efficiency and the accuracy of the Parkinson's disease diagnosis, reduces the medical cost, and is easy to popularize to the basic level.
Description
Technical Field
The invention relates to the technical field of data processing, in particular to an electrocardio-image-based intelligent method and system for Parkinson prediction.
Background
Parkinson's disease is one of the most common progressive neurodegenerative diseases affecting the health of old people, the morbidity is 17.4 people/10 ten thousand people in the population of 50-59 years, the morbidity is 93.1 people/10 ten thousand people in the population of 70-79 years, the lifelong risk of suffering from the Parkinson disease is about 1.5 percent, and according to statistics, nearly 300 ten thousand Parkinson disease patients account for half of the Parkinson disease patients in the world and are the most countries of the Parkinson disease patients in the world. With the gradual progress of China into an aging society, the incidence rate of Parkinson is further on the rising trend, and the disability rate of Parkinson is high, so that heavy burden is brought to patients, families of patients and the society.
At present, there is no thorough cure for Parkinson's disease, but if early detection and systemic treatment are performed, the disease condition can be controlled and the progress can be delayed, so that the life quality of patients can be improved. Therefore, in order to reduce the huge economic burden of parkinson on the family and society and improve the quality of life of patients, early diagnosis and effective treatment of parkinson are imperative. Improvements and innovations in diagnostic and therapeutic methods for parkinson have been difficult and hot to study in this field. Most of the traditional Parkinson diagnosis needs doctors to carry out clinical diagnosis on the clinical symptoms of patients to confirm the diagnosis, and the homogeneity of the diagnosis result is poor. The diagnosis process is inefficient, and omission is easily caused when early symptoms of the disease are not obvious.
The Convolutional Neural Network (CNN) is a feedforward neural network with convolutional computation, and is one of the most representative algorithms in deep learning. It was first proposed in 1989, being the most basic and most common deep learning method, and was first applied to speech recognition. With the advent of the big data age, the convolutional neural network has better application in image recognition and license plate recognition in recent years. Today, the application of convolutional neural networks to various fields of medicine has become a new trend in the development of artificial intelligence theory system, for example, the detection and classification of diabetic retinopathy and ovarian cancer tuberculosis by using CNN has shown very high accuracy and efficiency.
At present, an intelligent early screening scheme is researched and developed, so that the early screening method can be used as a method for early screening of the Parkinson's disease, the diagnosis accuracy is improved, the diagnosis tends to be homogeneous, a specialist doctor and a primary doctor can conveniently and quickly master the diagnosis method of the Parkinson's disease, and the early screening method is easy to popularize.
Through the above analysis, the problems and defects of the prior art are as follows:
(1) the traditional method for examining the condition of the Parkinson patients is to determine a Unified Parkinson's Disease Rating Scale (UPDRS), the diagnosis process needs to be completed by a doctor and the patients together, and the diagnosis result has certain subjectivity and poor homogeneity;
(2) due to the fact that the Parkinson patients are inconvenient to move, not only is the time and energy of the patients consumed, but also increasingly tense medical resources are occupied to a certain extent.
(3) The prior art does not have a related technology for carrying out the Parkinson prediction by utilizing a convolutional neural network.
The difficulty in solving the above problems and defects is:
the related art of applying artificial intelligence to the parkinson prediction is lacking, a large amount of image data is needed for support, and the whole training process needs guidance of neurology experts.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides an electrocardio-image-based intelligent method and system for Parkinson prediction.
The invention is realized in this way, a Parkinson's prediction intelligent method based on electrocardio-images mainly comprises the following steps:
acquiring electrocardiogram images of a Parkinson group and a normal contrast group and related data thereof; screening the acquired image data;
secondly, preprocessing the screened image data;
dividing the preprocessed image into a training set, a testing set and a verification set, and inputting the constructed convolutional neural network for training and verification;
optimizing the parameters of the trained convolutional neural network based on the verification result to obtain an optimized convolutional neural network;
testing the optimized convolutional neural network by using the test set data, and detecting the accuracy of the convolutional neural network;
and step six, predicting the Parkinson by using the detected convolutional neural network model.
Further, in the first step, the screening the acquired image data includes:
combining the obtained related data to eliminate dead electrocardiograms (electrocardiograms without continuous P-QRS-T wave groups in each lead), eliminate electrocardiograms of patients with heart diseases, eliminate electrocardiograms of patients who have undergone heart treatment, and eliminate electrocardiograms with serious noise or fuzzy or large limb lead interference; meanwhile, 3 certified neurologists evaluate all clinical data, divide patients into a Parkinson group and a normal control group, and exclude 3 patients with incomplete and consistent diagnosis.
Further, in the second step, the preprocessing the screened image data includes:
and cutting off the text information with invalid image edges by using a PIL image processing library, and only keeping the image information.
Further, the convolutional neural network structure includes:
a first layer: the input data is 400 × 400 × 1, the padding value is 4, 100 convolution kernels are used, the size of the convolution kernels is 11 × 11, the step size is 3, and therefore [ (400-11+2 × 2)/3] +1 ═ 132 features and the output feature is 132 × 132 × 100 are obtained, and then the ReLU activation function 1 processing is carried out, and 132 × 132 × 100 data is obtained;
a second layer: pooling 5 × 5 kernels in the pooling layer 1 at maximum with a step size of 3 to obtain [ (132-5+2 × 1)/3+1] ═ 44 features, resulting in 44 × 44 × 100 data;
and a third layer: the input data is 44 × 44 × 100, the padding value is 0, 150 convolution kernels, the size of the convolution kernels is 5 × 5, the step size is 3, and therefore [ (44-5+2 × 0)/3] +1 ═ 14 features and the output feature is 14 × 14 × 150 are obtained, and then the ReLU activation function 2 processing is carried out, and 14 × 14 × 150 data are obtained;
a fourth layer: pooling 3 × 3 kernels at maximum by step size 2 in pooling layer 2 to obtain [ (14-3)/2+1] ═ 7 features, resulting in 7 × 7 × 150 data;
and a fifth layer: inputting data 7 × 7 × 150, performing full connection after regularization by L2 to obtain 100 features, then performing ReLU activation function 5 processing, and performing dropout1 processing to obtain 150 data;
a sixth layer: input data 100 and L2 are subjected to regularization and full connection to obtain 50 features, then, the ReLU activation function 6 processing is carried out, and then, dropout2 processing is carried out to obtain 50 feature data.
The other purpose of the invention is to provide an electrocardio-image-based intelligent Parkinson prediction system for implementing the electrocardio-image-based intelligent Parkinson prediction method. The intelligent system for the Parkinson prediction based on the electrocardio-image comprises:
the electrocardiogram image data processing module is used for acquiring electrocardiogram images of the Parkinson group and the normal group and related data thereof; screening the acquired image data; preprocessing the screened image data;
the image training and verifying module is used for dividing the preprocessed image into a training set, a testing set and a verifying set and inputting the constructed convolutional neural network for training and verifying;
the convolutional neural network parameter optimization module is used for optimizing the parameters of the trained convolutional neural network based on the verification result to obtain an optimized convolutional neural network;
the convolutional neural network accuracy detection module is used for testing the optimized convolutional neural network model by using the test set data and detecting the accuracy of the convolutional neural network model;
and the Parkinson prediction module is used for predicting the Parkinson by using the detected convolutional neural network model.
Another object of the present invention is to provide a computer device, which includes a memory for storing instructions and a processor for operating according to the instructions to execute the steps of the above-mentioned parkinson's prediction intelligence method based on electrocardiographic images.
Another object of the present invention is to provide a computer-readable storage medium, which stores a computer program configured to execute the above-mentioned steps of the method for intelligent parkinson's prediction based on electrocardiographic images.
By combining all the technical schemes, the invention has the advantages and positive effects that: the invention discloses an electrocardiogram image-based intelligent method and system for Parkinson prediction. And (5) feeding back the detection accuracy of the model by using the verification set, and changing network related parameters to optimize the model so as to optimize the model. And finally, inputting the data of the test set into the model to obtain a prediction score for detecting the accuracy of the model. The model realizes intelligent identification of Parkinson.
Compared with the traditional mode, the invention innovatively adopts the computer to intelligently interpret the electrocardiogram, so that the Parkinson diagnosis tends to be homogeneous; the electrocardiogram reading method is quick and efficient in electrocardiogram reading; the invention rapidly and accurately extracts the electrocardiogram hidden information exceeding the eyesight limit of a human doctor and classifies the electrocardiogram hidden information through the computer, rapidly and non-invasively predicts the Parkinson, improves the efficiency and accuracy of the Parkinson diagnosis, reduces the medical cost and is easy to popularize to the basic level.
Technical effect or experimental effect of comparison. The method comprises the following steps:
at present, the Parkinson diagnosis depends on clinical examination, the workload of clinicians is huge, the homogeneity is poor, the research on the detection and diagnosis of the Parkinson is limited, and no relevant hospital carries out the diagnosis of the Parkinson by a computer automatic auxiliary diagnosis method.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments of the present application will be briefly described below, and it is obvious that the drawings described below are only some embodiments of the present application, and it is obvious for those skilled in the art that other drawings can be obtained from the drawings without creative efforts.
Fig. 1 is a schematic diagram of a parkinson prediction method based on a convolutional neural network according to an embodiment of the present invention.
Fig. 2 is a flowchart of a parkinson prediction method based on a convolutional neural network according to an embodiment of the present invention.
Fig. 3 is a schematic diagram of a standard neural network structure and a Dropout network structure according to an embodiment of the present invention.
Fig. 4 is a graph of an evaluation index of electrocardiogram accuracy versus ROC according to an embodiment of the present invention.
FIG. 5 is a confusion matrix as an example of an electrocardiogram accuracy evaluation index.
Fig. 6 is a schematic diagram of the severity and score of parkinson according to an embodiment of the invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail with reference to the following embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Aiming at the problems in the prior art, the invention provides an electrocardio-image-based intelligent method and system for Parkinson prediction, and the invention is described in detail below with reference to the accompanying drawings.
As shown in fig. 1-2, the intelligent parkinson prediction method based on electrocardiographic images according to the embodiment of the present invention includes the following steps:
s101, acquiring electrocardiogram images of a Parkinson group and a normal contrast group and related data thereof; screening the acquired image data;
s102, preprocessing the screened image data;
s103, dividing the preprocessed image into a training set, a testing set and a verification set, and inputting the constructed convolutional neural network for training and verification;
s104, optimizing the parameters of the trained convolutional neural network based on the verification result to obtain an optimized convolutional neural network;
s105, testing the optimized convolutional neural network by using the test set data, and detecting the accuracy of the convolutional neural network;
and S106, carrying out Parkinson prediction by using the detected convolutional neural network.
In step S101, the screening of the acquired image data provided by the embodiment of the present invention includes:
combining the obtained related data to eliminate dead electrocardiograms (electrocardiograms without continuous P-QRS-T wave groups in each lead), eliminate electrocardiograms of patients with heart diseases, eliminate electrocardiograms of patients who have undergone heart treatment, and eliminate electrocardiograms with serious noise or fuzzy or large limb lead interference; meanwhile, 3 certified neurologists evaluate all clinical data, divide patients into a Parkinson group and a normal control group, and exclude 3 patients with incomplete and consistent diagnosis.
In step S102, the preprocessing of the screened image data provided by the embodiment of the present invention includes:
and cutting off the text information with invalid image edges by using a PIL image processing library, and only keeping the image information.
The convolutional neural network structure provided by the embodiment of the invention comprises:
a first layer: the input data is 400 × 400 × 1, the padding value is 4, 100 convolution kernels are used, the size of the convolution kernels is 11 × 11, the step size is 3, and therefore [ (400-11+2 × 2)/3] +1 ═ 132 features and the output feature is 132 × 132 × 100 are obtained, and then the ReLU activation function 1 processing is carried out, and 132 × 132 × 100 data is obtained;
a second layer: pooling 5 × 5 kernels in the pooling layer 1 at maximum with a step size of 3 to obtain [ (132-5+2 × 1)/3+1] ═ 44 features, resulting in 44 × 44 × 100 data;
and a third layer: the input data is 44 × 44 × 100, the padding value is 0, 150 convolution kernels, the size of the convolution kernels is 5 × 5, the step size is 3, and therefore [ (44-5+2 × 0)/3] +1 ═ 14 features and the output feature is 14 × 14 × 150 are obtained, and then the ReLU activation function 2 processing is carried out, and 14 × 14 × 150 data are obtained;
a fourth layer: pooling 3 × 3 kernels at maximum by step size 2 in pooling layer 2 to obtain [ (14-3)/2+1] ═ 7 features, resulting in 7 × 7 × 150 data;
and a fifth layer: inputting data 7 × 7 × 150, performing full connection after regularization by L2 to obtain 100 features, then performing ReLU activation function 5 processing, and performing dropout1 processing to obtain 150 data;
a sixth layer: input data 100 and L2 are subjected to regularization and full connection to obtain 50 features, then, the ReLU activation function 6 processing is carried out, and then, dropout2 processing is carried out to obtain 50 feature data.
The invention provides an electrocardio-image-based intelligent Parkinson prediction system for implementing the electrocardio-image-based intelligent Parkinson prediction method. The intelligent system for the Parkinson prediction based on the electrocardio-image comprises:
the electrocardiogram image data processing module is used for acquiring electrocardiogram images of the Parkinson group and the normal contrast group and relevant data of the electrocardiogram images; screening the acquired image data; preprocessing the screened image data;
the image training and verifying module is used for dividing the preprocessed image into a training set, a testing set and a verifying set and inputting the constructed convolutional neural network for training and verifying;
the convolutional neural network parameter optimization module is used for optimizing the parameters of the trained convolutional neural network based on the verification result to obtain an optimized convolutional neural network;
the convolutional neural network accuracy detection module is used for testing the optimized convolutional neural network model by using the test set data and detecting the accuracy of the convolutional neural network model
And the Parkinson prediction module is used for predicting the Parkinson by using the detected convolutional neural network model.
The technical effects of the present invention will be further described with reference to specific embodiments.
Example (b):
data set collection
Standard twelve-lead electrocardiography for rest recumbent positions recorded in the second subsidiary hospital of the university at nanchang from 1/2017 to 12/2019 and 31/31 were collected. All electrocardiograms are screened, dead electrocardiograms (electrocardiograms without continuous P-QRS-T wave groups in each lead) are excluded, the electrocardiograms of patients with heart diseases are excluded, the electrocardiograms of patients who have undergone heart treatment are excluded, the electrocardiograms with serious noise or blurring or large limb lead interference are excluded, and relevant data of the researched objects are obtained.
Parkinson diagnosis:
all clinical data were evaluated by 3 certified neurologists, dividing patients into parkinson groups and normal control groups, and excluding patients whose diagnosis by 3 experts was not completely consistent. The present invention includes 1318 total ECG pictures, including the patient with Parkinson's disease picture 518 and the patient with non-Parkinson's disease picture 800. The data set was divided into a training set (64%), a validation set (16%), and a test set (20%).
Picture preprocessing
And cutting off invalid text information at the edge of the image by using a PIL image processing library, only reserving the image information, and inputting the processed image into a network for training or testing.
Convolutional neural network architecture:
a first layer: the input data is 400 × 400 × 1, the padding value is 4, 100 convolution kernels are used, the size of the convolution kernels is 11 × 11, the step size is 3, and therefore [ (400-11+2 × 2)/3] +1 ═ 132 features and the output feature is 132 × 132 × 100 are obtained, and then the ReLU activation function 1 processing is carried out, and 132 × 132 × 100 data is obtained;
a second layer: pooling 5 × 5 kernels in the pooling layer 1 at maximum with a step size of 3 to obtain [ (132-5+2 × 1)/3+1] ═ 44 features, resulting in 44 × 44 × 100 data;
and a third layer: the input data is 44 × 44 × 100, the padding value is 0, 150 convolution kernels, the size of the convolution kernels is 5 × 5, the step size is 3, and therefore [ (44-5+2 × 0)/3] +1 ═ 14 features and the output feature is 14 × 14 × 150 are obtained, and then the ReLU activation function 2 processing is carried out, and 14 × 14 × 150 data are obtained;
a fourth layer: pooling 3 × 3 kernels at maximum by step size 2 in pooling layer 2 to obtain [ (14-3)/2+1] ═ 7 features, resulting in 7 × 7 × 150 data;
and a fifth layer: inputting data 7 × 7 × 150, performing full connection after regularization by L2 to obtain 100 features, then performing ReLU activation function 5 processing, and performing dropout1 processing to obtain 150 data;
a sixth layer: input data 100 and L2 are subjected to regularization and full connection to obtain 50 features, then, the ReLU activation function 6 processing is carried out, and then, dropout2 processing is carried out to obtain 50 feature data.
Dropout is to randomly make the weights of some hidden layer nodes in the network not work during model training, and these nodes still exist in the network although the weights do not work, and only temporarily do not update their weights, and these weights can work again at the next updating time. As shown in fig. 3.
The technical effects of the present invention will be further described below with reference to experiments.
Accuracy analysis for predicting parkinson for electrocardiogram
The test set was tested and the effect is shown in fig. 4-5:
the area under the ROC curve (AUC) is 0.920(0.880,0.950), the sensitivity is 0.969, and the specificity is 0.808;
also, the severity of parkinson was found to correlate with the score, as shown in fig. 6.
The above description is only for the purpose of illustrating the present invention and the appended claims are not to be construed as limiting the scope of the invention, which is intended to cover all modifications, equivalents and improvements that are within the spirit and scope of the invention as defined by the appended claims.
Claims (7)
1. A Parkinson prediction intelligent method based on an electrocardiogram image is characterized by comprising the following steps: the method for automatically predicting the Parkinson disease risk by adopting the electrocardiogram comprises the following steps:
step 1, acquiring electrocardiogram images of a Parkinson group and a normal control group and related data thereof; screening the acquired image data; preprocessing the screened image data;
step 2, dividing the preprocessed image into a training set, a testing set and a verification set, and inputting the constructed convolutional neural network for training and verification;
step 3, optimizing the parameters of the trained convolutional neural network based on the verification result to obtain an optimized convolutional neural network; testing the optimized convolutional neural network by using the test set data, and detecting the accuracy of the convolutional neural network; and predicting the Parkinson by using the detected convolutional neural network model.
2. The intelligent method for parkinson's prediction based on electrocardiographic images of claim 1, wherein: the step 1 of screening the acquired image data comprises:
the dead electrocardiogram, the electrocardiogram of the patient with the heart disease and the electrocardiogram of the patient with the heart disease are eliminated, and the electrocardiogram with serious noise or fuzzy or large limb lead interference is eliminated by combining the obtained related data; meanwhile, 3 certified neurologists evaluate all clinical data, divide patients into a Parkinson group and a normal control group, and exclude 3 patients with incomplete and consistent diagnosis.
3. The intelligent method for parkinson's prediction based on electrocardiographic images of claim 1, wherein: the step 1 of preprocessing the screened image data comprises the following steps:
and cutting off the text information with invalid image edges by using a PIL image processing library, and only keeping the image information.
4. The intelligent method for parkinson's prediction based on electrocardiographic images of claim 1, wherein: step 2, the convolutional neural network structure comprises:
a first layer: the input data is 400 × 400 × 1, the padding value is 4, 100 convolution kernels are used, the size of the convolution kernels is 11 × 11, the step size is 3, and therefore [ (400-11+2 × 2)/3] +1 ═ 132 features and the output feature is 132 × 132 × 100 are obtained, and then the ReLU activation function 1 processing is carried out, and 132 × 132 × 100 data is obtained;
a second layer: pooling 5 × 5 kernels in the pooling layer 1 at maximum with a step size of 3 to obtain [ (132-5+2 × 1)/3+1] ═ 44 features, resulting in 44 × 44 × 100 data;
and a third layer: the input data is 44 × 44 × 100, the padding value is 0, 150 convolution kernels, the size of the convolution kernels is 5 × 5, the step size is 3, and therefore [ (44-5+2 × 0)/3] +1 ═ 14 features and the output feature is 14 × 14 × 150 are obtained, and then the ReLU activation function 2 processing is carried out, and 14 × 14 × 150 data are obtained;
a fourth layer: pooling 3 × 3 kernels at maximum by step size 2 in pooling layer 2 to obtain [ (14-3)/2+1] ═ 7 features, resulting in 7 × 7 × 150 data;
and a fifth layer: inputting data 7 × 7 × 150, performing full connection after regularization by L2 to obtain 100 features, then performing ReLU activation function 5 processing, and performing dropout1 processing to obtain 150 data;
a sixth layer: input data 100 and L2 are subjected to regularization and full connection to obtain 50 features, then, the ReLU activation function 6 processing is carried out, and then, dropout2 processing is carried out to obtain 50 feature data.
5. An electrocardiographic image-based intelligent parkinson prediction system for implementing the electrocardiographic image-based intelligent parkinson prediction method according to any one of claims 1 to 4, characterized in that: the intelligent Parkinson prediction system based on the electrocardiogram image mainly comprises:
the electrocardiogram image data processing module is used for acquiring electrocardiogram images of the Parkinson group and the normal group and related data thereof; screening the acquired image data; preprocessing the screened image data;
the image training and verifying module is used for dividing the preprocessed image into a training set, a testing set and a verifying set and inputting the constructed convolutional neural network for training and verifying;
the convolutional neural network parameter optimization module is used for optimizing the parameters of the trained convolutional neural network based on the verification result to obtain an optimized convolutional neural network;
the convolutional neural network accuracy detection module is used for testing the optimized convolutional neural network model by using the test set data and detecting the accuracy of the convolutional neural network model;
and the Parkinson prediction module is used for predicting the Parkinson by using the detected convolutional neural network model.
6. A computer device, characterized by: the computer device comprises a memory and a processor, wherein the memory is used for storing instructions, and the processor is used for operating according to the instructions to execute the steps of the intelligent method for Parkinson prediction based on electrocardio images according to any one of claims 1-4.
7. A computer-readable storage medium storing a computer program, characterized in that: the computer program is configured to execute the steps of the intelligent method for predicting Parkinson's disease based on electrocardio-images according to any one of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111580521.5A CN114159071A (en) | 2021-12-22 | 2021-12-22 | Parkinson prediction intelligent method and system based on electrocardiogram image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111580521.5A CN114159071A (en) | 2021-12-22 | 2021-12-22 | Parkinson prediction intelligent method and system based on electrocardiogram image |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114159071A true CN114159071A (en) | 2022-03-11 |
Family
ID=80487760
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111580521.5A Pending CN114159071A (en) | 2021-12-22 | 2021-12-22 | Parkinson prediction intelligent method and system based on electrocardiogram image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114159071A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114927216A (en) * | 2022-04-29 | 2022-08-19 | 中南大学湘雅医院 | Method and system for predicting treatment effect of PD-1 of melanoma patient based on artificial intelligence |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040225202A1 (en) * | 2003-01-29 | 2004-11-11 | James Skinner | Method and system for detecting and/or predicting cerebral disorders |
| US20100217144A1 (en) * | 2007-06-28 | 2010-08-26 | Arenare Brian | Diagnostic and predictive system and methodology using multiple parameter electrocardiography superscores |
| US20110160603A1 (en) * | 2008-06-13 | 2011-06-30 | The Parkinson's Institute | Diagnosis of neurogenerative disorders |
| US20140289172A1 (en) * | 2012-01-18 | 2014-09-25 | Brainscope Company, Inc. | Method and device for multimodal neurological evaluation |
| CN104902806A (en) * | 2012-11-10 | 2015-09-09 | 加利福尼亚大学董事会 | Systems and methods for evaluation of neuropathologies |
| CN109222902A (en) * | 2018-08-27 | 2019-01-18 | 上海铱硙医疗科技有限公司 | Parkinson's prediction technique, system, storage medium and equipment based on nuclear magnetic resonance |
| US20190223791A1 (en) * | 2017-12-28 | 2019-07-25 | Saleem Sayani | Wearable diagnostic device |
| CN111528842A (en) * | 2020-05-26 | 2020-08-14 | 复嶂环洲生物科技(上海)有限公司 | Quantitative assessment method for Parkinson disease symptoms based on physiological and behavioral indexes |
| US20210241908A1 (en) * | 2018-04-26 | 2021-08-05 | Mindmaze Holding Sa | Multi-sensor based hmi/ai-based system for diagnosis and therapeutic treatment of patients with neurological disease |
-
2021
- 2021-12-22 CN CN202111580521.5A patent/CN114159071A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040225202A1 (en) * | 2003-01-29 | 2004-11-11 | James Skinner | Method and system for detecting and/or predicting cerebral disorders |
| US20100217144A1 (en) * | 2007-06-28 | 2010-08-26 | Arenare Brian | Diagnostic and predictive system and methodology using multiple parameter electrocardiography superscores |
| US20110160603A1 (en) * | 2008-06-13 | 2011-06-30 | The Parkinson's Institute | Diagnosis of neurogenerative disorders |
| US20140289172A1 (en) * | 2012-01-18 | 2014-09-25 | Brainscope Company, Inc. | Method and device for multimodal neurological evaluation |
| CN104902806A (en) * | 2012-11-10 | 2015-09-09 | 加利福尼亚大学董事会 | Systems and methods for evaluation of neuropathologies |
| US20190223791A1 (en) * | 2017-12-28 | 2019-07-25 | Saleem Sayani | Wearable diagnostic device |
| US20210241908A1 (en) * | 2018-04-26 | 2021-08-05 | Mindmaze Holding Sa | Multi-sensor based hmi/ai-based system for diagnosis and therapeutic treatment of patients with neurological disease |
| CN109222902A (en) * | 2018-08-27 | 2019-01-18 | 上海铱硙医疗科技有限公司 | Parkinson's prediction technique, system, storage medium and equipment based on nuclear magnetic resonance |
| CN111528842A (en) * | 2020-05-26 | 2020-08-14 | 复嶂环洲生物科技(上海)有限公司 | Quantitative assessment method for Parkinson disease symptoms based on physiological and behavioral indexes |
Non-Patent Citations (2)
| Title |
|---|
| J WESTON HUGHES ET AL: "Performance of a Convolutional Neural Network and Explainability Technique for 12-Lead Electrocardiogram Interpretation", 《JAMA CARDIOL . 》, vol. 6, no. 11 * |
| 王广志，张立群，杨福生，丁海曙: "人工神经网络在肌肉痉挛分类中的应用", 《中国康复医学杂志》, vol. 15, no. 3 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114927216A (en) * | 2022-04-29 | 2022-08-19 | 中南大学湘雅医院 | Method and system for predicting treatment effect of PD-1 of melanoma patient based on artificial intelligence |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112037913B (en) | Periodontitis intelligent detection system based on convolutional neural network | |
| CN109948719B (en) | Automatic fundus image quality classification method based on residual dense module network structure | |
| CN111598867B (en) | Method, apparatus, and computer-readable storage medium for detecting specific facial syndrome | |
| CN108090509B (en) | Data length self-adaptive electrocardiogram classification method | |
| CN111785363A (en) | AI-guidance-based chronic disease auxiliary diagnosis system | |
| CN110674773A (en) | Dementia recognition system, device and storage medium | |
| CN114548158B (en) | Data processing method for blood sugar prediction | |
| Shaban | Automated screening of parkinson's disease using deep learning based electroencephalography | |
| CN112101424A (en) | Generation method, identification device and equipment of retinopathy identification model | |
| Gunasekaran et al. | A deep learning framework for earlier prediction of diabetic retinopathy from fundus photographs | |
| Soni et al. | Graphical representation learning-based approach for automatic classification of electroencephalogram signals in depression | |
| Sharma et al. | SzHNN: a novel and scalable deep convolution hybrid neural network framework for schizophrenia detection using multichannel EEG | |
| CN114159071A (en) | Parkinson prediction intelligent method and system based on electrocardiogram image | |
| Miao et al. | Classification of Diabetic Retinopathy Based on Multiscale Hybrid Attention Mechanism and Residual Algorithm | |
| CN113362944A (en) | Machine learning-based prediction of functional dyspepsia and acupuncture curative effect | |
| CN116386845A (en) | Schizophrenia diagnosis system based on convolutional neural network and facial dynamic video | |
| Latha et al. | Automated macular disease detection using retinal optical coherence tomography images by fusion of deep learning networks | |
| KR102340670B1 (en) | Deep learning-based psychophysiological test system and method | |
| CN113796850A (en) | Parathyroid MIBI image analysis system, computer device, and storage medium | |
| Meng et al. | Application of Improved U-Net Convolutional Neural Network for Automatic Quantification of the Foveal Avascular Zone in Diabetic Macular Ischemia | |
| Choudhry et al. | DarkNet-19 Based Intelligent Diagnostic System for Ocular Diseases | |
| Shaikh et al. | Automated classification of pneumonia from chest x-ray images using deep transfer learning efficientnet-b0 model | |
| Krishnapriya | EEG-Based Identification of Schizophrenia Using Deep Learning Techniques | |
| Muñoz et al. | Digital Transformation in Epilepsy Diagnosis Using Raw Images and Transfer Learning in Electroencephalograms. Sustainability 2022, 14, 11420 | |
| Ghanime et al. | Early Detection of Diabetic Retinopathy Using ResNet50 |
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 |