CN112971718A - Syndrome identification method and device, electronic equipment and storage medium - Google Patents
Syndrome identification method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN112971718A CN112971718A CN202110164591.6A CN202110164591A CN112971718A CN 112971718 A CN112971718 A CN 112971718A CN 202110164591 A CN202110164591 A CN 202110164591A CN 112971718 A CN112971718 A CN 112971718A
- Authority
- CN
- China
- Prior art keywords
- data
- syndrome
- physiological data
- feature
- target object
- 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
- A61B5/48—Other medical applications
- A61B5/4854—Diagnosis based on concepts of traditional oriental medicine
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/0059—Measuring for diagnostic purposes; Identification of persons using light, e.g. diagnosis by transillumination, diascopy, fluorescence
- A61B5/0077—Devices for viewing the surface of the body, e.g. camera, magnifying lens
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/01—Measuring temperature of body parts ; Diagnostic temperature sensing, e.g. for malignant or inflamed tissue
-
- 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
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Medical Informatics (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Pathology (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- Biophysics (AREA)
- Public Health (AREA)
- Veterinary Medicine (AREA)
- Artificial Intelligence (AREA)
- Alternative & Traditional Medicine (AREA)
- Evolutionary Computation (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physiology (AREA)
- Psychiatry (AREA)
- Signal Processing (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
- Image Analysis (AREA)
Abstract
The embodiment of the invention discloses a syndrome identification method, a syndrome identification device, electronic equipment and a storage medium, wherein the method comprises the following steps: acquiring current physiological data of a target object at the current moment; wherein the current physiological data at least comprises infrared temperature data and tongue image data; inputting the current physiological data into a pre-trained syndrome recognition model to obtain a target syndrome category of the target object at the current moment; the syndrome identification model is obtained by training a pre-established neural network model based on historical physiological data of a historical object and label data corresponding to the historical physiological data, wherein the label data is used for expressing the syndrome category of the historical physiological data. Through the technical scheme of the embodiment of the invention, the technical effect of accurately and quickly identifying the syndrome category of the target object by combining various physiological data is realized.
Description
Technical Field
The embodiment of the invention relates to the technical field of traditional Chinese medicine diagnosis and treatment, in particular to a syndrome identification method and device, electronic equipment and a storage medium.
Background
The doctor of traditional Chinese medicine usually diagnoses the patient by checking, smelling, asking and cutting to confirm the patient's syndrome. For example, doctors of traditional Chinese medicine determine the pulse strength, frequency and the like by pulse taking, determine the tongue color, the tongue quality and the like by observing the tongue of a patient, and judge the syndrome of the patient by inquiring the information of the physical constitution, the disease symptoms and the like of the patient.
At present, the diagnosis of a patient and the determination of the syndrome by the traditional Chinese medicine diagnosis method need to depend on the personal experience, diagnosis technique, cognition level and thinking ability of a doctor of traditional Chinese medicine. Moreover, the working ability and working time of the doctor of traditional Chinese medicine are limited, and it is difficult to determine the syndrome of the patient quickly and accurately, which leads to the patient to perform the inquiry of traditional Chinese medicine.
Disclosure of Invention
The embodiment of the invention provides a syndrome identification method, a syndrome identification device, electronic equipment and a storage medium, which are used for accurately and quickly identifying the syndrome type of a target object.
In a first aspect, an embodiment of the present invention provides a syndrome identification method, including:
acquiring current physiological data of a target object at the current moment; wherein the current physiological data at least comprises infrared temperature data and tongue image data;
inputting the current physiological data into a pre-trained syndrome recognition model to obtain a target syndrome category of the target object at the current moment;
the syndrome identification model is obtained by training a pre-established neural network model based on historical physiological data of a historical object and label data corresponding to the historical physiological data, wherein the label data is used for expressing the syndrome category of the historical physiological data.
In a second aspect, an embodiment of the present invention further provides a syndrome identification apparatus, where the apparatus includes:
the physiological data acquisition module is used for acquiring current physiological data of the target object at the current moment; wherein the current physiological data at least comprises infrared temperature data and tongue image data;
the target syndrome determining module is used for inputting the current physiological data into a syndrome recognition model which is trained in advance to obtain a target syndrome category of the target object at the current moment;
the syndrome identification model is obtained by training a pre-established neural network model based on historical physiological data of a historical object and label data corresponding to the historical physiological data, wherein the label data is used for expressing the syndrome category of the historical physiological data.
In a third aspect, an embodiment of the present invention further provides an electronic device, where the electronic device includes:
one or more processors;
a storage device for storing one or more programs,
when executed by the one or more processors, cause the one or more processors to implement a syndrome identification method as in any one of the embodiments of the present invention.
In a fourth aspect, the present invention further provides a computer-readable storage medium, on which a computer program is stored, where the computer program is executed by a processor to implement the syndrome identification method according to any one of the embodiments of the present invention.
According to the technical scheme of the embodiment of the invention, the current physiological data of the target object at the current moment is acquired, and the current physiological data is input into the pre-trained syndrome identification model to acquire the target syndrome type of the target object at the current moment, so that the problem that the syndrome type of the target object is difficult to determine quickly and accurately due to experience and time limitation of a doctor in traditional Chinese medicine is solved, and the technical effect of accurately and quickly identifying the syndrome type of the target object by combining various physiological data is realized.
Drawings
In order to more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, a brief description is given below of the drawings used in describing the embodiments. It should be clear that the described figures are only views of some of the embodiments of the invention to be described, not all, and that for a person skilled in the art, other figures can be derived from these figures without inventive effort.
Fig. 1 is a schematic flow chart of a syndrome identification method according to an embodiment of the present invention;
fig. 2 is a schematic diagram of three yin and three yang provided by the first embodiment of the invention;
FIG. 3 is a structural thermodynamic diagram of a top-hot and bottom-cold structure according to an embodiment of the present invention;
FIG. 4 is a diagram of a structure thermodynamic diagram of a lower heat and an upper cold according to an embodiment of the present invention;
FIG. 5 is a flowchart illustrating a syndrome identification method according to a second embodiment of the present invention;
FIG. 6 is an infrared temperature chart for different body potentials according to a second embodiment of the present invention;
FIG. 7 is a diagram illustrating a convolution structure according to a second embodiment of the present invention;
FIG. 8 is a diagram of a syndrome identification model according to a third embodiment of the present invention;
fig. 9 is a schematic structural diagram of a syndrome identification device according to a fourth embodiment of the present invention;
fig. 10 is a schematic structural diagram of an electronic device according to a fifth embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and are not limiting of the invention. It should be further noted that, for the convenience of description, only some of the structures related to the present invention are shown in the drawings, not all of the structures.
Example one
Fig. 1 is a schematic flow chart of a syndrome identification method according to an embodiment of the present invention, which is applicable to determining a syndrome of a target object during diagnosis and treatment of the target object, and the method may be executed by a syndrome identification device, and the device may be implemented in a form of software and/or hardware, where the hardware may be an electronic device, and optionally, the electronic device may be a mobile terminal, and the like.
As shown in fig. 1, the method of this embodiment specifically includes the following steps:
and S110, acquiring current physiological data of the target object at the current moment.
Wherein, the current physiological data at least comprises infrared temperature data and tongue image data. The target object may be a living individual who performs syndrome recognition, and typically, the target object may be a human or an animal, or the like. For ease of understanding, in the embodiment of the present invention, the target object is exemplified as a human. The physiological data may be various data associated with the health status of the target subject. The infrared temperature data may be an infrared temperature image of the target object scanned by an infrared detection device such as an infrared temperature scanning instrument. The tongue image data may be a tongue image of the target object acquired by a photographing apparatus.
It should be noted that in the field of non-equilibrium thermodynamics, the body can be divided into yin and yang according to the temperature, and further into yang, shaoyang, yangming, taiyin, shaoyin and jueyin according to the relationship between the pathology and the temperature, as shown in fig. 2. The traditional Chinese medicine yin-yang theory is used for describing the temperature gradient of the unbalanced thermodynamic system of the human body, and the temperature gradient can be divided into three yin and three yang. The highest temperature area is sunny, the area around the highest temperature area is sunny, and the junction of the high temperature area and the low temperature area (namely the cold-heat junction) is shaoyang; the lowest temperature region is taiyin, the region with lower temperature than taiyin region is shaoyin, and the edge region of shaoyin region is jueyin. Therefore, the human body can be regarded as a heat transfer circulation system, and the heat transfer in the system has the characteristics of ascending and descending of the sun and the cathode, transferring along a path with high heat conductivity coefficient, and transferring from high temperature to low temperature.
As shown in the structural thermodynamic diagram of fig. 3, the temperature of the region corresponding to the symbol a in fig. 3 is the highest, and the heat transfer is performed from the region a in the direction of the arrow. As shown in the thermodynamic diagram of the structure with the lower heat and the upper cold in fig. 4, the temperature of the region corresponding to the mark B in fig. 4 is the highest, and the heat transfer is performed by the region B along the arrow direction. Moreover, the human body can be regarded as a whole circulatory system, the functions of each organ are different, the working principle is different, and therefore the temperature of each organ is different. The temperature of each organ of a healthy person is usually maintained within a specific range. If a person is ill, the work of the organs of the person is uncoordinated, the temperature emitted by the organs is abnormal, and the syndrome information of the patient can be reflected by detecting the organs with abnormal temperature changes. Therefore, the infrared temperature data can be used as one of the basic data for syndrome identification.
It should be noted that, in the field of the traditional Chinese medicine, the tongue is divided into four parts, namely a tongue tip, a tongue middle part, a tongue root and a tongue edge, wherein the tongue tip reflects the pathological changes of the heart and the lung, the tongue reflects the pathological changes of the spleen and the stomach, the tongue root reflects the pathological changes of the kidney, and the tongue edge reflects the pathological changes of the liver and the gallbladder. If the patient is ill, the tongue image of the patient shows changes such as the tongue coating quality, the tongue coating color, the tongue shape and the like, so that the tongue image data is used as one of the basic data for syndrome identification in the embodiment of the invention.
Specifically, at the present moment, the infrared temperature image of the target object obtained by scanning with the infrared detection device and the tongue image of the target object acquired by the photographing instrument are obtained.
Optionally, the current physiological data of the target object at the current time further includes: pulse condition data, blood pressure, heart rate, and disease data.
The pulse condition data may include pulse position, pulse force, pulse rate and other data, and the disease condition data may include patient's self-describing symptoms, past history, family history and other data.
And S120, inputting the current physiological data into a syndrome recognition model trained in advance to obtain the target syndrome category of the target object at the current moment.
The syndrome identification model is obtained by training a pre-established neural network model based on historical physiological data of a historical object and label data corresponding to the historical physiological data, wherein the label data is used for expressing the syndrome category of the historical physiological data. The history object may be all users, may also be the current user, and may also be a history user corresponding to the current user determined based on the user data of the current user, such as a user having the same constitution as the current user. The user data may be physiological data related to the health of the user, such as gender, age, and physical index data. The historical physiological data is physiological data corresponding to a historical subject. The target syndrome type may be an identification result of the syndrome identification model, and may be syndrome label data, where each syndrome label data corresponds to one syndrome type.
In traditional Chinese medicine, the syndromes can be divided into rare or absent syndromes, deficiency of both qi and yin, deficiency of both qi and blood, deficiency of lung and spleen qi, deficiency of lung and kidney qi, deficiency of both heart and spleen, syndrome of stomach heat, damp-heat accumulation in skin, damp-heat in large intestine, damp-heat in liver and gallbladder, damp-heat in uterus, damp-heat in tendons and bones, deficiency of spleen-yang, deficiency of both heart-yang and kidney-yang, deficiency of water and qi accompanied by deficiency of both qi and blood stasis, deficiency of kidney essence, cold-damp in bones and muscles, deficiency of liver-yin and kidney-yin, deficiency of stomach-yin, deficiency of lung-qi and blood stasis (heart-yang deficiency accompanied by blood stasis, deficiency of heart-yang deficiency with blood stasis), qi stagnation and blood stasis, qi stagnation, stagnation of liver-qi, liver-qi stagnation transforming into fire, liver-stagnation and spleen-deficiency, incoordination between liver and stomach, spleen-deficiency with phlegm-damp-dampness, phlegm-damp-stagnation of lung (phlegm turbidity obstruction of lung), phlegm stagnation of phlegm-core (phlegm stagnation syndrome of phlegm stagnation of phlegm), wind-phlegm, Spleen-kidney yang deficiency, etc.
It should be noted that, if the historical objects are all users, the training data for training the syndrome recognition model is abundant, and the recognition effect of the syndrome recognition model can be improved, but the syndrome types corresponding to the same or similar physiological data of users with different constitutions may be different, which may affect the individual adaptability of the syndrome recognition model; if the historical object is the current user, the syndrome recognition model can be more applied to each user, but the physiological data of the current user may not relate to the data corresponding to all syndrome categories, so that the training data volume is low, the model training effect is influenced, and in addition, one syndrome recognition model is constructed for each user, so that the workload is large; if the historical object is a user with the same constitution as the current user, a syndrome recognition model can be trained for users with different constitutions, so that the model can balance the training data volume and the personalized effect of the model, but the syndrome recognition model needs to be classified according to the user constitution, and larger workload is increased. Therefore, when the syndrome recognition model is trained, what kind of historical physiological data of the historical object is used may be selected according to actual needs, and is not specifically limited in this embodiment.
Specifically, the current physiological data is input into a pre-trained syndrome recognition model, and the syndrome recognition model can process the current physiological data to extract high-dimensional features. And then, predicting according to the high-dimensional characteristics, and outputting a syndrome type label to determine a target syndrome type.
The syndrome identification model is obtained by training a pre-established neural network model, and optionally, the neural network model is at least one of a computer vision group, a residual error network and a long-short term memory artificial neural network.
Among them, the advantage of the computer vision Group (VGG) is that the model has a deeper network structure, a smaller convolution kernel and a pooling sampling domain, so that it can obtain more image features while controlling the number of features, avoiding excessive calculation and excessively complex structure. Because a Residual Network (ResNet) has a Residual, the Residual can be regarded as information of an image and then transmitted, so that the Residual structure reduces the knowledge that the Residual Network needs to learn, and the Residual Network is easier to learn. The Long-Short Term Memory artificial neural network (LSTM) has a Long-Term Memory function, is simple to realize, and can solve the problems of gradient disappearance and gradient explosion in the Long sequence training process.
On the basis of the technical scheme of the embodiment, the syndrome recognition model can be trained by using the infrared temperature data and the tongue image data of the label data corresponding to the predetermined syndrome category.
According to the technical scheme of the embodiment of the invention, the current physiological data of the target object at the current moment is acquired, and the current physiological data is input into the pre-trained syndrome identification model to acquire the target syndrome type of the target object at the current moment, so that the problem that the syndrome type of the target object is difficult to determine quickly and accurately due to experience and time limitation of a doctor in traditional Chinese medicine is solved, and the technical effect of accurately and quickly identifying the syndrome type of the target object by combining various physiological data is realized.
Example two
Fig. 5 is a schematic flow chart of a syndrome identification method according to a second embodiment of the present invention, and the present embodiment refers to the technical solution of the present embodiment for a specific implementation manner of obtaining a target syndrome type of a target object at a current time according to current physiological data based on a pre-trained syndrome identification model based on the above embodiments. The same or corresponding terms as those in the above embodiments are not explained in detail herein.
As shown in fig. 5, the method of this embodiment specifically includes the following steps:
s210, acquiring an infrared temperature image of a target object under at least one posture based on an infrared temperature acquisition device; a tongue image of a target object is acquired based on an imaging device.
The infrared temperature acquisition device may be a device for acquiring infrared temperature data of a human body, and may be, for example, an infrared temperature scanner or the like. The image pickup device may be a device having a function of taking an image, and may be, for example, a video camera, a still camera, or the like. The posture may be a body posture, and the at least one posture may be at least one of a hand-lifting front position, a chest back position, a whole front position, a whole back position, a whole right side position, a whole left side position, and the like.
Specifically, the infrared temperature acquisition device acquires body surface temperature data of the target object, and an infrared temperature image of at least one posture can be obtained. The tongue image of the target object is captured by the camera device, and may be a tongue photograph or the like of the target object. The acquired infrared temperature image and tongue image can be used for predicting the syndrome category through a syndrome identification model.
It should be noted that the temperature images in different body postures can show the temperature images of the target object in different postures, and the temperature images can be acquired from different angles. The infrared temperature maps under different body potentials are shown in fig. 6, wherein the left image in fig. 6 is an integral front infrared temperature image, the middle image is an integral rear infrared temperature image, and the right image is an integral left infrared temperature image. The advantage of using infrared temperature image under at least one posture lies in can the comprehensive consideration human body position infrared temperature data under different angles for the recognition effect of syndrome identification model promotes.
S220, inputting the infrared temperature data into a pre-trained syndrome recognition model through a first channel, and dividing the infrared temperature data into at least two groups of sub-region temperature data based on a region division algorithm in the syndrome recognition model and an infrared temperature acquisition region of a target object.
The syndrome identification model can be a two-channel input model, infrared temperature data is input through a first channel, and tongue image data is input through a second channel. The region division algorithm may be an algorithm that separates regions of different features, such as a threshold-based image segmentation algorithm, a watershed algorithm, and the like. The infrared temperature acquisition region may be a temperature acquisition region corresponding to an infrared temperature image, such as a head image, a chest image, or a whole body image. The sub-region temperature data may be temperature data of each region after the region division is performed based on a region division algorithm.
Specifically, the infrared temperature data is used as the input of a first channel of the syndrome recognition model and is input into the syndrome recognition model trained in advance. The region division algorithm in the syndrome identification model can perform region division on the infrared temperature acquisition region of the target object, and the infrared temperature data of each region is used as sub-region temperature data.
It should be noted that the human body regions include the regions corresponding to the triple energizer, the trunk, the meridians and collaterals, and the five sense organs, and can be divided according to the specific requirements in practical application. The temperature of human triple energizer is in a small-amplitude temperature difference thermal structure with the hottest lower energizer and the coolest upper energizer, and when the functions of the five viscera and six bowels are normal, the qi activity of the viscera is lifted and maintained to maintain the stable thermal structure; when the meridians and collaterals are obstructed and the viscera are abnormally lifted, the upper heat and the lower cold of the human body can cause the qi movement to be disordered, and diseases can be caused. Meridians and collaterals are the paths for energy circulation and also the important ways for internal organs to generate heat and dissipate heat. By comparing the temperature levels of the trunk and the respective regions, the normality or abnormality of the respective regions can be determined. Specifically, the heat value calculation may be performed on each area, that is, a difference between the average temperature value of each area and the average temperature value of the trunk is calculated, and further, the heat values of the areas are sorted, and may be used to determine whether the areas are normal or abnormal.
And S230, respectively performing feature extraction on each group of sub-region temperature data based on the feature extraction network layer of the first channel, and determining temperature feature data.
The feature extraction network layer of the first channel is a layer for performing feature extraction on data input by the first channel, and may include a convolutional layer, a pooling layer, an active layer, a residual connection layer, and the like. The temperature characteristic data may be characteristic data output by a feature extraction network layer of the first channel.
Specifically, feature extraction may be performed on each set of sub-region temperature data input by the first channel based on a convolutional layer, where the convolutional layer includes a plurality of convolutional kernels, and each element constituting the convolutional kernel corresponds to a weight coefficient and a deviation amount, and is similar to a neuron of a feedforward neural network. Each neuron in the convolutional layer is connected to a plurality of neurons in a closely located region in the previous layer. Convolutional layer parameters include convolutional kernel size, step size, and padding, which together determine the dimensions of the convolutional layer output signature. Convolutional layer parameters may be determined when training a syndrome recognition model. The convolutional layer is used to learn local features of image data, and then combine the local features into complex and abstract features, and the schematic diagram of the structure of convolution is shown in fig. 7.
After feature extraction by the convolutional layer, the output data is passed to the pooling layer for feature selection and information filtering. The pooling layer comprises a preset pooling function, the function of the pooling layer is to keep the overall characteristics of data, simultaneously reduce the dimensionality of the data, reduce the parameters of a model and enable the training speed to be faster. Further, the data output by the pooling layer is input to an active layer, through which the data can be non-linearly transformed to enable the data to represent features in a multi-dimensional space. Furthermore, the data output by the active layer is input into the residual connecting layer, so that the degradation problem of the neural network can be better solved, the convergence of the neural network is faster, and the final result is not obviously influenced. Data output by a residual connecting layer of the first channel feature extraction network layer can be used as temperature feature data.
And S240, inputting the tongue image data into a pre-trained syndrome recognition model through a second channel, and performing feature extraction on the tongue image data based on a feature extraction network layer of the second channel in the syndrome recognition model to determine the tongue image feature data.
The feature extraction network layer of the second channel is a layer for performing feature extraction on data input by the second channel, and may include a convolutional layer, a pooling layer, an active layer, a residual connection layer, and the like. The tongue image feature data may be feature data output by a feature extraction network layer of the second channel.
Specifically, the tongue image data input in the second channel may be subjected to feature extraction based on the convolutional layer, and output data may be obtained sequentially through the pooling layer, the activation layer, and the residual connection layer, and the output data may be used as the tongue image feature data. The specific implementation of acquiring the tongue image feature data according to the feature extraction network layer is similar to the acquisition of the temperature feature data in S230, and is not described in detail in this step.
And S250, performing feature fusion on the temperature feature data and the tongue image feature data based on a feature fusion network layer in the syndrome identification model, and determining fusion feature data.
The feature fusion network layer is a layer for fusing temperature feature data and tongue image feature data. The fused feature data may be feature data output by a feature fusion network layer.
Specifically, the temperature characteristic data and the tongue image characteristic data may be input to a characteristic fusion network layer in the syndrome identification model to perform fusion of the multidimensional multi-channel characteristic data, that is, the high-dimensional temperature characteristic data and the high-dimensional tongue image characteristic data are spliced and fused to obtain fusion characteristic data. The fused feature data may be used to identify a syndrome category of the target object through subsequent network layers.
Optionally, based on the feature fusion network layer in the syndrome identification model, the specific steps of performing feature fusion on the temperature feature data and the tongue image feature data may be:
and step one, connecting the temperature characteristic data and the tongue image characteristic data to obtain connection characteristic data.
Specifically, the temperature characteristic data and the tongue image characteristic data are connected in series, so that the high-dimensional data of the two channels are processed into the high-dimensional characteristic data of the single channel, and the characteristic data obtained after connection is used as the connection characteristic data.
And secondly, performing feature fusion on the connection feature data based on the full connection layer to determine first fusion feature data.
Specifically, the connection feature data is input into the full connection layer so as to reduce the connection feature data from a high dimension to a low dimension, and the obtained low-dimension data is used as the first fusion feature data. The benefit of obtaining the first fused feature data is: the syndrome type can be analyzed and identified according to the temperature characteristic data and the tongue image characteristic data, and the syndrome type can be analyzed and identified according to the combined characteristic data between the temperature characteristic data and the tongue image characteristic number, so that the identification effect of the syndrome identification model is improved.
It should be noted that the full-link layer can be regarded as matrix multiplication, which is equivalent to a feature space transformation, that is, useful information can be extracted and integrated. An important role of the fully connected layer is dimension transformation, in particular, to change a high dimension to a low dimension while retaining useful information.
And thirdly, processing the first fusion characteristic data based on the neuron processing layer to determine second fusion characteristic data.
Specifically, the first fusion characteristic data is input into the neuron processing layer, processed data is obtained to avoid overfitting of the syndrome identification model, and the processed data is used as second fusion characteristic data.
It should be noted that the characteristic of the neuron processing layer, i.e., the Dropout layer, is to make some neurons temporarily inoperative, and to prevent the model from being over-fitted, and is generally used when training the model. The Dropout layer is used for temporarily removing the neural network training neurons from the network according to a certain probability in the process of syndrome recognition training so as to avoid the occurrence of model overfitting caused by too few training samples.
And S260, classifying and identifying the fusion characteristic data based on a classification identification network layer in the syndrome identification model, and determining the target syndrome category.
The classification and identification network layer can be a data processing layer constructed based on a classification algorithm in a neural network model.
Specifically, based on a high-order feature classification algorithm preset in the convolutional neural network, the fusion feature data can be classified and identified to determine target syndrome label data. Further, the target syndrome category of the target object may be determined based on the correspondence between the target syndrome tag data and the syndrome type.
The syndrome identification model of the embodiment of the invention can extract and fuse high-order features through infrared temperature data and tongue image data, thereby achieving the effect of automatic syndrome identification of the model. The syndrome identification model can be obtained by training a large amount of infrared temperature data, tongue image data and corresponding syndrome types in advance. When the infrared temperature data and the tongue image data of the target object are input into the syndrome identification model, the syndrome type of the target object can be rapidly predicted. Compared with the traditional Chinese medicine diagnosis and treatment method, the method can improve the speed of syndrome identification, save time for doctors of traditional Chinese medicine, and provide the reference for syndrome diagnosis for the doctors of traditional Chinese medicine.
According to the technical scheme of the embodiment of the invention, the infrared temperature image and the tongue image of the target object at the current moment are acquired, the infrared temperature image and the tongue image are input into a pre-trained syndrome recognition model, the infrared temperature data are input into the pre-trained syndrome recognition model through a first channel, the infrared temperature data are divided into at least two groups of subarea temperature data based on a region division algorithm in the syndrome recognition model and an infrared temperature acquisition region of the target object, the characteristic extraction network layer of the first channel is used for respectively extracting the characteristics of each group of subarea temperature data so as to extract high-order temperature characteristic data, the tongue image data are input into the pre-trained syndrome recognition model through a second channel, and the network layer is used for extracting the characteristics of the tongue image data based on the characteristics of the second channel in the syndrome recognition model, the method comprises the steps of extracting high-order tongue image characteristic data, performing characteristic fusion on temperature characteristic data and tongue image characteristic data based on a characteristic fusion network layer, determining fusion characteristic data to obtain characteristic data in each physiological data and among the physiological data, performing classification identification on the fusion characteristic data based on a classification identification network layer, and determining the target syndrome type to obtain the target syndrome type of a target object at the current moment, so that the problem that the target object syndrome type is difficult to determine quickly and accurately due to experience and time limitation of doctors in traditional Chinese medicine is solved, and the technical effect of accurately and quickly identifying the syndrome type of the target object by combining various physiological data is achieved.
EXAMPLE III
Fig. 8 is a schematic diagram of a syndrome identification model provided in the third embodiment of the present invention, and as a preferred embodiment, reference may be made to the technical solution of this embodiment for a specific implementation manner of construction and use of the syndrome identification model. The same or corresponding terms as those in the above embodiments are not explained in detail herein.
As shown in FIG. 8, the syndrome identification model adopts a neural network, such as VGG, ResNet or LSTM. The model may consist of the following network layers: convolution (Convolution), Pooling (Pooling), Activation (Activation), residual concatenation (Skip Connect), Dropout, etc.
The infrared temperature image and the tongue image are respectively input into the syndrome identification model through double-channel input (one channel inputs the infrared temperature image, and the other channel inputs the tongue image). And the model is divided into two channels, and high-order features are extracted by technologies such as convolution, pooling, activation, residual connection and the like, then the high-order features of the two channels are connected (connected), and the connected features pass through a Full connection layer (Full connection) to obtain a pre-constructed syndrome identification model.
For a pre-constructed syndrome identification model, an infrared temperature image and a tongue image of pre-determined syndrome category label data can be used for training so as to improve the accuracy of model prediction.
Specifically, the training syndrome recognition model includes: and acquiring a plurality of training sample data, wherein the training sample data comprises an infrared temperature image, a tongue image and label data corresponding to the infrared temperature image and the tongue image. Wherein the tag data may be determined based on a predetermined mapping table. For each training sample data, processing the training sample data through convolution, pooling, residual connection and the like to obtain vectors with uniform input dimensions, and then inputting the processed training sample data into a pre-constructed syndrome identification model to obtain a training result. And correcting a loss function in the pre-constructed syndrome identification model based on the training result and the real label data result of the training sample data, and training to obtain the to-be-used syndrome identification model by taking the convergence of the loss function as a training target. And checking the to-be-used syndrome recognition model based on the check sample data, and when the check result meets the preset condition, taking the to-be-used syndrome recognition model as a pre-trained syndrome recognition model.
It should be noted that, in order to improve the accuracy of the model, training sample data may be acquired as much as possible.
The model parameters in the pre-constructed syndrome recognition model may be set as default values before the syndrome recognition model is trained to correct the model parameters in the pre-constructed syndrome recognition model when trained based on training sample data.
Specifically, training sample data can be input into a pre-constructed syndrome recognition model to obtain an output value corresponding to the training sample data, a loss value between a standard value and the output value can be calculated based on the standard value and the training output value in the training sample data, and model parameters in the classification model to be trained are determined based on the loss value. The output value is a training result corresponding to the training sample data, the standard value is a real label data result corresponding to the training sample data, and the loss value is a difference value between the output value and the standard value. The syndrome recognition model may be modified according to the convergence condition of the loss function, that is, the loss parameter is used as a condition for detecting whether the loss function reaches the convergence currently, for example, whether the training error is smaller than a preset error or whether the error variation trend tends to be stable, or whether the current iteration number is equal to the preset number. If the detection reaches the convergence condition, for example, the training error of the loss function is smaller than the preset error or the error change tends to be stable, indicating that the model training is finished, the iterative training may be stopped at this time. If the current condition is not met, sample data can be further obtained to train the model until the training error of the loss function is within the preset range. When the training error of the loss function reaches convergence, the model can be used as a syndrome identification model.
The specific manner when using the syndrome recognition model for syndrome recognition is as follows:
the infrared temperature data (for example, the data size may be 384 × 288) is subjected to feature extraction through network layers such as the neural network CNN, firing, Activation, Skip connect, and the like, so as to obtain temperature feature data of a first preset dimension (for example, 512 dimensions).
The tongue image data (for example, the data size may be 384 × 288) is subjected to feature extraction through network layers such as neural networks CNN, firing, Activation, Skip connect, and the like, so as to obtain tongue image feature data of a second preset dimension (for example, 512 dimensions).
Connecting (Connect) the infrared temperature data (512-dimensional infrared temperature data) with preset dimensionality and the tongue image data (512-dimensional tongue image data), performing feature fusion through a neural network Full connection (Full Connect) layer, preventing overfitting through a Dropout layer, and finally outputting feature data of a third preset dimensionality (such as 1000 dimensionality) for syndrome identification and obtaining a corresponding syndrome category label through classification identification.
It should be noted that, the size of the infrared temperature data, the size of the tongue image data, the first preset dimension, the second preset dimension, and the third preset dimension related in this embodiment may be adjusted according to actual requirements, and are not specifically limited in this embodiment.
The syndrome identification model has the advantages that the judgment of syndrome categories is carried out by adopting a neural network algorithm and combining a mode of fusing multiple characteristics, and compared with the traditional single characteristic, the feasibility and the accuracy are higher.
According to the technical scheme of the embodiment of the invention, the infrared temperature image and the tongue image of the target object at the current moment are acquired and input into the pre-trained syndrome identification model to obtain the target syndrome type of the target object at the current moment, so that the problem that the syndrome type of the target object is difficult to determine quickly and accurately due to experience and time limitation of a doctor in traditional Chinese medicine is solved, and the technical effect of accurately and quickly identifying the syndrome type of the target object by combining various physiological data is realized.
Example four
Fig. 9 is a schematic structural diagram of a syndrome identification device according to a fourth embodiment of the present invention, where the device includes: a physiological data acquisition module 310 and a target syndrome determination module 320.
The physiological data acquiring module 310 is configured to acquire current physiological data of the target object at a current time; wherein the current physiological data at least comprises infrared temperature data and tongue image data; a target syndrome determining module 320, configured to input the current physiological data into a syndrome recognition model trained in advance, so as to obtain a target syndrome category of the target object at the current time; the syndrome identification model is obtained by training a pre-established neural network model based on historical physiological data of a historical object and label data corresponding to the historical physiological data, wherein the label data is used for expressing the syndrome category of the historical physiological data.
Optionally, the physiological data acquiring module 310 is further configured to acquire an infrared temperature image of the target object in at least one posture based on an infrared temperature acquiring device; and acquiring a tongue image of the target object based on a camera device.
Optionally, the target syndrome determining module 320 is further configured to input the infrared temperature data into a pre-trained syndrome recognition model through a first channel, and divide the infrared temperature data into at least two sets of sub-region temperature data based on a region division algorithm in the syndrome recognition model and an infrared temperature acquisition region of the target object; respectively extracting the characteristics of each group of sub-region temperature data based on the characteristic extraction network layer of the first channel to determine temperature characteristic data; inputting the tongue image data into a pre-trained syndrome recognition model through a second channel, and performing feature extraction on the tongue image data based on a feature extraction network layer of the second channel in the syndrome recognition model to determine tongue image feature data; performing feature fusion on the temperature feature data and the tongue image feature data based on a feature fusion network layer in the syndrome identification model to determine fusion feature data; and based on a classification and identification network layer in the syndrome identification model, classifying and identifying the fusion characteristic data, and determining the target syndrome category.
Optionally, the feature extraction network layer of the first channel and/or the second channel includes: convolutional layer, pooling layer, activation layer, and residual connecting layer.
Optionally, the target syndrome determining module 320 is further configured to connect the temperature characteristic data and the tongue image characteristic data to obtain connection characteristic data; performing feature fusion on the connection feature data based on a full connection layer, and determining first fusion feature data; and processing the first fusion characteristic data based on a neuron processing layer to determine second fusion characteristic data.
Optionally, the current physiological data of the target object at the current time further includes: pulse condition data and/or disease state data.
Optionally, the neural network model is at least one of a computer vision group, a residual error network and a long-short term memory artificial neural network.
According to the technical scheme of the embodiment of the invention, the current physiological data of the target object at the current moment is acquired, and the current physiological data is input into the pre-trained syndrome identification model to acquire the target syndrome type of the target object at the current moment, so that the problem that the syndrome type of the target object is difficult to determine quickly and accurately due to experience and time limitation of a doctor in traditional Chinese medicine is solved, and the technical effect of accurately and quickly identifying the syndrome type of the target object by combining various physiological data is realized.
The syndrome identification device provided by the embodiment of the invention can execute the syndrome identification method provided by any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.
It should be noted that the units and modules included in the syndrome identification apparatus are merely divided according to functional logic, but are not limited to the above division as long as the corresponding functions can be implemented; in addition, specific names of the functional units are only for convenience of distinguishing from each other, and are not used for limiting the protection scope of the embodiment of the invention.
EXAMPLE five
Fig. 10 is a schematic structural diagram of an electronic device according to a fifth embodiment of the present invention. FIG. 10 illustrates a block diagram of an exemplary electronic device 40 suitable for use in implementing embodiments of the present invention. The electronic device 40 shown in fig. 10 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiment of the present invention.
As shown in fig. 10, electronic device 40 is embodied in the form of a general purpose computing device. The components of electronic device 40 may include, but are not limited to: one or more processors or processing units 401, a system memory 402, and a bus 403 that couples the various system components (including the system memory 402 and the processing unit 401).
The system memory 402 may include computer system readable media in the form of volatile memory, such as Random Access Memory (RAM)404 and/or cache memory 405. The electronic device 40 may further include other removable/non-removable, volatile/nonvolatile computer system storage media. By way of example only, storage system 406 may be used to read from and write to non-removable, nonvolatile magnetic media (not shown in FIG. 10, and commonly referred to as a "hard drive"). Although not shown in FIG. 10, a magnetic disk drive for reading from and writing to a removable, nonvolatile magnetic disk (e.g., a "floppy disk") and an optical disk drive for reading from or writing to a removable, nonvolatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to the bus 403 by one or more data media interfaces. Memory 402 may include at least one program product having a set (e.g., at least one) of program modules that are configured to carry out the functions of embodiments of the invention.
A program/utility 408 having a set (at least one) of program modules 407 may be stored, for example, in memory 402, such program modules 407 including, but not limited to, an operating system, one or more application programs, other program modules, and program data, each of which examples or some combination thereof may comprise an implementation of a network environment. Program modules 407 generally perform the functions and/or methods of the described embodiments of the invention.
The electronic device 40 may also communicate with one or more external devices 409 (e.g., keyboard, pointing device, display 410, etc.), with one or more devices that enable a user to interact with the electronic device 40, and/or with any devices (e.g., network card, modem, etc.) that enable the electronic device 40 to communicate with one or more other computing devices. Such communication may be through input/output (I/O) interface 411. Also, the electronic device 40 may communicate with one or more networks (e.g., a Local Area Network (LAN), a Wide Area Network (WAN), and/or a public network, such as the internet) via the network adapter 412. As shown, the network adapter 412 communicates with the other modules of the electronic device 40 over the bus 403. It should be appreciated that although not shown in FIG. 10, other hardware and/or software modules may be used in conjunction with electronic device 40, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, among others.
The processing unit 401 executes various functional applications and data processing by executing programs stored in the system memory 402, for example, to implement the syndrome identification method provided by the embodiment of the present invention.
EXAMPLE six
An embodiment of the present invention further provides a storage medium containing computer-executable instructions, which when executed by a computer processor, perform a syndrome identification method, including:
acquiring current physiological data of a target object at the current moment; wherein the current physiological data at least comprises infrared temperature data and tongue image data;
inputting the current physiological data into a pre-trained syndrome recognition model to obtain a target syndrome category of the target object at the current moment;
the syndrome identification model is obtained by training a pre-established neural network model based on historical physiological data of a historical object and label data corresponding to the historical physiological data, wherein the label data is used for expressing the syndrome category of the historical physiological data.
Computer storage media for embodiments of the invention may employ any combination of one or more computer-readable media. The computer readable medium may be a computer readable signal medium or a computer readable storage medium. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium would include the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of this document, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.
A computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.
Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
Computer program code for carrying out operations for embodiments of the present invention may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider).
It is to be noted that the foregoing is only illustrative of the preferred embodiments of the present invention and the technical principles employed. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, although the present invention has been described in greater detail by the above embodiments, the present invention is not limited to the above embodiments, and may include other equivalent embodiments without departing from the spirit of the present invention, and the scope of the present invention is determined by the scope of the appended claims.
Claims (10)
1. A syndrome identification method is characterized by comprising the following steps:
acquiring current physiological data of a target object at the current moment; wherein the current physiological data at least comprises infrared temperature data and tongue image data;
inputting the current physiological data into a pre-trained syndrome recognition model to obtain a target syndrome category of the target object at the current moment;
the syndrome identification model is obtained by training a pre-established neural network model based on historical physiological data of a historical object and label data corresponding to the historical physiological data, wherein the label data is used for expressing the syndrome category of the historical physiological data.
2. The method of claim 1, wherein the obtaining current physiological data of the target subject at the current time comprises:
acquiring an infrared temperature image of the target object under at least one posture based on an infrared temperature acquisition device;
and acquiring a tongue image of the target object based on a camera device.
3. The method according to claim 1, wherein the inputting the current physiological data into a pre-trained syndrome recognition model to obtain a target syndrome category of the target subject at the current time comprises:
inputting the infrared temperature data into a pre-trained syndrome identification model through a first channel, and dividing the infrared temperature data into at least two groups of sub-region temperature data based on a region division algorithm in the syndrome identification model and an infrared temperature acquisition region of the target object;
respectively extracting the characteristics of each group of sub-region temperature data based on the characteristic extraction network layer of the first channel to determine temperature characteristic data;
inputting the tongue image data into a pre-trained syndrome recognition model through a second channel, and performing feature extraction on the tongue image data based on a feature extraction network layer of the second channel in the syndrome recognition model to determine tongue image feature data;
performing feature fusion on the temperature feature data and the tongue image feature data based on a feature fusion network layer in the syndrome identification model to determine fusion feature data;
and based on a classification and identification network layer in the syndrome identification model, classifying and identifying the fusion characteristic data, and determining the target syndrome category.
4. The method of claim 3, wherein the feature extraction network layer of the first channel and/or the second channel comprises: convolutional layer, pooling layer, activation layer, and residual connecting layer.
5. The method according to claim 3, wherein the determining fused feature data by performing feature fusion on the temperature feature data and the tongue image feature data based on a feature fusion network layer in the syndrome identification model comprises:
connecting the temperature characteristic data and the tongue image characteristic data to obtain connection characteristic data;
performing feature fusion on the connection feature data based on a full connection layer, and determining first fusion feature data;
and processing the first fusion characteristic data based on a neuron processing layer to determine second fusion characteristic data.
6. The method of claim 1, wherein the current physiological data of the target subject at the current time further comprises: pulse condition data and/or disease state data.
7. The method of claim 1, wherein the neural network model is at least one of a computer vision suite, a residual network, and a long-short term memory artificial neural network.
8. A syndrome identification device, comprising:
the physiological data acquisition module is used for acquiring current physiological data of the target object at the current moment; wherein the current physiological data at least comprises infrared temperature data and tongue image data;
the target syndrome determining module is used for inputting the current physiological data into a syndrome recognition model which is trained in advance to obtain a target syndrome category of the target object at the current moment;
the syndrome identification model is obtained by training a pre-established neural network model based on historical physiological data of a historical object and label data corresponding to the historical physiological data, wherein the label data is used for expressing the syndrome category of the historical physiological data.
9. An electronic device, characterized in that the electronic device comprises:
one or more processors;
a storage device for storing one or more programs,
when executed by the one or more processors, cause the one or more processors to implement a syndrome identification method as recited in any one of claims 1-7.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out a syndrome identification method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110164591.6A CN112971718A (en) | 2021-02-05 | 2021-02-05 | Syndrome identification method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110164591.6A CN112971718A (en) | 2021-02-05 | 2021-02-05 | Syndrome identification method and device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112971718A true CN112971718A (en) | 2021-06-18 |
Family
ID=76348439
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110164591.6A Pending CN112971718A (en) | 2021-02-05 | 2021-02-05 | Syndrome identification method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112971718A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113822199A (en) * | 2021-09-23 | 2021-12-21 | 浙江大华技术股份有限公司 | Object attribute identification method, device, storage medium and electronic device |
| CN113842116A (en) * | 2021-10-14 | 2021-12-28 | 北京鹰之眼智能健康科技有限公司 | Automatic positioning method and device for human acupuncture points and electronic equipment |
| CN114098640A (en) * | 2021-11-18 | 2022-03-01 | 北京鹰之眼智能健康科技有限公司 | Constitution identification method, constitution identification device, storage medium and equipment |
| CN116309501A (en) * | 2023-03-27 | 2023-06-23 | 北京鹰之眼智能健康科技有限公司 | Sore surface type prediction method, electronic equipment and storage medium |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1759804A (en) * | 2005-11-02 | 2006-04-19 | 浙江大学 | Intelligent analyzing and differentiating method of herbalist doctor through integrated references form four parts of diagnosis |
| CN103284697A (en) * | 2013-03-04 | 2013-09-11 | 北京中医药大学 | Qi-blood body fluid network dynamic identification device and method for biological information of salivary secretion, sweating and urine excretion |
| CN108986912A (en) * | 2018-07-12 | 2018-12-11 | 北京三医智慧科技有限公司 | Chinese medicine stomach trouble tongue based on deep learning is as information intelligent processing method |
| CN111598865A (en) * | 2020-05-14 | 2020-08-28 | 上海锘科智能科技有限公司 | Hand-foot-and-mouth disease detection method, device and system based on thermal infrared and RGB (red, green and blue) double-taking |
| CN111597918A (en) * | 2020-04-26 | 2020-08-28 | 北京金山云网络技术有限公司 | Training and detecting method and device of human face living body detection model and electronic equipment |
| CN112070737A (en) * | 2020-09-02 | 2020-12-11 | 张书臣 | Traditional Chinese medicine tongue observation and syndrome differentiation intelligent identification method based on tongue picture image processing |
| US20210007606A1 (en) * | 2019-07-10 | 2021-01-14 | Compal Electronics, Inc. | Method of and imaging system for clinical sign detection |
| CN112289441A (en) * | 2020-11-19 | 2021-01-29 | 吾征智能技术(北京)有限公司 | Multimode-based medical biological characteristic information matching system |
-
2021
- 2021-02-05 CN CN202110164591.6A patent/CN112971718A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1759804A (en) * | 2005-11-02 | 2006-04-19 | 浙江大学 | Intelligent analyzing and differentiating method of herbalist doctor through integrated references form four parts of diagnosis |
| CN103284697A (en) * | 2013-03-04 | 2013-09-11 | 北京中医药大学 | Qi-blood body fluid network dynamic identification device and method for biological information of salivary secretion, sweating and urine excretion |
| CN108986912A (en) * | 2018-07-12 | 2018-12-11 | 北京三医智慧科技有限公司 | Chinese medicine stomach trouble tongue based on deep learning is as information intelligent processing method |
| US20210007606A1 (en) * | 2019-07-10 | 2021-01-14 | Compal Electronics, Inc. | Method of and imaging system for clinical sign detection |
| CN111597918A (en) * | 2020-04-26 | 2020-08-28 | 北京金山云网络技术有限公司 | Training and detecting method and device of human face living body detection model and electronic equipment |
| CN111598865A (en) * | 2020-05-14 | 2020-08-28 | 上海锘科智能科技有限公司 | Hand-foot-and-mouth disease detection method, device and system based on thermal infrared and RGB (red, green and blue) double-taking |
| CN112070737A (en) * | 2020-09-02 | 2020-12-11 | 张书臣 | Traditional Chinese medicine tongue observation and syndrome differentiation intelligent identification method based on tongue picture image processing |
| CN112289441A (en) * | 2020-11-19 | 2021-01-29 | 吾征智能技术(北京)有限公司 | Multimode-based medical biological characteristic information matching system |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113822199A (en) * | 2021-09-23 | 2021-12-21 | 浙江大华技术股份有限公司 | Object attribute identification method, device, storage medium and electronic device |
| CN113842116A (en) * | 2021-10-14 | 2021-12-28 | 北京鹰之眼智能健康科技有限公司 | Automatic positioning method and device for human acupuncture points and electronic equipment |
| CN113842116B (en) * | 2021-10-14 | 2022-09-27 | 北京鹰之眼智能健康科技有限公司 | Automatic positioning method and device for human acupuncture points and electronic equipment |
| CN114098640A (en) * | 2021-11-18 | 2022-03-01 | 北京鹰之眼智能健康科技有限公司 | Constitution identification method, constitution identification device, storage medium and equipment |
| CN116309501A (en) * | 2023-03-27 | 2023-06-23 | 北京鹰之眼智能健康科技有限公司 | Sore surface type prediction method, electronic equipment and storage medium |
| CN116309501B (en) * | 2023-03-27 | 2024-02-02 | 北京鹰之眼智能健康科技有限公司 | Sore surface type prediction method, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112971718A (en) | Syndrome identification method and device, electronic equipment and storage medium | |
| CN109166130B (en) | Image processing method and image processing device | |
| CN109919928B (en) | Medical image detection method and device and storage medium | |
| Sajjad et al. | Leukocytes classification and segmentation in microscopic blood smear: a resource-aware healthcare service in smart cities | |
| WO2021093448A1 (en) | Image processing method and apparatus, server, medical image processing device and storage medium | |
| Wang et al. | Automated interpretation of congenital heart disease from multi-view echocardiograms | |
| CN111755118B (en) | Medical information processing method, device, electronic equipment and storage medium | |
| WO2020019614A1 (en) | Image processing method, electronic device and storage medium | |
| Huan et al. | Deep convolutional neural networks for classifying body constitution based on face image | |
| CN112926537B (en) | Image processing method, device, electronic equipment and storage medium | |
| WO2021151302A1 (en) | Drug quality-control analysis method, apparatus, device, and medium based on machine learning | |
| CN112690802B (en) | Method, device, terminal and storage medium for detecting electrocardiosignals | |
| Uddin et al. | Machine learning based diabetes detection model for false negative reduction | |
| KR20190087681A (en) | A method for determining whether a subject has an onset of cervical cancer | |
| CN112270240A (en) | Signal processing method and device, electronic equipment and storage medium | |
| CN112869753A (en) | Analysis method, equipment, medium and electrocardiograph for QRST waveform of electrocardiogram | |
| CN116110566A (en) | Data processing method, device and equipment | |
| Raut et al. | Transfer learning based video summarization in wireless capsule endoscopy | |
| Golrizkhatami et al. | Multi-scale features for heartbeat classification using directed acyclic graph CNN | |
| CN114098640A (en) | Constitution identification method, constitution identification device, storage medium and equipment | |
| CN110403611B (en) | Method and apparatus for predicting glycated hemoglobin component value in blood, computer device, and storage medium | |
| WO2024005542A1 (en) | Method and device for predicting disease through wrinkle detection | |
| WO2023108418A1 (en) | Brain atlas construction and neural circuit detection method and related product | |
| Chen et al. | Camera-based peripheral edema measurement using machine learning | |
| Kumari et al. | Automated Diabetic Retinopathy Grading based on the Modified Capsule Network Architecture |
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 |