CN111507964B - Integrated learning grading method for new coronary pneumonia, electronic equipment and storage medium - Google Patents
Integrated learning grading method for new coronary pneumonia, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN111507964B CN111507964B CN202010304517.5A CN202010304517A CN111507964B CN 111507964 B CN111507964 B CN 111507964B CN 202010304517 A CN202010304517 A CN 202010304517A CN 111507964 B CN111507964 B CN 111507964B
- Authority
- CN
- China
- Prior art keywords
- data
- loss
- ensemble learning
- grading method
- ordered
- 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.)
- Active
Links
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
- G06N3/08—Learning methods
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H30/00—ICT specially adapted for the handling or processing of medical images
- G16H30/20—ICT specially adapted for the handling or processing of medical images for handling medical images, e.g. DICOM, HL7 or PACS
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
-
- 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
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
- G06T2207/10081—Computed x-ray tomography [CT]
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- 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
- G06T2207/30061—Lung
-
- 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
- G06T2207/30061—Lung
- G06T2207/30064—Lung nodule
-
- 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
- G06T2207/30096—Tumor; Lesion
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Medical Informatics (AREA)
- General Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- Biomedical Technology (AREA)
- Public Health (AREA)
- General Physics & Mathematics (AREA)
- Primary Health Care (AREA)
- Epidemiology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Software Systems (AREA)
- Computing Systems (AREA)
- Radiology & Medical Imaging (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Biophysics (AREA)
- Databases & Information Systems (AREA)
- Mathematical Physics (AREA)
- Pathology (AREA)
- Molecular Biology (AREA)
- Computational Linguistics (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Quality & Reliability (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses an integrated learning grading method for new coronary pneumonia, which comprises the following steps: acquiring CT data of a plurality of patients and preprocessing the CT data to generate lung window data with a set format; constructing an ensemble learning framework based on an ordered regression residual error network; inputting the lung window data into an ensemble learning frame for training so as to output a training model; inputting new CT data into the training model to output the grading result of the pneumonia symptoms corresponding to the CT data. The integrated learning framework built by the invention adopts the ordered regression regularization and combines the ordered structure among the new coronary pneumonia grading data, so that the ordered structure among the data can be learned, the model performance is improved, the large deviation of model prediction is prevented, and the data result output is more accurate.
Description
Technical Field
The invention relates to a data processing technology, in particular to an integrated learning grading method for new coronary pneumonia, electronic equipment and a storage medium.
Background
The main difficulty of the new coronary pneumonia lies in the difficulty of detection, nucleic acid detection is the most important basis for the accurate diagnosis of patients at present, and effective clinical pointers and experiences become important reference indexes. The chest CT is a conventional tool for diagnosing pneumonia, the examination speed is high, and the image can be used for detecting pneumonia lesions, judging properties, integrating ranges, diagnosing and evaluating and the like. Meanwhile, the new coronary pneumonia has the rapid change of lung infection, and is mainly characterized by the change of the distribution of the outer zone, multiple segments and the interstitiality of ground glass on the image, and the change appears in two or three days. Studies have shown that the sensitivity of thoracic CT to identification of new coronary pneumonia infection is 98%. The problem that exists at present is that to chest CT all need artifical discernment, and medical staff's amount of labour is huge, influences detection efficiency.
Disclosure of Invention
In order to overcome the defects of the prior art, the invention aims to provide an integrated learning grading method for new coronary pneumonia, which can solve the problem that the prior art depends on manual work to check CT data.
Another object of the present invention is to provide an electronic device, which can solve the problem of the prior art that the CT data is checked manually.
It is a further object of the present invention to provide a storage medium that can solve the problem of the prior art that relies on manual review of CT data.
One of the purposes of the invention is realized by adopting the following technical scheme:
the integrated learning grading method of the new coronary pneumonia comprises the following steps:
acquiring CT data of a plurality of patients and preprocessing the CT data to generate lung window data with a set format;
inputting pneumonia symptom grading parameters, and constructing an integrated learning frame based on an ordered regression residual error network;
inputting the lung window data into an ensemble learning frame for training so as to output a training model;
inputting new CT data into the training model to output the grading result of the pneumonia symptoms corresponding to the CT data.
Preferably, the CT data includes an image format and pixel values.
Preferably, the image format is a DCM format.
Preferably, the preprocessing is to convert the PIXEL value of the CT data into the HU value of the CT by the formula HU ═ SLOPE × PIXEL + interval, where SLOPE is the scaling SLOPE of the CT data, PIXEL value is PIXEL value, and interval is the scaling INTERCEPT of the CT data.
Preferably, the range of CT data is adjusted to lung window data with a window level of-300 HU and a window width of 1400 HU.
Preferably, the lung window data is normalized.
Preferably, the "building an ensemble learning framework based on the ordered regression residual error network" specifically includes:
constructing a 3DRESNet34 network, wherein the 3DRESNet34 network has a tail feature layer, and the loss function of the 3DRESNet34 network is cross entropy loss;
adding an ordered regression canonical to an ending feature layer of the 3 dressnet 34 network to form an ordered regression learning framework; the loss function of the order regression regularization is an absolute value loss.
Integrating the model on the basis of the ordered regression learning framework, extracting macroscopic features of data, performing feature fusion on the features and the features extracted by the ordered regression learning framework, and performing XG boosting on the fused features;
preferably, the loss of the ensemble learning framework is defined as: loss ═ (1- λ) cross import Loss + λ L1Loss, where Loss is the Loss of the ensemble learning framework; λ is a constant parameter, and 0< λ < 1; cross EntropyLoss is the cross entropy Loss and L1Loss is the absolute value Loss of the order regression regularization.
The second purpose of the invention is realized by adopting the following technical scheme:
an electronic device having a memory, a processor, and a computer readable program stored in the memory and executable by the processor, wherein the computer readable program, when processed by the processor, implements a ensemble learning grading method as in any one of the objects of the invention.
The third purpose of the invention is realized by adopting the following technical scheme:
a storage medium having a computer readable program stored thereon for execution by a processor, wherein the computer readable program, when processed by the processor, implements a ensemble learning grading method according to any one of the objects of the present invention.
Compared with the prior art, the invention has the beneficial effects that:
according to the invention, the model is constructed to automatically detect lung CT, judge whether the new coronary pneumonia is infected, and output the grading result in combination with the infection degree, so that the workload of medical workers is reduced, and the medical workers can conveniently judge the detection and treatment needed in the next step according to the grading result, thereby more fully utilizing medical resources and avoiding unnecessary waste; the built integrated learning framework adopts the ordered regression regularization and combines the ordered structure among the new coronary pneumonia grading data, so that the ordered structure among the data can be learned, the performance of the model is improved, the large deviation of model prediction is prevented, and the data result output is more accurate.
Drawings
FIG. 1 is a flow chart of the integrated learning grading method for new coronary pneumonia according to the present invention;
FIG. 2 is a CT data of a first embodiment of the present invention;
FIG. 3 is a structural diagram of an ensemble learning framework according to the present invention.
Detailed Description
The invention will be further described with reference to the accompanying drawings and the detailed description below:
example one
The present embodiment provides a new coronary pneumonia integrated learning and rating method, as shown in fig. 1, the present embodiment includes the following steps:
CT data for a number of patients is acquired and pre-processed to generate lung window data in a set format S1.
The embodiment of the invention is used for detecting lung CT of suspected patients. Perhaps enough training data before the model is built. The training data may be obtained by collecting CT of the lungs of a patient who has been diagnosed and processing it as CT data. Through the lung CT of the suspected patient, whether the suspected patient is infected with the new coronary pneumonia is judged, the grade of the infection degree is given, and according to different grades, a doctor can judge whether the patient needs further detection and treatment according to the grading result, so that medical resources can be fully utilized, more resources can be used for the patient who needs more, and unnecessary waste is avoided.
As shown in fig. 2, the CT data, i.e., the CT image, can be used in the present embodiment. In this step, the CT data includes an image format and corresponding pixel values.
Furthermore, the image format of the conventional CT image is the DCM format, and the DCM format adopted in the embodiment is more suitable for the conventional application, and is convenient to implement.
This step treats the suspected lesion area, i.e., pulmonary CT. In this embodiment, the preprocessing is to convert the PIXEL value of the CT data into the HU value of the CT by the formula HU ═ SLOPE × PIXEL + interrupt, where SLOPE is the scaling SLOPE of the CT data, PIXEL is the PIXEL value, and interrupt is the scaling INTERCEPT of the CT data.
In a more preferred embodiment, the present embodiment adjusts the range of the CT data to be lung window data having a window level of-1000 to 400 HU. The lung window data relates to window width and window level. On the basis, the lung window data are normalized, and the precision of a subsequent model is improved.
And S2, inputting pneumonia symptom grading parameters, and constructing an integrated learning framework based on the ordered regression residual error network.
The ensemble learning framework of this step is shown in fig. 3, input is an input of data, and ResNet is a network to be constructed subsequently. In this step, "building an ensemble learning framework based on an ordered regression residual error network" specifically includes:
constructing a 3DRESNet34 network, wherein the 3DRESNet34 network has a tail feature layer, and the loss function of the 3DRESNet34 network is cross entropy loss;
adding an ordered regression canonical to an ending feature layer of the 3 dressnet 34 network to form an ordered regression learning framework; the loss function of the order regression regularization is an absolute value loss.
Integrating the model on the basis of the ordered regression learning framework, extracting macroscopic features of data, namely crop in the graph 3, performing feature fusion on the features and the features extracted by the ordered regression learning framework, and performing XGBOSTING on the fused features;
further, the loss of the ensemble learning framework is defined as: loss ═ (1- λ) cross import Loss + λ L1Loss, where Loss is the Loss of the ensemble learning framework; λ is a constant parameter, and 0< λ < 1; cross EntropyLoss is the cross entropy Loss and L1Loss is the absolute value Loss of the order regression regularization.
At present, the grade of the new coronary pneumonia is divided into 6 grades, and the grades are divided from 0 to 5, and the grades are specifically as follows:
and 0, no abnormal sign of lung CT.
1 point, slight calcification, lung bullae, nodules, etc. The clinical significance at this level suggests no association with infection by the novel coronavirus (2019-nCoV).
2, extensive pulmonary emphysema, pulmonary fibrosis, lung malignant tumor, etc. The clinical significance at this level is that patients may be more susceptible to new coronaviruses or to develop severe illness after infection.
And 3, limited leaf segment lesions, leaf segment small plaque shadows or multiple nodules, which indicate lung infectious lesions (infected by bacteria or tuberculosis or other viruses which can be difficult to identify) and have smaller possibility of infection by the novel coronavirus. The clinical significance of the grade is household isolation and community isolation, and the follow-up examination after treatment such as targeted antibacterial or anti-tuberculosis treatment and the like; if the treatment is not effective, the nucleic acid detection judges whether the novel coronavirus (2019-nCoV) is infected.
And 4, grinding glass density shadows around the central actual change of a single leaf segment, or grinding glass density shadows or grinding small spot patches frequently generated by two lungs, wherein the shapes are different, and the lung is prompted to have infectious lesion, and the probability of infection of the novel coronavirus is higher. The clinical significance is hospital isolation treatment, nucleic acid detection is applied, and 2019-nCoV infection can be diagnosed with certainty when the nucleic acid is positive.
And 5, the smear-shaped frosted glass density shadow with multiple speckles under the pleura of the double lungs has clear boundary, can be changed into large solid or gradually absorbed and fibrillated in a short period (within a few days), and strongly suggests the novel coronavirus infectious pneumonia.
Since the ranked data of the new coronary pneumonia has an ordered structure, and a general classification model is difficult to learn the ordered relationship among the data, in the embodiment, an ordered regression rule is creatively added to the network feature layer to learn the ordered relationship among the data.
The 3 dressnet 34 network is a residual network, the 3 dressnet 34 network is used as a base network in the present embodiment, on this basis, the cross entropy loss is used as a loss function, and an ordered regression regularization part is added to the ending feature layer, the ordered regression regularization term adopts L1loss, the whole model loss function is a convex combination of the two, that is, the loss function of the whole model is the above formula: loss ═ (1- λ) crossEntropyLoss + λ L1 Loss.
In this step, the optimal parameter can be determined through a five-fold experiment or a ten-fold experiment by adjusting the value of λ. In an actual experiment, the application of the embodiment has a better effect than a classification model without using the ordered regular term, the model prediction result does not deviate too much, and the accuracy is higher.
An integrated learning mode is added on the basis of the ordered regression, learning characteristics are data macro characteristics and network micro detail characteristics, data are cut through the macro characteristics, the data macro characteristics and the network micro detail characteristics are combined for learning, the data characteristics are fully utilized, and the model performance and the robustness are further improved through an integrated method. The integrated learning method considers the ordered relation among the data, and can fully learn and utilize the ordered structure information among the data, thereby improving the performance of the model.
And S3, inputting the lung window data into the ensemble learning framework for training so as to output a training model.
Through the extraction and processing of the data and the establishment of the model, in the step, the training model is obtained by inputting the data into the frame for learning and training on the basis of the two steps. The model is trained through sufficient data, so that the model has more perfect processing capacity, and the model can have better accuracy in the actual processing of the subsequent input data. The trained model has the capability of identifying the grade division of the new coronary pneumonia and has the capability of preprocessing CT data.
And S4, inputting new CT data into the training model to output the grading result of the pneumonia symptoms corresponding to the CT data.
Based on the previous steps, new CT data are input, the CT data are preprocessed, and a model is trained to identify a new coronary pneumonia grade corresponding to the CT data.
Example two
The embodiment provides an electronic device, on which a memory, a processor and a computer readable program stored in the memory and executable by the processor are stored, wherein the computer readable program, when processed by the processor, implements the following steps:
acquiring CT data of a plurality of patients and preprocessing the CT data to generate lung window data with a set format;
inputting pneumonia symptom grading parameters, and constructing an integrated learning frame based on an ordered regression residual error network;
inputting the lung window data into an ensemble learning frame for training so as to output a training model;
inputting new CT data into the training model to output the grading result of the pneumonia symptoms corresponding to the CT data.
EXAMPLE III
The present embodiment provides a storage medium having a computer readable program stored thereon, wherein the computer readable program when processed by a processor implements the following steps:
acquiring CT data of a plurality of patients and preprocessing the CT data to generate lung window data with a set format;
inputting pneumonia symptom grading parameters, and constructing an integrated learning frame based on an ordered regression residual error network;
inputting the lung window data into an ensemble learning frame for training so as to output a training model;
inputting new CT data into the training model to output the grading result of the pneumonia symptoms corresponding to the CT data.
Various other modifications and changes may be made by those skilled in the art based on the above-described technical solutions and concepts, and all such modifications and changes should fall within the scope of the claims of the present invention.
Claims (9)
1. The integrated learning grading method for the new coronary pneumonia is characterized by comprising the following steps of:
acquiring CT data of a plurality of patients and preprocessing the CT data to generate lung window data with a set format;
inputting pneumonia symptom grading parameters, and constructing an integrated learning frame based on an ordered regression residual error network;
inputting the lung window data into an ensemble learning frame for training so as to output a training model;
inputting new CT data into a training model to output a pneumonia symptom grading result corresponding to the CT data;
the specific steps of building the integrated learning framework based on the ordered regression residual error network are as follows:
constructing a 3DRESNet34 network, wherein the 3DRESNet34 network has a tail feature layer, and the loss function of the 3DRESNet34 network is cross entropy loss;
adding an ordered regression canonical to an ending feature layer of the 3 dressnet 34 network to form an ordered regression learning framework; the loss function of the order regression is absolute value loss;
and integrating the model on the basis of the ordered regression learning framework, extracting the macroscopic features of the data, performing feature fusion on the features and the features extracted by the ordered regression learning framework, and performing XGBOSTING on the fused features.
2. The ensemble learning grading method according to claim 1, wherein said CT data comprises image format and pixel values.
3. The ensemble learning grading method according to claim 2, wherein said image format is DCM format.
4. The ensemble learning grading method according to claim 2, wherein said preprocessing is to convert the digital PIXEL values of the CT data into HU values of CT by the formula HU-SLOPE PIXEL + INTERCEPT, where SLOPE is the scaling SLOPE of CT data, PIXEL is the PIXEL value, and INTERCEPT is the scaling INTERCEPT of CT data.
5. The ensemble learning grading method according to claim 4, wherein the range of CT data is adjusted to lung window data with a window level of-300 HU and a window width of 1400 HU.
6. The ensemble learning grading method according to claim 5, characterized in that said lung window data is normalized.
7. The ensemble learning grading method according to claim 1, wherein the loss of the ensemble learning framework is defined as: loss ═ (1- λ) cross import Loss + λ L1Loss, where Loss is the Loss of the ensemble learning framework; λ is a constant parameter, and 0< λ < 1; cross EntropyLoss is the cross entropy Loss and L1Loss is the absolute value Loss of the order regression regularization.
8. An electronic device having stored thereon a memory, a processor, and a computer readable program stored in the memory and executable by the processor, wherein the computer readable program, when processed by the processor, implements the ensemble learning grading method of any of claims 1-7.
9. A storage medium having a computer readable program stored thereon for execution by a processor, the computer readable program, when processed by the processor, implementing the ensemble learning grading method of any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010304517.5A CN111507964B (en) | 2020-04-17 | 2020-04-17 | Integrated learning grading method for new coronary pneumonia, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010304517.5A CN111507964B (en) | 2020-04-17 | 2020-04-17 | Integrated learning grading method for new coronary pneumonia, electronic equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111507964A CN111507964A (en) | 2020-08-07 |
| CN111507964B true CN111507964B (en) | 2021-03-19 |
Family
ID=71871034
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010304517.5A Active CN111507964B (en) | 2020-04-17 | 2020-04-17 | Integrated learning grading method for new coronary pneumonia, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111507964B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112766345A (en) * | 2021-01-12 | 2021-05-07 | 合肥黎曼信息科技有限公司 | Classification method based on ordered regression regularization |
| CN112700867A (en) * | 2021-01-13 | 2021-04-23 | 浙江大学医学院附属第一医院 | Classification diagnosis system for COVID-19 patient based on transfer learning |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107280697A (en) * | 2017-05-15 | 2017-10-24 | 北京市计算中心 | Lung neoplasm grading determination method and system based on deep learning and data fusion |
| CN109544534A (en) * | 2018-11-26 | 2019-03-29 | 上海联影智能医疗科技有限公司 | A kind of lesion image detection device, method and computer readable storage medium |
| CN110309860A (en) * | 2019-06-06 | 2019-10-08 | 昆明理工大学 | The method classified based on grade malignancy of the convolutional neural networks to Lung neoplasm |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106096632A (en) * | 2016-06-02 | 2016-11-09 | 哈尔滨工业大学 | Based on degree of depth study and the ventricular function index prediction method of MRI image |
| CN107103187B (en) * | 2017-04-10 | 2020-12-29 | 四川省肿瘤医院 | Lung nodule detection grading and management method and system based on deep learning |
| CN109009222A (en) * | 2018-06-19 | 2018-12-18 | 杨成伟 | Intelligent evaluation diagnostic method and system towards heart disease type and severity |
-
2020
- 2020-04-17 CN CN202010304517.5A patent/CN111507964B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107280697A (en) * | 2017-05-15 | 2017-10-24 | 北京市计算中心 | Lung neoplasm grading determination method and system based on deep learning and data fusion |
| CN109544534A (en) * | 2018-11-26 | 2019-03-29 | 上海联影智能医疗科技有限公司 | A kind of lesion image detection device, method and computer readable storage medium |
| CN110309860A (en) * | 2019-06-06 | 2019-10-08 | 昆明理工大学 | The method classified based on grade malignancy of the convolutional neural networks to Lung neoplasm |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111507964A (en) | 2020-08-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Shan et al. | Lung infection quantification of COVID-19 in CT images with deep learning | |
| Shan et al. | Abnormal lung quantification in chest CT images of COVID‐19 patients with deep learning and its application to severity prediction | |
| Yun et al. | Improvement of fully automated airway segmentation on volumetric computed tomographic images using a 2.5 dimensional convolutional neural net | |
| CN111028224B (en) | Data labeling method, model training device, image processing method, image processing device and storage medium | |
| CN111127466A (en) | Medical image detection method, device, equipment and storage medium | |
| CN111507964B (en) | Integrated learning grading method for new coronary pneumonia, electronic equipment and storage medium | |
| CN105139390A (en) | Image processing method for detecting pulmonary tuberculosis focus in chest X-ray DR film | |
| Li et al. | COVID-GATNet: A deep learning framework for screening of COVID-19 from chest X-ray images | |
| EP3971830B1 (en) | Pneumonia sign segmentation method and apparatus, medium and electronic device | |
| WO2021209887A1 (en) | Rapid, accurate and machine-agnostic segmentation and quantification method and device for coronavirus ct-based diagnosis | |
| Raajan et al. | Non-invasive technique-based novel corona (COVID-19) virus detection using CNN | |
| Bhosale et al. | Deep convolutional neural network based Covid-19 classification from radiology X-Ray images for IoT enabled devices | |
| AU2020103779A4 (en) | Covid-19 classification recognition method based on ct images of lungs | |
| CN111584046A (en) | AI (Artificial intelligence) processing method for medical image data | |
| Xu et al. | CARes‐UNet: Content‐aware residual UNet for lesion segmentation of COVID‐19 from chest CT images | |
| CN116864119B (en) | Pneumonia risk prediction method, equipment, system and storable medium based on infection period | |
| CN111798437A (en) | Novel coronavirus pneumonia AI rapid diagnosis method based on CT image | |
| Nurmaini et al. | Automated Detection of COVID-19 Infected Lesion on Computed Tomography Images Using Faster-RCNNs. | |
| CN110911008A (en) | Method for establishing biliary tract occluded alignment chart prediction model and application thereof | |
| CN112489022A (en) | COVID-19 rapid feature extraction system based on radiologics, rapid diagnosis system and disease course prediction system | |
| Yumeng et al. | Pneumonia Detection in chest X-rays: A deep learning approach based on ensemble RetinaNet and Mask R-CNN | |
| Razi | An analysis of COVID-19 using X-ray image segmentation based graph cut and box counting fractal dimension | |
| Rajaraman et al. | Semantically redundant training data removal and deep model classification performance: A study with chest X-rays | |
| Mayya et al. | A novel medical support deep learning fusion model for the diagnosis of COVID-19 | |
| Nneji et al. | COVID-19 Identification Using Deep Capsule Network: A Perspective of Super-Resolution CNN on Low-Quality CXR Images |
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 |