CN113907710A - Skin lesion classification system based on model-independent image enhancement meta-learning - Google Patents
Skin lesion classification system based on model-independent image enhancement meta-learning Download PDFInfo
- Publication number
- CN113907710A CN113907710A CN202111153549.0A CN202111153549A CN113907710A CN 113907710 A CN113907710 A CN 113907710A CN 202111153549 A CN202111153549 A CN 202111153549A CN 113907710 A CN113907710 A CN 113907710A
- Authority
- CN
- China
- Prior art keywords
- meta
- model
- skin lesion
- data
- learning
- 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
- 206010040882 skin lesion Diseases 0.000 title claims abstract description 49
- 231100000444 skin lesion Toxicity 0.000 title claims abstract description 49
- 238000013145 classification model Methods 0.000 claims abstract description 21
- 201000001441 melanoma Diseases 0.000 claims abstract description 21
- 238000001514 detection method Methods 0.000 claims abstract description 17
- 238000000034 method Methods 0.000 claims abstract description 15
- 238000007781 pre-processing Methods 0.000 claims abstract description 12
- 238000012549 training Methods 0.000 claims description 47
- 238000012360 testing method Methods 0.000 claims description 35
- 230000006870 function Effects 0.000 claims description 7
- 206010041823 squamous cell carcinoma Diseases 0.000 claims description 5
- 206010004146 Basal cell carcinoma Diseases 0.000 claims description 4
- 206010020649 Hyperkeratosis Diseases 0.000 claims description 4
- 208000001126 Keratosis Diseases 0.000 claims description 4
- 208000009621 actinic keratosis Diseases 0.000 claims description 4
- 238000011176 pooling Methods 0.000 claims description 4
- 206010053717 Fibrous histiocytoma Diseases 0.000 claims description 3
- 208000009077 Pigmented Nevus Diseases 0.000 claims description 3
- 230000004913 activation Effects 0.000 claims description 3
- 208000001119 benign fibrous histiocytoma Diseases 0.000 claims description 3
- 201000010305 cutaneous fibrous histiocytoma Diseases 0.000 claims description 3
- 238000005516 engineering process Methods 0.000 claims description 3
- 201000002699 melanoma in congenital melanocytic nevus Diseases 0.000 claims description 3
- 231100000216 vascular lesion Toxicity 0.000 claims description 3
- 239000003623 enhancer Substances 0.000 claims 9
- 208000017520 skin disease Diseases 0.000 abstract description 6
- 230000006978 adaptation Effects 0.000 abstract 1
- 238000013528 artificial neural network Methods 0.000 abstract 1
- 238000010586 diagram Methods 0.000 description 9
- 238000004590 computer program Methods 0.000 description 7
- 238000012545 processing Methods 0.000 description 5
- 238000004458 analytical method Methods 0.000 description 3
- 230000003902 lesion Effects 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 206010028980 Neoplasm Diseases 0.000 description 2
- 201000011510 cancer Diseases 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 238000003745 diagnosis Methods 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 206010016629 fibroma Diseases 0.000 description 2
- 210000003491 skin Anatomy 0.000 description 2
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 1
- 206010059245 Angiopathy Diseases 0.000 description 1
- 206010027145 Melanocytic naevus Diseases 0.000 description 1
- 208000007256 Nevus Diseases 0.000 description 1
- 208000037014 Rare skin disease Diseases 0.000 description 1
- 208000000453 Skin Neoplasms Diseases 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004195 computer-aided diagnosis Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 210000002752 melanocyte Anatomy 0.000 description 1
- 239000012528 membrane Substances 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
- 230000008569 process Effects 0.000 description 1
- 201000000849 skin cancer Diseases 0.000 description 1
- 230000004083 survival effect Effects 0.000 description 1
- 208000024891 symptom Diseases 0.000 description 1
- 238000011282 treatment Methods 0.000 description 1
- 238000011269 treatment regimen Methods 0.000 description 1
- 238000009827 uniform distribution Methods 0.000 description 1
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/44—Detecting, measuring or recording for evaluating the integumentary system, e.g. skin, hair or nails
- A61B5/441—Skin evaluation, e.g. for skin disorder diagnosis
- A61B5/443—Evaluating skin constituents, e.g. elastin, melanin, water
-
- 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)
- Artificial Intelligence (AREA)
- Physics & Mathematics (AREA)
- Animal Behavior & Ethology (AREA)
- Pathology (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Biophysics (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Veterinary Medicine (AREA)
- Dermatology (AREA)
- Evolutionary Computation (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physiology (AREA)
- Psychiatry (AREA)
- Signal Processing (AREA)
- Image Analysis (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
Abstract
The invention provides a skin lesion classification system based on model-independent image enhancement meta-learning, comprising a data acquisition module, a classification module and a classification module, wherein the data acquisition module is configured to acquire melanoma detection data; a data pre-processing module configured to regularize melanoma detection data using image enhancement techniques; the data classification module is configured to classify the preprocessed data by using a trained skin lesion classification model irrelevant to the image enhancement element learning to obtain a classification result; the invention selects model-independent meta-learning small sample classification to improve the generalization capability of the classification model and the confidence coefficient of the model, and utilizes meta-learning and image enhancement to promote the rapid adaptation and generalization of a deep neural network trained on the basis of common skin disease data so as to identify skin diseases with less annotation data and improve the classification accuracy and the generalization capability of classification.
Description
Technical Field
The invention belongs to the technical field of machine learning, and particularly relates to a skin lesion classification system based on model-independent image enhancement meta-learning.
Background
The statements in this section merely provide background information related to the present disclosure and may not necessarily constitute prior art.
If skin lesions can be diagnosed early and appropriate treatment is selected, the survival rate is very high. Therefore, it is absolutely necessary to know as early as possible whether the patient's symptoms correspond to cancer. Conventionally, doctors have been detecting skin cancer using naked eyes, especially when the cancer is still in an early stage, which is difficult to confirm by even experts.
Therefore, computer-aided diagnosis is important for improving the treatment strategy of skin diseases, and deep learning for solving the diagnosis task of skin diseases is very effective. At present, the number of annotation images for diagnosing rare skin diseases is still small, and the distribution of long-tail classes often appears in a skin disease data set, so that the generalization degree of the distributed tail classes is poor.
Disclosure of Invention
The invention aims to solve the problems and provides a skin lesion classification system based on model-independent image enhancement meta-learning.
According to some embodiments, the invention adopts the following technical scheme:
a system for skin lesion classification based on model-independent image-enhanced meta-learning, comprising:
a data acquisition module configured to acquire melanoma detection data;
a data pre-processing module configured to regularize melanoma detection data using image enhancement techniques;
and the data classification module is configured to classify the preprocessed data by using the trained skin lesion classification model of the image enhancement meta-learning irrelevant to the model to obtain a classification result.
As an alternative embodiment, the training of the trained skin lesion classification model comprises:
acquiring known melanoma detection data, and dividing the data into a meta-training set and a meta-testing set;
regularizing the meta-training set data by using an image enhancement technology;
constructing a meta-training task and a meta-testing task by utilizing the meta-training set and the meta-testing set;
performing gradient descent updating parameters for several times on a meta-training task, and constructing a skin lesion classification model of image enhancement meta-learning irrelevant to the model;
setting hyper-parameters of a skin lesion classification model;
and training the skin lesion classification model by using the meta-training set, and performing iterative test and fine adjustment on the trained model by using the meta-testing set until a test result meets a condition.
As a further limitation, when constructing the meta training task and the meta testing task by using the meta training set and the meta testing set, constructing the meta training task and the meta testing task by using the meta training set and the meta testing set according to an N-way K-shot mode; the meta training task and the meta testing task are respectively composed of two categories of a meta training set and a meta testing set which are randomly sampled; the meta-test set and the meta-training set both comprise a query set and a support set.
By way of further limitation, the hyper-parameters include meta-learning rate, batch size, gradient update steps, optimizer, and iteration number.
By way of further limitation, melanoma detection data is known to include melanoma, melanocytic nevi, basal cell carcinoma, actinic keratosis, benign keratosis, dermatofibroma, vascular lesions, and squamous cell carcinoma data, each of an order of magnitude.
As an alternative embodiment, the skin lesion classification model includes an input layer, a convolutional layer including four convolutional layers, an activation function, a pooling layer, a full-link layer, and an output layer.
As an alternative embodiment, the data preprocessing module is configured to perform sample enhancement on the image using CutOut, MixUp, or CutMix techniques when regularizing melanoma detection data using image enhancement techniques.
As an alternative embodiment, the data preprocessing module performs CutOut enhancement on the image, the CutOut enhancement randomly generates a square mask, and the pixel value in the generated mask is zero.
As an alternative embodiment, the data preprocessing module performs MixUp enhancement on the images, and creates a virtual training sample by randomly selecting one sample from the same batch of images and performing linear interpolation with another sample.
As an alternative embodiment, the data preprocessing module performs CutMix enhancement on the image, cuts off a segment of the image, pastes the cut segment of the image into different images among training images to generate a new sample, and proportionally mixes the background real label into the area of the cut region to generate the new sample.
Compared with the prior art, the invention has the beneficial effects that:
1. the present invention proposes to solve the problem of medical image dermatology classification lacking high quality annotation data using model-independent meta-learning and to simulate this problem by treating it as a small learning problem;
2. the invention uses advanced enhancement technology such as MixUp, CutOut and CutMix in the meta-training stage to standardize the model; the information added to model learning by conventional enhancement strategies such as turning, rotating and the like is very little, new virtual samples and classes can be generated by using some important enhancement methods such as MixUp and CutMix, and the image enhancement strategy is only used in the meta-training stage to avoid the over-fitting problem;
3. the model acquires the priori knowledge through training on the meta-training task, so that the skin lesion classification model can be quickly adapted to a new small sample classification task and has a better classification effect;
4. the invention greatly reduces the requirements of mass data collection and labeling based on deep learning in the medical field.
In order to make the aforementioned and other objects, features and advantages of the present invention comprehensible, preferred embodiments accompanied with figures are described in detail below.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, are included to provide a further understanding of the invention, and are incorporated in and constitute a part of this specification, illustrate exemplary embodiments of the invention and together with the description serve to explain the invention and not to limit the invention.
FIG. 1 is a schematic diagram of a rare skin lesion classification system based on image enhancement element learning according to the present invention;
FIG. 2 is a flow chart illustrating an implementation of a rare skin lesion classification system based on image enhancement meta-learning;
fig. 3 is a comparison of the image enhancement performed on the skin lesion image by the rare skin lesion classification system based on image enhancement meta-learning according to the present invention.
The specific implementation mode is as follows:
the invention is further described with reference to the following figures and examples.
It is to be understood that the following detailed description is exemplary and is intended to provide further explanation of the invention as claimed. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of exemplary embodiments according to the invention. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, and it should be understood that when the terms "comprises" and/or "comprising" are used in this specification, they specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof, unless the context clearly indicates otherwise.
As an exemplary embodiment, a skin lesion classification system based on model-independent image-enhanced meta-learning includes:
a data acquisition module configured to acquire melanoma detection data;
a data pre-processing module configured to regularize melanoma detection data using image enhancement techniques;
and the data classification module is configured to classify the preprocessed data by using the trained skin lesion classification model of the image enhancement meta-learning irrelevant to the model to obtain a classification result.
As an exemplary embodiment, a skin lesion classification system based on model-independent image enhancement meta-learning performs the following steps, as shown in fig. 2:
by adopting an ISIC-2019 skin lesion analysis melanoma detection data set, the skin lesion classification model of the embodiment utilizes prior knowledge to realize rapid and accurate rare skin lesion classification in a meta-test stage.
Analysis of melanoma detection data sets using ISIC-2019 skin lesions includes eight categories: melanoma, melanocytic nevi, basal cell carcinoma, actinic keratosis, benign keratosis, dermatofibroma, vasculopathy, squamous cell carcinoma. The number of images of the eight categories is 4522, 12875, 3323, 867, 2624, 239, 253, and 628, respectively. The method comprises the steps of dividing an ISIC-2019 skin lesion analysis melanoma detection data set into a meta-training set and a meta-testing set, using five lesions of melanoma, a melanocyte nevus, basal cell carcinoma, actinic keratosis and benign keratosis as the meta-training set, and using three lesions of skin fibroma, angiopathy and squamous cell carcinoma as the meta-testing set.
In the meta-training stage, CutOut enhancement is performed on the image, the CutOut enhancement randomly generates a square mask, and the pixel value in the generated mask is zero. In the case of CutOut, since features are deleted at the input stage, it can be guaranteed that no feature map contains features corresponding to the mask area.
The images are subjected to MixUp enhancement, which enhances the generalization capability of the deep learning model by generating virtual samples from a given data distribution. It creates virtual training samples by randomly selecting one sample (index i) from the same batch of images and linearly interpolating with another sample (index m (i)).
MixUp:
Wherein lambda is Beta (zeta ).
The images were subjected to CutMix:
the core idea of CutMix is to cut a section of image, paste it into different images between training images, generate a new sample, and proportionally mix the background real label into the area of the cut area. Suppose (x)i,yi) And (x)m(i),ym(i)) Are two samples. The new virtual samples are then:
x′＝F⊙xi+(1-F)⊙xm(i)
y′＝λyi+(1-λ)ym(i)
where λ β (ζ, ζ), F denote a randomly generated binary mask and indicate where to exit and fill in from the two images, an element intelligent multiplication. To generate the binary mask F, the invention first randomly generates a rectangular bounding box R ═ x1,y1,x2,y2B) represents the region x to be cropped in the two imagesiAnd xm(i). If W and H are the height and width, x, of the image, respectively1～Unifo rm(0,W),y1～Unifo rm(0,H),
And constructing a meta-training task and a meta-testing task by utilizing the meta-training set and the meta-testing set according to the N-way K-shot mode. The meta training task and the meta testing task are respectively composed of two categories of a meta training set and a meta testing set which are randomly sampled. In a 2-way K-shot manner, the support set is set to K samples per category, where K is set to 3, 5, and 10 in this embodiment.
The meta-test set and the meta-training set both comprise a query set and a support set.
The rare skin lesion classification model of the model-independent image enhancement meta-learning is constructed, and as shown in fig. 1, the model is composed of four convolutional layers, including an input layer, a convolutional layer, an activation function, a pooling layer, a full-link layer and an output layer. Each layer contains 32 kernels of size 3 x 3, followed by a 2 x 2 max pooling layer, step size 2.
Setting hyper-parameters of a skin lesion classification model, processing a skin lesion image into a size of 80 multiplied by 80, wherein the meta-learning rate is 0.001, the batch size is 5, an Adam optimizer is adopted, the iteration times are 10000, and the gradient updating step number is 4.
The rare skin lesion classification model of the model-independent image enhancement meta-learning utilizes the optimal initialization parameters to finely adjust the support set on the meta-test task of skin fibroma, vascular lesion and squamous cell carcinoma membrane lesion, and completes the classification of the model-independent image enhancement meta-learning on the rare skin lesion on the query set.
In this embodiment, advanced enhancement techniques, such as MixUp, CutOut, and CutMix, are used in the meta-training stage, as shown in fig. 3, so as to normalize the model, generate new virtual samples and classes, and only use the image enhancement strategy in the meta-training stage, so as to avoid the over-fitting problem; the model absorbs the capability of the system to adapt to new tasks and environments quickly, and the influence caused by insufficient sample data is relieved. The skin disease diagnosis performance is improved by utilizing image enhancement and meta-learning, and the rapid adaptability and the classification accuracy of the skin lesion classification model under low data are improved.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
Although the embodiments of the present invention have been described with reference to the accompanying drawings, it is not intended to limit the scope of the present invention, and it should be understood by those skilled in the art that various modifications and variations can be made without inventive efforts by those skilled in the art based on the technical solution of the present invention.
Claims (10)
1. A skin lesion classification system based on image enhancement meta-learning irrelevant to models is characterized in that: the method comprises the following steps:
a data acquisition module configured to acquire melanoma detection data;
a data pre-processing module configured to regularize melanoma detection data using image enhancement techniques;
and the data classification module is configured to classify the preprocessed data by using the trained skin lesion classification model of the image enhancement meta-learning irrelevant to the model to obtain a classification result.
2. The system of claim 1, wherein the skin lesion classification system based on model-independent image enhancer learning comprises: the training of the trained skin lesion classification model comprises:
acquiring known melanoma detection data, and dividing the data into a meta-training set and a meta-testing set;
regularizing the meta-training set data by using an image enhancement technology;
constructing a meta-training task and a meta-testing task by utilizing the meta-training set and the meta-testing set;
performing gradient descent updating parameters for several times on a meta-training task, and constructing a skin lesion classification model of image enhancement meta-learning irrelevant to the model;
setting hyper-parameters of a skin lesion classification model;
and training the skin lesion classification model by using the meta-training set, and performing iterative test and fine adjustment on the trained model by using the meta-testing set until a test result meets a condition.
3. The system of claim 2, wherein the skin lesion classification system based on model-independent image enhancer learning comprises: when constructing a meta-training task and a meta-testing task by using a meta-training set and a meta-testing set, constructing the meta-training task and the meta-testing task by using the meta-training set and the meta-testing set according to an N-way K-shot mode; the meta training task and the meta testing task are respectively composed of two categories of a meta training set and a meta testing set which are randomly sampled; the meta-test set and the meta-training set both comprise a query set and a support set.
4. The system of claim 2, wherein the skin lesion classification system based on model-independent image enhancer learning comprises: the hyper-parameters include meta-learning rate, batch size, gradient update steps, optimizer, and iteration number.
5. The system of claim 2, wherein the skin lesion classification system based on model-independent image enhancer learning comprises: melanoma detection data is known to include melanoma, melanocytic nevi, basal cell carcinoma, actinic keratosis, benign keratosis, dermatofibroma, vascular lesions, and squamous cell carcinoma data, each of an order of magnitude.
6. The system of claim 1, wherein the skin lesion classification system based on model-independent image enhancer learning comprises: the skin lesion classification model comprises an input layer, a convolution layer, an activation function, a pooling layer, a full-connection layer and an output layer, wherein the convolution layer comprises four convolution layers.
7. The system of claim 1, wherein the skin lesion classification system based on model-independent image enhancer learning comprises: the data preprocessing module is configured to perform sample enhancement on the image using CutOut, MixUp, or CutMix techniques when regularizing melanoma detection data using image enhancement techniques.
8. The system of claim 1, wherein the skin lesion classification system based on model-independent image enhancer learning comprises: the data preprocessing module carries out CutOut enhancement on the image, the CutOut enhancement randomly generates a square mask, and the pixel value in the generated mask is zero.
9. The system of claim 1, wherein the skin lesion classification system based on model-independent image enhancer learning comprises: the data preprocessing module performs MixUp enhancement on the images, and creates a virtual training sample by randomly selecting one sample from the same batch of images and performing linear interpolation on the other sample.
10. The system of claim 1, wherein the skin lesion classification system based on model-independent image enhancer learning comprises: the data preprocessing module carries out CutMix enhancement on the image, cuts off a section of image, pastes the cut section of image into different images among training images to generate a new sample, and proportionally mixes the background real label into the area of a cutting area to generate the new sample.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111153549.0A CN113907710A (en) | 2021-09-29 | 2021-09-29 | Skin lesion classification system based on model-independent image enhancement meta-learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111153549.0A CN113907710A (en) | 2021-09-29 | 2021-09-29 | Skin lesion classification system based on model-independent image enhancement meta-learning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113907710A true CN113907710A (en) | 2022-01-11 |
Family
ID=79236984
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111153549.0A Pending CN113907710A (en) | 2021-09-29 | 2021-09-29 | Skin lesion classification system based on model-independent image enhancement meta-learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113907710A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114663679A (en) * | 2022-05-25 | 2022-06-24 | 山东师范大学 | Blood coagulation index abnormity classification method based on feature fusion meta-learning |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108734138A (en) * | 2018-05-24 | 2018-11-02 | 浙江工业大学 | A kind of melanoma skin disease image classification method based on integrated study |
| CN110991402A (en) * | 2019-12-19 | 2020-04-10 | 湘潭大学 | Skin disease classification device and method based on deep learning |
| CN112861995A (en) * | 2021-03-15 | 2021-05-28 | 中山大学 | Unsupervised few-sample image classification method and system based on model independent meta learning and storage medium |
| CN113240113A (en) * | 2021-06-04 | 2021-08-10 | 北京富通东方科技有限公司 | Method for enhancing network prediction robustness |
-
2021
- 2021-09-29 CN CN202111153549.0A patent/CN113907710A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108734138A (en) * | 2018-05-24 | 2018-11-02 | 浙江工业大学 | A kind of melanoma skin disease image classification method based on integrated study |
| CN110991402A (en) * | 2019-12-19 | 2020-04-10 | 湘潭大学 | Skin disease classification device and method based on deep learning |
| CN112861995A (en) * | 2021-03-15 | 2021-05-28 | 中山大学 | Unsupervised few-sample image classification method and system based on model independent meta learning and storage medium |
| CN113240113A (en) * | 2021-06-04 | 2021-08-10 | 北京富通东方科技有限公司 | Method for enhancing network prediction robustness |
Non-Patent Citations (2)
| Title |
|---|
| RISHAV SINGH 等: "MetaMed: Few-shot medical image classification using gradient-based meta-learning", 《PATTERN RECOGNITION》, 17 June 2021 (2021-06-17), pages 2 - 4 * |
| XIAOMENG LI 等: "Di culty-aware Meta-learning for Rare Disease Diagnosis", 《SPRINGER》, 15 July 2020 (2020-07-15), pages 1 - 10 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114663679A (en) * | 2022-05-25 | 2022-06-24 | 山东师范大学 | Blood coagulation index abnormity classification method based on feature fusion meta-learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110689543A (en) | Improved convolutional neural network brain tumor image segmentation method based on attention mechanism | |
| Putra et al. | Enhanced skin condition prediction through machine learning using dynamic training and testing augmentation | |
| CN113256641B (en) | Skin lesion image segmentation method based on deep learning | |
| WO2021114130A1 (en) | Unsupervised self-adaptive mammary gland lesion segmentation method | |
| CN110619641A (en) | Automatic segmentation method of three-dimensional breast cancer nuclear magnetic resonance image tumor region based on deep learning | |
| CN111179237A (en) | Image segmentation method and device for liver and liver tumor | |
| CN112365980A (en) | Brain tumor multi-target point auxiliary diagnosis and prospective treatment evolution visualization method and system | |
| CN110414481A (en) | A kind of identification of 3D medical image and dividing method based on Unet and LSTM | |
| Popescu et al. | Retinal blood vessel segmentation using pix2pix gan | |
| Nie et al. | Pulse coupled neural network based MRI image enhancement using classical visual receptive field for smarter mobile healthcare | |
| CN112381846A (en) | Ultrasonic thyroid nodule segmentation method based on asymmetric network | |
| CN113421240A (en) | Mammary gland classification method and device based on ultrasonic automatic mammary gland full-volume imaging | |
| CN113907710A (en) | Skin lesion classification system based on model-independent image enhancement meta-learning | |
| CN109215035B (en) | Brain MRI hippocampus three-dimensional segmentation method based on deep learning | |
| CN116664590B (en) | Automatic segmentation method and device based on dynamic contrast enhancement magnetic resonance image | |
| CN116228753B (en) | Tumor prognosis evaluation method, device, computer equipment and storage medium | |
| CN116486156A (en) | Full-view digital slice image classification method integrating multi-scale feature context | |
| CN116228731A (en) | Multi-contrast learning coronary artery high-risk plaque detection method, system and terminal | |
| CN114999661A (en) | Construction method of skin cancer identification model, skin cancer identification device and electronic equipment | |
| CN115239655A (en) | Thyroid ultrasonic image tumor segmentation and classification method and device | |
| Khani | Medical image segmentation using machine learning | |
| CN111612762B (en) | MRI brain tumor image generation method and system | |
| Rao et al. | Weight pruning-UNet: Weight pruning UNet with depth-wise separable convolutions for semantic segmentation of kidney tumors | |
| CN113012129A (en) | System and device for counting area positioning and marked nerve cells of brain slice image | |
| Wei et al. | Application of U-net with variable fractional order gradient descent method in rectal tumor segmentation |
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 |