CN113592842A - Sample serum quality identification method and identification device based on deep learning - Google Patents
Sample serum quality identification method and identification device based on deep learning Download PDFInfo
- Publication number
- CN113592842A CN113592842A CN202110910127.7A CN202110910127A CN113592842A CN 113592842 A CN113592842 A CN 113592842A CN 202110910127 A CN202110910127 A CN 202110910127A CN 113592842 A CN113592842 A CN 113592842A
- Authority
- CN
- China
- Prior art keywords
- biochemical sample
- neural network
- convolutional neural
- network model
- deep convolutional
- 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.)
- Granted
Links
- 210000002966 serum Anatomy 0.000 title claims abstract description 50
- 238000000034 method Methods 0.000 title claims abstract description 45
- 238000013135 deep learning Methods 0.000 title claims abstract description 14
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 158
- 238000012549 training Methods 0.000 claims abstract description 85
- 206010018910 Haemolysis Diseases 0.000 claims abstract description 65
- 206010023126 Jaundice Diseases 0.000 claims abstract description 65
- 230000008588 hemolysis Effects 0.000 claims abstract description 64
- 238000012545 processing Methods 0.000 claims description 21
- 238000012795 verification Methods 0.000 claims description 21
- 239000000126 substance Substances 0.000 claims description 4
- 230000035945 sensitivity Effects 0.000 abstract description 7
- 230000000694 effects Effects 0.000 abstract description 5
- 238000013461 design Methods 0.000 description 20
- 230000008569 process Effects 0.000 description 11
- 238000010586 diagram Methods 0.000 description 8
- 230000009466 transformation Effects 0.000 description 8
- 238000004891 communication Methods 0.000 description 7
- 230000006870 function Effects 0.000 description 5
- 230000004913 activation Effects 0.000 description 4
- 239000008280 blood Substances 0.000 description 2
- 238000013500 data storage Methods 0.000 description 2
- 238000003709 image segmentation Methods 0.000 description 2
- 238000013519 translation Methods 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 235000002566 Capsicum Nutrition 0.000 description 1
- 239000006002 Pepper Substances 0.000 description 1
- 235000016761 Piper aduncum Nutrition 0.000 description 1
- 235000017804 Piper guineense Nutrition 0.000 description 1
- 244000203593 Piper nigrum Species 0.000 description 1
- 235000008184 Piper nigrum Nutrition 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 210000004369 blood Anatomy 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 150000002632 lipids Chemical class 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 150000003839 salts Chemical class 0.000 description 1
- 238000011426 transformation method Methods 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- 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
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/048—Activation functions
-
- 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
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30101—Blood vessel; Artery; Vein; Vascular
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Biomedical Technology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
- Investigating Or Analysing Biological Materials (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The application discloses a sample serum quality identification method and identification equipment based on deep learning, wherein the method comprises the following steps: acquiring a preprocessed biochemical sample image data set; constructing a deep convolutional neural network model frame, and performing learning training on the deep convolutional neural network model frame based on the preprocessed biochemical sample image data set to obtain a deep convolutional neural network model; acquiring a biochemical sample image to be identified, inputting the biochemical sample image to be identified into a deep convolutional neural network model, and acquiring the probability of hemolysis, jaundice and lipemia of a first biochemical sample corresponding to the biochemical sample image to be identified; and determining the judgment condition of hemolysis, jaundice and lipemia corresponding to the first biochemical sample based on the probability of hemolysis, jaundice and lipemia of the first biochemical sample. The method can improve the sensitivity and specificity of identifying the serum quality, obtain good anti-interference capability and improve the identification effect of the serum quality.
Description
Technical Field
The application relates to the technical field of medical detection, in particular to a sample serum quality identification method and identification equipment based on deep learning.
Background
Poor sample quality is one of the main causes of errors in clinical examination, and sample failure accounts for 60% of errors before analysis, so that high attention needs to be paid to how to identify failed samples.
In the identification of serum quality, the conditions of hemolysis, jaundice, blood lipid, etc. of a sample need to be highly regarded. At present, visual evaluation of sample quality is widely applied in clinical laboratories, but due to the influence of environmental and physiological factors (such as dawdon), visual results of different individuals have large difference, which easily results in low accuracy. At present, part of sample pretreatment equipment adopts a traditional image segmentation algorithm and combines a color difference model to identify the serum quality.
However, when the serum quality is identified by combining the traditional image segmentation algorithm with a color difference model, the problems of poor specificity, weak anti-interference capability and the like exist, so that the identification effect of the serum quality is poor.
Disclosure of Invention
Based on the method, the sample serum quality identification method and the sample serum quality identification equipment based on deep learning are provided, and are used for improving the identification effect of the serum quality.
In a first aspect, an embodiment of the present application provides a sample serum quality identification method based on deep learning, including:
obtaining a preprocessed biochemical sample image dataset, the preprocessed biochemical sample image dataset comprising: a plurality of biochemical sample images and serum indexes corresponding to the plurality of biochemical sample images respectively;
constructing a deep convolutional neural network model frame, and performing learning training on the deep convolutional neural network model frame based on the preprocessed biochemical sample image data set to obtain a deep convolutional neural network model;
acquiring a biochemical sample image to be identified, inputting the biochemical sample image to be identified into the deep convolutional neural network model, and acquiring the probability of hemolysis, jaundice and lipemia of a first biochemical sample corresponding to the biochemical sample image to be identified;
determining the judgment condition of hemolysis, jaundice and lipemia corresponding to the first biochemical sample based on the probability of hemolysis, jaundice and lipemia of the first biochemical sample;
and determining the first biochemical sample to be a qualified sample or an unqualified sample based on the judgment condition.
In one possible design, performing learning training on the deep convolutional neural network model framework based on the preprocessed biochemical sample image dataset to obtain a deep convolutional neural network model, including:
randomly dividing the preprocessed biochemical sample image data set into a biochemical sample training data set and a biochemical sample verification data set according to a preset proportion;
training the deep convolutional neural network model framework based on the biochemical sample training data set to obtain an initial deep convolutional neural network model;
respectively inputting the biochemical sample images in the biochemical sample verification data set into the initial deep convolutional neural network model, and performing probability precision verification on the initial deep convolutional neural network model;
and if the probability precision of the initial deep convolutional neural network model meets the preset requirement, taking the initial deep convolutional neural network model as the deep convolutional neural network model.
In one possible design, training the deep convolutional neural network model framework based on the biochemical sample training data set to obtain an initial deep convolutional neural network model, including:
setting system parameters of a tensorfow system, wherein the system parameters comprise an initial learning rate and iteration parameters of the initial learning rate;
performing image enhancement processing on the biochemical sample training data set to obtain a processed biochemical sample training data set;
training the deep convolutional neural network model framework on the tensortow system based on the processed biochemical sample training data set to obtain the initial deep convolutional neural network model.
In one possible design, training the deep convolutional neural network model framework on the tensortow system based on the processed biochemical sample training dataset to obtain the initial deep convolutional neural network model, comprising:
training the deep convolutional neural network model framework on the tensortow system based on the processed biochemical sample training data set to obtain the probabilities of hemolysis, jaundice and lipemia corresponding to the biochemical sample images in the processed biochemical sample training data set;
determining the probability sum of the probabilities of hemolysis, jaundice and lipemia corresponding to the biochemical sample images in the processed biochemical sample training data set based on a preset classification network;
and determining model parameters of the deep convolutional neural network model framework for judging hemolysis, jaundice and lipemia based on the probability sum to obtain the initial deep convolutional neural network model.
In one possible design, the predetermined classification network is a two-class network of Sigmoid activation functions.
In one possible design, constructing a deep convolutional neural network model framework includes:
constructing 782 layers of initial deep convolutional neural network model frames by combining 1-by-1 convolutional kernels with a residual error network;
adding a global average pool (2D) layer and a final output layer behind the initial convolutional neural network model framework to construct the deep convolutional neural network model framework; wherein the content of the first and second substances,
the global average pool 2D layer is used for outputting a feature map of an input biochemical sample image of the deep convolutional neural network model, and the final output layer is used for outputting probabilities of hemolysis, jaundice and lipemia of the biochemical sample.
In one possible design, acquiring the preprocessed image dataset of the biochemical sample includes:
acquiring an original biochemical sample image dataset;
manually marking a biochemical sample image with interference information in a serum part in an original biochemical sample image dataset to obtain a biochemical sample image dataset marked with the interference information;
and setting the image resolution of the biochemical sample images in the biochemical sample image dataset after marking the interference information to be N M Z, wherein N, M, Z is an integer larger than 1, and obtaining the preprocessed biochemical sample image dataset.
In a second aspect, an embodiment of the present application provides an identification device, including:
a processing unit to: obtaining a preprocessed biochemical sample image dataset, the preprocessed biochemical sample image dataset comprising: a plurality of biochemical sample images and serum indexes corresponding to the plurality of biochemical sample images respectively;
constructing a deep convolutional neural network model frame, and performing learning training on the deep convolutional neural network model frame based on the preprocessed biochemical sample image data set to obtain a deep convolutional neural network model;
a determination unit configured to: acquiring a biochemical sample image to be identified, inputting the biochemical sample image to be identified into the deep convolutional neural network model, and acquiring the probability of hemolysis, jaundice and lipemia of a first biochemical sample corresponding to the biochemical sample image to be identified; determining the judgment condition of hemolysis, jaundice and lipemia corresponding to the first biochemical sample based on the probability of hemolysis, jaundice and lipemia of the first biochemical sample; and determining the first biochemical sample to be a qualified sample or an unqualified sample based on the judgment condition.
In one possible design, the processing unit is specifically configured to:
randomly dividing the preprocessed biochemical sample image data set into a biochemical sample training data set and a biochemical sample verification data set according to a preset proportion;
training the deep convolutional neural network model framework based on the biochemical sample training data set to obtain an initial deep convolutional neural network model;
respectively inputting the biochemical sample images in the biochemical sample verification data set into the initial deep convolutional neural network model, and performing probability precision verification on the initial deep convolutional neural network model;
and if the probability precision of the initial deep convolutional neural network model meets the preset requirement, taking the initial deep convolutional neural network model as the deep convolutional neural network model.
In one possible design, the processing unit is specifically configured to:
setting system parameters of a tensorfow system, wherein the system parameters comprise an initial learning rate and iteration parameters of the initial learning rate;
performing image enhancement processing on the biochemical sample training data set to obtain a processed biochemical sample training data set;
training the deep convolutional neural network model framework on the tensortow system based on the processed biochemical sample training data set to obtain the initial deep convolutional neural network model.
In one possible design, the processing unit is specifically configured to:
training the deep convolutional neural network model framework on the tensortow system based on the processed biochemical sample training data set to obtain the probabilities of hemolysis, jaundice and lipemia corresponding to the biochemical sample images in the processed biochemical sample training data set;
determining the probability sum of the probabilities of hemolysis, jaundice and lipemia corresponding to the biochemical sample images in the processed biochemical sample training data set based on a preset classification network;
and determining model parameters of the deep convolutional neural network model framework for judging hemolysis, jaundice and lipemia based on the probability sum to obtain the initial deep convolutional neural network model.
In one possible design, the predetermined classification network is a two-class network of Sigmoid activation functions.
In one possible design, the processing unit is specifically configured to:
constructing 782 layers of initial deep convolutional neural network model frames by combining 1-by-1 convolutional kernels with a residual error network;
adding a global average pool (2D) layer and a final output layer behind the initial convolutional neural network model framework to construct the deep convolutional neural network model framework; wherein the content of the first and second substances,
the global average pool 2D layer is used for outputting a feature map of an input biochemical sample image of the deep convolutional neural network model, and the final output layer is used for outputting probabilities of hemolysis, jaundice and lipemia of the biochemical sample.
In one possible design, the processing unit is specifically configured to:
acquiring an original biochemical sample image dataset;
manually marking a biochemical sample image with interference information in a serum part in an original biochemical sample image dataset to obtain a biochemical sample image dataset marked with the interference information;
and setting the image resolution of the biochemical sample images in the biochemical sample image dataset after marking the interference information to be N M Z, wherein N, M, Z is an integer larger than 1, and obtaining the preprocessed biochemical sample image dataset.
In a third aspect, an embodiment of the present application provides an identification device, where the identification device includes: at least one memory and at least one processor;
the at least one memory is for storing one or more programs;
the one or more programs, when executed by the at least one processor, implement the method as recited in any one of the possible designs of the first aspect above.
In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium storing at least one program; the at least one program, when executed by a processor, performs the method of any one of the possible designs of the first aspect.
The beneficial effect of this application is as follows:
in the technical scheme provided by the application, a preprocessed biochemical sample image data set is obtained, and the preprocessed biochemical sample image data set comprises: the serum indexes of the multiple biochemical sample images and the multiple biochemical sample images respectively correspond to each other; constructing a deep convolutional neural network model frame, and performing learning training on the deep convolutional neural network model frame based on the preprocessed biochemical sample image data set to obtain a deep convolutional neural network model; acquiring a biochemical sample image to be identified, inputting the biochemical sample image to be identified into a deep convolutional neural network model, and acquiring the probability of hemolysis, jaundice and lipemia of a first biochemical sample corresponding to the biochemical sample image to be identified; and determining the judgment condition of hemolysis, jaundice and lipemia corresponding to the first biochemical sample based on the probability of hemolysis, jaundice and lipemia of the first biochemical sample. Through the mode, when the serum quality of the biochemical sample image is identified, the sensitivity and the specificity for identifying the serum quality can be improved, so that good anti-interference capability for identifying the serum quality can be obtained, and the identification effect of the serum quality can be improved.
Drawings
Fig. 1 is a schematic flowchart of a sample serum quality identification method based on deep learning according to an embodiment of the present disclosure;
FIG. 2 is a schematic flow chart of acquiring an image dataset of a preprocessed biochemical sample according to an embodiment of the present disclosure;
fig. 3 is a schematic diagram of a learning and training process for a deep convolutional neural network model according to an embodiment of the present disclosure;
FIG. 4 is a schematic diagram of a qualified sample and a non-qualified sample according to an embodiment of the present application;
FIG. 5 is a schematic diagram illustrating the classification of a rejected biochemical sample according to an embodiment of the present application;
FIG. 6a is a schematic representation of a ROC curve corresponding to one hemolysis classification provided in an example of the present application;
fig. 6b is a schematic diagram of an ROC curve corresponding to a jaundice classification according to an embodiment of the present disclosure;
FIG. 6c is a schematic diagram of a ROC curve corresponding to a kind of lipemia classification provided in the present application;
FIG. 6d is a schematic diagram of an ROC curve corresponding to the sum of classification probabilities of hemolysis, jaundice, and lipemia according to an embodiment of the present application;
fig. 7 is a schematic structural diagram of an identification device according to an embodiment of the present application;
fig. 8 is a schematic structural diagram of an identification device according to an embodiment of the present application.
Detailed Description
In order to facilitate understanding of the technical solutions provided by the embodiments of the present application, the technical solutions of the present application are described in detail below with reference to the accompanying drawings.
The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. Rather, they are merely examples of methods consistent with aspects of the present application, as detailed in the appended claims.
Before describing the embodiments of the present application, some terms in the present application will be explained to facilitate understanding for those skilled in the art.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used in this application and the appended claims, the singular forms "a", "an", and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items. It is also to be understood that the term "at least one" as used herein includes one or more, "a plurality" includes two and more.
Unless otherwise stated, the embodiments of the present application refer to the ordinal numbers "first" to "third" for distinguishing a plurality of objects, and do not limit the sequence, timing, priority, or importance of the plurality of objects.
Any of the biochemical samples referred to herein includes both plasma and serum fractions. Illustratively, referring to FIG. 4, a biochemical sample may include two portions, plasma and serum.
The sample serum quality identification method based on deep learning provided by the embodiment of the present application will be specifically described below with reference to fig. 1 to 6 d.
Please refer to fig. 1, which is a schematic flow chart of a sample serum quality identification method based on deep learning according to an embodiment of the present application. The main execution body of the method flow shown in fig. 1 is an identification device. As shown in fig. 1, the method flow may include the following steps:
s101, acquiring a preprocessed biochemical sample image data set.
In some embodiments, the preprocessed biochemical sample image dataset may include: and the plurality of biochemical sample images and the serum indexes corresponding to the plurality of biochemical sample images respectively. The serum indexes corresponding to the multiple biochemical sample images can be used for representing the serum quality of the biochemical samples corresponding to the multiple biochemical sample images respectively.
In some embodiments, as shown in fig. 2, executing step S101 may specifically include the following steps:
s201, acquiring an original biochemical sample image data set.
In some embodiments, the plurality of images of biochemical samples and the respective serum indices of the plurality of images of biochemical samples for a hospital or a plurality of hospitals over a certain time period may be collected as the raw biochemical sample image dataset. For example, 10667 images of a biochemical sample and 10667 images of a biochemical sample of a hospital emergency within three months may each be collected as a raw biochemical sample image dataset.
S202, marking the biochemical sample image with interference information in the serum part in the original biochemical sample image data set manually to obtain the biochemical sample image data set marked with the interference information.
In some embodiments, the interference information may include, but is not limited to: label, handwriting, sticker, etc.
In some embodiments, the biochemical sample image with interference information in the serum part of the original biochemical sample image data set can be artificially marked, so that the degree of influence of the interference information on the sensitivity of the deep convolutional neural network model and the interference resistance of the deep convolutional neural network model can be judged when the deep convolutional neural network model is trained according to the obtained biochemical sample image data set marked with the interference information.
And S203, setting the image resolution of the biochemical sample images in the biochemical sample image data set with the marked interference information to be N M Z, wherein N, M, Z is an integer larger than 1, and obtaining the preprocessed biochemical sample image data set.
Illustratively, N × M × Z may be set to 120 × 500 × 32.
In the embodiment of the application, the image resolution of the biochemical sample image in the biochemical sample image data set after the marking of the interference information is set to be N M Z, so that the deep convolutional neural network model can be conveniently trained subsequently.
S102, constructing a deep convolutional neural network model framework, and performing learning training on the deep convolutional neural network model framework based on the preprocessed biochemical sample image data set to obtain a deep convolutional neural network model.
In some embodiments, a 1 × 1 convolution kernel in combination with a residual network may be used to construct an initial deep convolutional neural network model framework of 782 layers, and then a global average pool 2D layer and a final output layer are added after the initial convolutional neural network model framework to construct the deep convolutional neural network model framework. Wherein, the global average pool 2D layer can be used for outputting a feature map of an input biochemical sample image of the deep convolutional neural network model. The final output layer may be used to output the probability of hemolysis, jaundice, and lipemia of the biochemical sample.
In some embodiments, as shown in fig. 3, after the deep convolutional neural network model framework is constructed, the learning training process for the deep convolutional neural network model framework may include the following steps:
s301, randomly dividing the preprocessed biochemical sample image data set into a biochemical sample training data set and a biochemical sample verification data set according to a preset proportion.
In some embodiments, the preset ratio may be 8: 2.
In some embodiments, the preprocessed biochemical sample image dataset may be automatically randomly divided into a biochemical sample training dataset and a biochemical sample validation dataset according to a preset ratio by Keras (an open source artificial neural network library written by Python).
S302, training the deep convolutional neural network model frame based on the biochemical sample training data set to obtain an initial deep convolutional neural network model.
In some embodiments, a tensortow system (a symbolic mathematical system based on data stream programming) can be employed as the back-end of the deep convolutional neural network model framework for learning training.
In a specific implementation, system parameters of the tensorfw system may be set. For example, the initial learning rate of the tensorfow system can be set to 0.0001, the iteration parameter of the initial learning rate is 1/2 iterated every 10 cycles (epoch), and then the training time of the deep convolutional neural network model framework can be 120 cycles, and the algorithm iteration time is short.
In a specific implementation process, the image enhancement processing may be performed on the biochemical sample training data set to obtain a processed biochemical sample training data set. For example, the image enhancement processing may be performed on the training data set of the biochemical sample by using a geometric transformation (including translation, inversion, rotation, scaling, etc.) to obtain a processed training data set of the biochemical sample. In the geometric transformation method, one or more combinations of image enhancement methods such as rotation/reflection transformation, inversion transformation, scaling transformation, translation transformation, scale transformation, contrast transformation, noise disturbance (salt and pepper noise and gaussian noise can be adopted), and miscut transformation can be adopted. Of course, other image enhancement methods may also be used to perform image enhancement processing on the biochemical sample training data set, and the embodiment of the present application is not limited. Other image enhancement processing methods may include, but are not limited to: random brightness adjustment methods, random contrast adjustment, etc. In a specific implementation process, an imagedata generator function in an open source code library Keras can be adopted to implement image enhancement processing on a biochemical sample training data set by adopting a corresponding image enhancement method. It should be understood that the processed biochemical sample training data set contains a greater number of biochemical sample images than the biochemical sample training data set.
In a specific implementation process, a deep convolutional neural network model framework can be trained on a tensorfw system based on a processed biochemical sample training data set, so as to obtain an initial deep convolutional neural network model.
For example, as shown in FIG. 5, biochemical samples may have overlapping parts when they are subjected to hemolysis, jaundice and lipid-blood classification, for example, a biochemical sample may be a hemolysis-combined jaundice sample. In the embodiment of the application, a deep convolutional neural network model framework can be trained on a tensorhow system based on a processed biochemical sample training data set, and probabilities of hemolysis, jaundice and lipemia corresponding to a biochemical sample image in the processed biochemical sample training data set are obtained. Then, a sum of probabilities between the probabilities of hemolysis, jaundice, and lipemia corresponding to the biochemical sample images in the processed biochemical sample training dataset may be determined based on a preset classification network (e.g., a two-classification network of Sigmoid activation functions). Then, based on the probability sum, model parameters of a deep convolutional neural network model framework for judging hemolysis, jaundice and lipemia can be determined, and an initial deep convolutional neural network model is obtained, so that the sensitivity of the deep convolutional neural network model can be improved, and the anti-interference capability of the deep convolutional neural network model can be enhanced.
For example, in the course of learning and training by using the deep convolutional neural network model framework provided in the embodiment of the present application with 10667 biochemical sample images as described above, the ROC curve corresponding to hemolysis probability may be shown in fig. 6a, the ROC curve corresponding to jaundice probability may be shown in fig. 6b, the ROC curve corresponding to lipemia probability may be shown in fig. 6c, and the ROC curve corresponding to calculating the sum of the probabilities of hemolysis, jaundice, and lipemia may be shown in fig. 6 d.
As shown in fig. 6a to 6d, when the deep convolutional neural network model provided by the embodiment of the present application is used for identifying the serum quality of a biochemical sample image, although interference information exists in a biochemical sample image data set used for learning and training the deep convolutional neural network model, higher sensitivity and specificity can be obtained, which indicates that the interference information has little influence on the sensitivity of the deep convolutional neural network model and the anti-interference capability of the enhanced deep convolutional neural network model, and the anti-interference capability of the deep convolutional neural network model is better.
And S303, respectively inputting the biochemical sample images in the biochemical sample verification data set into the initial deep convolutional neural network model, and performing probability precision verification on the initial deep convolutional neural network model.
In a specific implementation process, a probability accuracy threshold of the initial deep convolutional neural network model may be set. When the biochemical sample images in the biochemical sample verification data set are respectively input into the initial deep convolutional neural network model, the probabilities of hemolysis, jaundice and lipemia corresponding to the second biochemical sample corresponding to the biochemical sample image in the biochemical sample verification data set can be predicted and obtained. Then, the probability of hemolysis, jaundice and lipemia corresponding to the second biochemical sample obtained based on the serum index actual judgment of the hemolysis, jaundice and lipemia corresponding to the second biochemical sample is compared with the predicted probability of hemolysis, jaundice and lipemia corresponding to the second biochemical sample, and the probability accuracy of the initial deep convolutional neural network model can be obtained. And then, comparing the probability precision of the initial deep convolutional neural network model with a probability precision threshold value, and verifying whether the probability precision of the initial deep convolutional neural network model meets a preset requirement.
And S304, if the probability precision of the initial deep convolutional neural network model meets the preset requirement, taking the initial deep convolutional neural network model as the deep convolutional neural network model.
In a specific implementation, an allowable error range of the probability accuracy may be set. If the error between the probability accuracy of the deep convolutional neural network model and the probability accuracy threshold is within the allowable error range, the probability accuracy of the initial deep convolutional neural network can be determined to pass verification, and the preset requirement is met.
In a specific implementation process, if the error between the probability accuracy of the deep convolutional neural network model and the probability accuracy threshold is not within the allowable error range, it can be determined that the probability accuracy of the initial deep convolutional neural network does not pass verification and does not meet the preset requirement. When the probability accuracy of the initial deep convolutional neural network does not pass the verification, the step S101 may be executed back until the probability accuracy of the initial deep convolutional neural network passes the verification.
S103, acquiring a biochemical sample image to be identified.
It should be noted that, in a specific implementation process, the embodiment of the present application does not limit the execution sequence between steps S101 to S102 and step S103, for example, the identification device may execute steps S101 to S102 first and then step S103, or may execute step S103 first and then step S101 to S102, or may execute steps S101 to S102 and step S103 at the same time.
S104, inputting the biochemical sample image to be identified into the deep convolutional neural network model, and obtaining the probability of hemolysis, jaundice and lipemia of the first biochemical sample corresponding to the biochemical sample image to be identified.
And S105, determining the judgment condition of hemolysis, jaundice and lipemia corresponding to the first biochemical sample based on the probability of hemolysis, jaundice and lipemia of the first biochemical sample.
In some embodiments, the determination may include a determination that the first biochemical sample is a qualified biochemical sample or a non-qualified biochemical sample, and whether the first biochemical sample is a biochemical sample that combines one or more of hemolysis, jaundice, and lipemia when the first biochemical sample is a non-qualified biochemical sample.
In a specific implementation process, the probabilities of hemolysis, jaundice, and lipemia of the first biochemical sample may be compared with preset threshold probabilities of hemolysis, jaundice, and lipemia, respectively. When it is determined that one or more of the probabilities of hemolysis, jaundice, and lipemia for the first biochemical sample is greater than or equal to the respective probability threshold, the first biochemical sample may be determined to be a combined biochemical sample of one or more of hemolysis, jaundice, and lipemia. When it is determined that any of the probabilities of hemolysis, jaundice, and lipemia for the first biochemical sample is less than the respective probability threshold, the first biochemical sample may be determined to be a combined biochemical sample of one or more of hemolysis, jaundice, and lipemia.
As can be seen from the above description, in the technical solution provided in the embodiment of the present application, the obtaining of the preprocessed biochemical sample image data set includes: the serum indexes of the multiple biochemical sample images and the multiple biochemical sample images respectively correspond to each other; constructing a deep convolutional neural network model frame, and performing learning training on the deep convolutional neural network model frame based on the preprocessed biochemical sample image data set to obtain a deep convolutional neural network model; acquiring a biochemical sample image to be identified, inputting the biochemical sample image to be identified into a deep convolutional neural network model, and acquiring the probability of hemolysis, jaundice and lipemia of a first biochemical sample corresponding to the biochemical sample image to be identified; and determining the judgment condition of hemolysis, jaundice and lipemia corresponding to the first biochemical sample based on the probability of hemolysis, jaundice and lipemia of the first biochemical sample. Through the mode, when the serum quality of the biochemical sample image is identified, the sensitivity and the specificity of identifying the serum quality can be improved, so that good anti-interference capability can be obtained, and the identification effect of the serum quality can be improved.
Based on the same inventive concept, an embodiment of the present application further provides an identification device, as shown in fig. 7, the identification device 600 may include:
a processing unit 601, configured to: obtaining a preprocessed biochemical sample image dataset, the preprocessed biochemical sample image dataset comprising: a plurality of biochemical sample images and serum indexes corresponding to the plurality of biochemical sample images respectively;
constructing a deep convolutional neural network model frame, and performing learning training on the deep convolutional neural network model frame based on the preprocessed biochemical sample image data set to obtain a deep convolutional neural network model;
a determining unit 602, configured to: acquiring a biochemical sample image to be identified, inputting the biochemical sample image to be identified into the deep convolutional neural network model, and acquiring the probability of hemolysis, jaundice and lipemia of a first biochemical sample corresponding to the biochemical sample image to be identified; determining the judgment condition of hemolysis, jaundice and lipemia corresponding to the first biochemical sample based on the probability of hemolysis, jaundice and lipemia of the first biochemical sample; and determining the first biochemical sample to be a qualified sample or an unqualified sample based on the judgment condition.
In one possible design, the processing unit 601 is specifically configured to:
randomly dividing the preprocessed biochemical sample image data set into a biochemical sample training data set and a biochemical sample verification data set according to a preset proportion;
training the deep convolutional neural network model framework based on the biochemical sample training data set to obtain an initial deep convolutional neural network model;
respectively inputting the biochemical sample images in the biochemical sample verification data set into the initial deep convolutional neural network model, and performing probability precision verification on the initial deep convolutional neural network model;
and if the probability precision of the initial deep convolutional neural network model meets the preset requirement, taking the initial deep convolutional neural network model as the deep convolutional neural network model.
In one possible design, the processing unit 601 is specifically configured to:
setting system parameters of a tensorfow system, wherein the system parameters comprise an initial learning rate and iteration parameters of the initial learning rate;
performing image enhancement processing on the biochemical sample training data set to obtain a processed biochemical sample training data set;
training the deep convolutional neural network model framework on the tensortow system based on the processed biochemical sample training data set to obtain the initial deep convolutional neural network model.
In one possible design, the processing unit 601 is specifically configured to:
training the deep convolutional neural network model framework on the tensortow system based on the processed biochemical sample training data set to obtain the probabilities of hemolysis, jaundice and lipemia corresponding to the biochemical sample images in the processed biochemical sample training data set;
determining the probability sum of the probabilities of hemolysis, jaundice and lipemia corresponding to the biochemical sample images in the processed biochemical sample training data set based on a preset classification network;
and determining model parameters of the deep convolutional neural network model framework for judging hemolysis, jaundice and lipemia based on the probability sum to obtain the initial deep convolutional neural network model.
In one possible design, the predetermined classification network is a two-class network of Sigmoid activation functions.
In one possible design, the processing unit 601 is specifically configured to:
constructing 782 layers of initial deep convolutional neural network model frames by combining 1-by-1 convolutional kernels with a residual error network;
adding a global average pool (2D) layer and a final output layer behind the initial convolutional neural network model framework to construct the deep convolutional neural network model framework; wherein the content of the first and second substances,
the global average pool 2D layer is used for outputting a feature map of an input biochemical sample image of the deep convolutional neural network model, and the final output layer is used for outputting probabilities of hemolysis, jaundice and lipemia of the biochemical sample.
In one possible design, the processing unit 601 is specifically configured to:
acquiring an original biochemical sample image dataset;
manually marking a biochemical sample image with interference information in a serum part in an original biochemical sample image dataset to obtain a biochemical sample image dataset marked with the interference information;
and setting the image resolution of the biochemical sample images in the biochemical sample image dataset after marking the interference information to be N M Z, wherein N, M, Z is an integer larger than 1, and obtaining the preprocessed biochemical sample image dataset.
The identification device 600 in the embodiment of the present application and the sample serum quality identification method based on deep learning shown in fig. 1 are based on the same concept, and through the foregoing detailed description of the sample serum quality identification method based on deep learning, a person skilled in the art can clearly understand the implementation process of the identification device 600 in the embodiment, so for brevity of the description, details are not repeated here.
Based on the same inventive concept, an embodiment of the present application further provides an identification device, as shown in fig. 8, an identification device 700 may include: at least one memory 701 and at least one processor 702. Wherein:
the at least one memory 701 is used to store one or more programs.
The one or more programs, when executed by the at least one processor 702, implement the method for sample serum quality identification based on deep learning illustrated in fig. 1 and described above.
The identification device 700 may also preferably include a communication interface (not shown in fig. 8) for communicating with external devices and data exchange.
It should be noted that the memory 701 may include a high-speed RAM memory, and may also include a nonvolatile memory (nonvolatile memory), such as at least one disk memory.
In a specific implementation process, if the memory, the processor and the communication interface are integrated on one chip, the memory, the processor and the communication interface can complete mutual communication through the internal interface. If the memory, the processor and the communication interface are implemented independently, the memory, the processor and the communication interface may be connected to each other through a bus and perform communication with each other.
Based on the same inventive concept, the present application further provides a computer-readable storage medium, where at least one program is stored, and when the at least one program is executed by a processor, the method for identifying serum quality of a sample based on deep learning shown in fig. 1 is implemented.
It should be understood that the computer-readable storage medium is any data storage device that can store data or programs which can thereafter be read by a computer system. Examples of computer-readable storage media include: read-only memory, random access memory, CD-ROM, HDD, DVD, magnetic tape, optical data storage devices, and the like.
The computer readable storage medium can also be distributed over network coupled computer systems so that the computer readable code is stored and executed in a distributed fashion.
Program code embodied on a computer readable storage medium may be transmitted using any appropriate medium, including but not limited to: wireless, wire, fiber optic cable, Radio Frequency (RF), etc., or any suitable combination of the foregoing.
The above examples only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the claims. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application.
Claims (10)
1. A sample serum quality identification method based on deep learning is characterized by comprising the following steps:
obtaining a preprocessed biochemical sample image dataset, the preprocessed biochemical sample image dataset comprising: a plurality of biochemical sample images and serum indexes corresponding to the plurality of biochemical sample images respectively;
constructing a deep convolutional neural network model frame, and performing learning training on the deep convolutional neural network model frame based on the preprocessed biochemical sample image data set to obtain a deep convolutional neural network model;
acquiring a biochemical sample image to be identified, inputting the biochemical sample image to be identified into the deep convolutional neural network model, and acquiring the probability of hemolysis, jaundice and lipemia of a first biochemical sample corresponding to the biochemical sample image to be identified;
and determining the judgment condition of hemolysis, jaundice and lipemia corresponding to the first biochemical sample based on the probability of hemolysis, jaundice and lipemia of the first biochemical sample.
2. The method of claim 1, wherein learning training the deep convolutional neural network model framework based on the preprocessed biochemical sample image dataset to obtain a deep convolutional neural network model comprises:
randomly dividing the preprocessed biochemical sample image data set into a biochemical sample training data set and a biochemical sample verification data set according to a preset proportion;
training the deep convolutional neural network model framework based on the biochemical sample training data set to obtain an initial deep convolutional neural network model;
respectively inputting the biochemical sample images in the biochemical sample verification data set into the initial deep convolutional neural network model, and performing probability precision verification on the initial deep convolutional neural network model;
and if the probability precision of the initial deep convolutional neural network model meets the preset requirement, taking the initial deep convolutional neural network model as the deep convolutional neural network model.
3. The method of claim 2, wherein training the deep convolutional neural network model framework based on the biochemical sample training dataset to obtain an initial deep convolutional neural network model comprises:
setting system parameters of a tensorfow system, wherein the system parameters comprise an initial learning rate and iteration parameters of the initial learning rate;
performing image enhancement processing on the biochemical sample training data set to obtain a processed biochemical sample training data set;
training the deep convolutional neural network model framework on the tensortow system based on the processed biochemical sample training data set to obtain the initial deep convolutional neural network model.
4. The method of claim 3, wherein training the deep convolutional neural network model framework on the tensorfow system based on the processed biochemical sample training dataset to obtain the initial deep convolutional neural network model comprises:
training the deep convolutional neural network model framework on the tensortow system based on the processed biochemical sample training data set to obtain the probabilities of hemolysis, jaundice and lipemia corresponding to the biochemical sample images in the processed biochemical sample training data set;
determining the probability sum of the probabilities of hemolysis, jaundice and lipemia corresponding to the biochemical sample images in the processed biochemical sample training data set based on a preset classification network;
and determining model parameters of the deep convolutional neural network model framework for judging hemolysis, jaundice and lipemia based on the probability sum to obtain the initial deep convolutional neural network model.
5. The method of claim 2, wherein the predetermined classification network is a Sigmoid-enabled binary network.
6. The method of any one of claims 1-5, wherein constructing a deep convolutional neural network model framework comprises:
constructing 782 layers of initial deep convolutional neural network model frames by combining 1-by-1 convolutional kernels with a residual error network;
adding a global average pool (2D) layer and a final output layer behind the initial convolutional neural network model framework to construct the deep convolutional neural network model framework; wherein the content of the first and second substances,
the global average pool 2D layer is used for outputting a feature map of an input biochemical sample image of the deep convolutional neural network model, and the final output layer is used for outputting probabilities of hemolysis, jaundice and lipemia of the biochemical sample.
7. The method of any one of claims 1-5, wherein obtaining the pre-processed biochemical sample image dataset comprises:
acquiring an original biochemical sample image dataset;
manually marking a biochemical sample image with interference information in a serum part in an original biochemical sample image dataset to obtain a biochemical sample image dataset marked with the interference information;
and setting the image resolution of the biochemical sample images in the biochemical sample image dataset after marking the interference information to be N M Z, wherein N, M, Z is an integer larger than 1, and obtaining the preprocessed biochemical sample image dataset.
8. An identification device, comprising:
a processing unit to: obtaining a preprocessed biochemical sample image dataset, the preprocessed biochemical sample image dataset comprising: a plurality of biochemical sample images and serum indexes corresponding to the plurality of biochemical sample images respectively;
constructing a deep convolutional neural network model frame, and performing learning training on the deep convolutional neural network model frame based on the preprocessed biochemical sample image data set to obtain a deep convolutional neural network model;
a determination unit configured to: acquiring a biochemical sample image to be identified, inputting the biochemical sample image to be identified into the deep convolutional neural network model, and acquiring the probability of hemolysis, jaundice and lipemia of a first biochemical sample corresponding to the biochemical sample image to be identified; determining the judgment condition of hemolysis, jaundice and lipemia corresponding to the first biochemical sample based on the probability of hemolysis, jaundice and lipemia of the first biochemical sample; and determining the first biochemical sample to be a qualified sample or an unqualified sample based on the judgment condition.
9. An identification device, comprising: at least one memory and at least one processor;
the at least one memory is for storing one or more programs;
the one or more programs, when executed by the at least one processor, implement the method of any of claims 1-7.
10. A computer-readable storage medium characterized in that the computer-readable storage medium stores at least one program; the at least one program, when executed by a processor, implements the method of any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110910127.7A CN113592842B (en) | 2021-08-09 | 2021-08-09 | Sample serum quality identification method and identification equipment based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110910127.7A CN113592842B (en) | 2021-08-09 | 2021-08-09 | Sample serum quality identification method and identification equipment based on deep learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113592842A true CN113592842A (en) | 2021-11-02 |
| CN113592842B CN113592842B (en) | 2024-05-24 |
Family
ID=78256524
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110910127.7A Active CN113592842B (en) | 2021-08-09 | 2021-08-09 | Sample serum quality identification method and identification equipment based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113592842B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114638803A (en) * | 2022-03-15 | 2022-06-17 | 四川大学华西医院 | Serum index intelligent interpretation method and system based on deep learning |
| CN114878844A (en) * | 2022-05-20 | 2022-08-09 | 上海捷程医学科技有限公司 | Method, system and equipment for automatically detecting quality of centrifuged blood sample |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109766780A (en) * | 2018-12-20 | 2019-05-17 | 武汉理工大学 | A kind of ship smog emission on-line checking and method for tracing based on deep learning |
| CN110199172A (en) * | 2016-11-14 | 2019-09-03 | 美国西门子医学诊断股份有限公司 | For detecting the haemolysis of sample, jaundice, piarhemia or method, equipment and the quality inspection module of normality |
| CN110573883A (en) * | 2017-04-13 | 2019-12-13 | 美国西门子医学诊断股份有限公司 | Method and apparatus for determining tag count during sample characterization |
| CN110573859A (en) * | 2017-04-13 | 2019-12-13 | 美国西门子医学诊断股份有限公司 | Method and apparatus for HILN characterization using convolutional neural networks |
| CN111223088A (en) * | 2020-01-16 | 2020-06-02 | 东南大学 | Casting surface defect identification method based on deep convolutional neural network |
| CN111462076A (en) * | 2020-03-31 | 2020-07-28 | 湖南国科智瞳科技有限公司 | Method and system for detecting fuzzy area of full-slice digital pathological image |
| US20210064927A1 (en) * | 2018-01-10 | 2021-03-04 | Siemens Healthcare Diagnostics Inc. | Methods and apparatus for bio-fluid specimen characterization using neural network having reduced training |
| CN112639482A (en) * | 2018-06-15 | 2021-04-09 | 美国西门子医学诊断股份有限公司 | Sample container characterization using single depth neural networks in an end-to-end training manner |
| CN112673261A (en) * | 2018-09-20 | 2021-04-16 | 美国西门子医学诊断股份有限公司 | Method and apparatus for HILN determination using a deep adaptation network for both serum and plasma sampling |
-
2021
- 2021-08-09 CN CN202110910127.7A patent/CN113592842B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110199172A (en) * | 2016-11-14 | 2019-09-03 | 美国西门子医学诊断股份有限公司 | For detecting the haemolysis of sample, jaundice, piarhemia or method, equipment and the quality inspection module of normality |
| US20190277870A1 (en) * | 2016-11-14 | 2019-09-12 | Siemens Healthcare Diagnostics Inc. | Methods, apparatus, and quality check modules for detecting hemolysis, icterus, lipemia, or normality of a specimen |
| CN110573883A (en) * | 2017-04-13 | 2019-12-13 | 美国西门子医学诊断股份有限公司 | Method and apparatus for determining tag count during sample characterization |
| CN110573859A (en) * | 2017-04-13 | 2019-12-13 | 美国西门子医学诊断股份有限公司 | Method and apparatus for HILN characterization using convolutional neural networks |
| US20200151498A1 (en) * | 2017-04-13 | 2020-05-14 | Siemens Healthcare Diagnostics Inc. | Methods and apparatus for hiln characterization using convolutional neural network |
| US20210064927A1 (en) * | 2018-01-10 | 2021-03-04 | Siemens Healthcare Diagnostics Inc. | Methods and apparatus for bio-fluid specimen characterization using neural network having reduced training |
| CN112639482A (en) * | 2018-06-15 | 2021-04-09 | 美国西门子医学诊断股份有限公司 | Sample container characterization using single depth neural networks in an end-to-end training manner |
| CN112673261A (en) * | 2018-09-20 | 2021-04-16 | 美国西门子医学诊断股份有限公司 | Method and apparatus for HILN determination using a deep adaptation network for both serum and plasma sampling |
| CN109766780A (en) * | 2018-12-20 | 2019-05-17 | 武汉理工大学 | A kind of ship smog emission on-line checking and method for tracing based on deep learning |
| CN111223088A (en) * | 2020-01-16 | 2020-06-02 | 东南大学 | Casting surface defect identification method based on deep convolutional neural network |
| CN111462076A (en) * | 2020-03-31 | 2020-07-28 | 湖南国科智瞳科技有限公司 | Method and system for detecting fuzzy area of full-slice digital pathological image |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114638803A (en) * | 2022-03-15 | 2022-06-17 | 四川大学华西医院 | Serum index intelligent interpretation method and system based on deep learning |
| CN114878844A (en) * | 2022-05-20 | 2022-08-09 | 上海捷程医学科技有限公司 | Method, system and equipment for automatically detecting quality of centrifuged blood sample |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113592842B (en) | 2024-05-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110807788B (en) | Medical image processing method, medical image processing device, electronic equipment and computer storage medium | |
| Kibriya et al. | A novel and effective brain tumor classification model using deep feature fusion and famous machine learning classifiers | |
| US11409992B2 (en) | Data slicing for machine learning performance testing and improvement | |
| CN113159147B (en) | Image recognition method and device based on neural network and electronic equipment | |
| CN112966074A (en) | Emotion analysis method and device, electronic equipment and storage medium | |
| CN113592842A (en) | Sample serum quality identification method and identification device based on deep learning | |
| WO2020190480A1 (en) | Classifying an input data set within a data category using multiple data recognition tools | |
| US11120297B2 (en) | Segmentation of target areas in images | |
| CN113065525A (en) | Age recognition model training method, face age recognition method and related device | |
| CN117011274A (en) | Automatic glass bottle detection system and method thereof | |
| Wang et al. | SERR‐U‐Net: Squeeze‐and‐Excitation Residual and Recurrent Block‐Based U‐Net for Automatic Vessel Segmentation in Retinal Image | |
| CN111161884A (en) | Disease prediction method, device, equipment and medium for unbalanced data | |
| CN114093435A (en) | Chemical molecule related water solubility prediction method based on deep learning | |
| CN113764101A (en) | CNN-based breast cancer neoadjuvant chemotherapy multi-modal ultrasonic diagnosis system | |
| CN113887559A (en) | Brain-computer information fusion classification method and system for brain off-loop application | |
| CN111582647A (en) | User data processing method and device and electronic equipment | |
| CN110826616A (en) | Information processing method and device, electronic equipment and storage medium | |
| US20220172370A1 (en) | Method for detecting white matter lesions based on medical image | |
| US20230143701A1 (en) | Systems and methods for predicting expression levels | |
| CN114649092A (en) | Auxiliary diagnosis method and device based on semi-supervised learning and multi-scale feature fusion | |
| CN113836297A (en) | Training method and device for text emotion analysis model | |
| US20230298326A1 (en) | Image augmentation method, electronic device and readable storage medium | |
| CN112989032A (en) | Entity relationship classification method, apparatus, medium and electronic device | |
| CN110689112A (en) | Data processing method and device | |
| CN116185843B (en) | Two-stage neural network testing method and device based on neuron coverage rate guidance |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |