CN112017184B - New coronary pneumonia CT image processing method based on lung non-uniform pooling - Google Patents
New coronary pneumonia CT image processing method based on lung non-uniform pooling Download PDFInfo
- Publication number
- CN112017184B CN112017184B CN202011184333.6A CN202011184333A CN112017184B CN 112017184 B CN112017184 B CN 112017184B CN 202011184333 A CN202011184333 A CN 202011184333A CN 112017184 B CN112017184 B CN 112017184B
- Authority
- CN
- China
- Prior art keywords
- pooling
- lung
- image
- windows
- pooling windows
- 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
-
- 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
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/187—Segmentation; Edge detection involving region growing; involving region merging; involving connected component labelling
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (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)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Biophysics (AREA)
- Biomedical Technology (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 Processing (AREA)
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The invention discloses a new coronary pneumonia CT image processing method based on lung non-uniform pooling, which comprises the following steps: s1, performing a full-automatic lung segmentation algorithm based on the FPN; s2, identifying the center lines of the two lungs; s3, lung pooling operation; s4, convolution neural network based on lung pooling. Aiming at the problems that a uniform pooling layer is used in a traditional convolutional neural network, important regions in the lung cannot be focused when CT images are classified, and features in the lung and focus features are not excavated, the method provides the lung pooling layer, amplifies the features of the lung regions when the convolutional neural network is pooled, compresses the regions outside the lung, eliminates redundant features, strengthens image information in the lung, improves the precision of the CT image processing method of the new coronary pneumonia, does not depend on any manual labeled image, and improves the practicability of an algorithm.
Description
Technical Field
The invention relates to a medical technology, in particular to a new coronary pneumonia CT image processing method based on lung non-uniform pooling.
Background
The nucleic acid detection is used as a gold standard for diagnosing the new coronary pneumonia, and has the defects of high false negative, high requirement on the stability of a kit, longer test time and the like. The sensitivity of diagnosis can be improved by carrying out new coronary pneumonia diagnosis through the CT image, the rapid new coronary pneumonia diagnosis is realized, and the CT image processing method has important significance on the precision of the CT image processing method. The existing CT image classification algorithm based on the convolutional neural network has the following two defects: 1) depending on manual or semi-automatic lesion segmentation, lesion tissues need to be segmented from the CT images and then classified for diagnosis. Due to wide distribution of inflammation areas in the lung and large morphological change, manual segmentation of lesion areas is time-consuming and has subjective difference, and a semi-automatic segmentation algorithm is difficult to ensure that all inflammation areas are correctly segmented. 2) The traditional convolutional neural network uses the uniform pooling operation of maximum value pooling or mean value pooling, the spatial position information of the image is ignored in the mode, the intra-lung region and the extra-lung irrelevant region are treated equally, and the loss of the intra-lung information and the interference of the extra-lung irrelevant information are caused in the pooling process.
Therefore, there is a need for a fully automatic image analysis method that does not rely on manual annotation data, can analyze all verification regions in the lung, and further, a pooling method that can take into account spatial location information in the image, and that retains more features for regions in the lung and less features for irrelevant regions outside the lung to reduce noise interference.
For the problems in the prior art, the inventor finds that when the feature pyramid full convolution network FPN constructed based on the DenseNet added to the lung non-pooling processing is applied to the new coronary pneumonia CT image processing, the defects of the existing CT image classification algorithm based on other convolution neural networks can be overcome well, and an unexpected effect is achieved. DenseNet is a convolutional neural network with dense connections, as compared to other convolutional neural networks. In the network, any two layers are directly connected, namely, the input of each layer of the network is the union of the outputs of all the previous layers, and the feature map learned by the layer is directly transmitted to all the subsequent layers as input, so that the parameter quantity of the Densenet is greatly reduced compared with other models, and the model precision can be further improved by adding the lung heterogeneous pooling treatment on the basis.
The application of DenseNet added with lung non-pooling treatment to the new coronary pneumonia CT image processing belongs to the initiative in the field of processing CT images based on a deep learning neural network feature extraction model.
Disclosure of Invention
The invention aims to provide a new coronary pneumonia CT image analysis method based on lung non-uniform pooling, which can be used for fully automatically segmenting lung regions from a CT image without depending on an artificial labeling image, automatically identifying the central lines of the left lung and the right lung, constructing a convolutional neural network, gradually amplifying information in the lungs through lung pooling operation, and removing irrelevant information outside the lungs, so that the full-automatic image analysis is carried out on the full-lung CT image, and the high-precision diagnosis of the new coronary pneumonia CT image is realized. By adopting the CT image analysis method, full-automatic auxiliary diagnosis can be realized, the popularization and the application are convenient, and subjective difference and labor consumption caused by manually delineating the region of interest are avoided.
In order to achieve the aim, the invention provides a new coronary pneumonia CT image analysis method based on lung non-uniform pooling, which comprises the following steps:
s1, a full-automatic lung segmentation algorithm based on a feature pyramid full convolution network FPN:
constructing a characteristic pyramid full convolution network FPN based on DenseNet, and fully automatically segmenting lung regions from the new coronary pneumonia CT image;
s2, identifying the center lines of the two lungs:
after segmenting the lung region from the new coronary pneumonia CT image fully automatically by step S1, first, performing connected component screening on the segmented lung region image to obtain connected components 1 and 2 corresponding to the left and right lungs, respectively, and then detecting minimum abscissas a1 and b1 and maximum abscissas a2 and b2 of the connected components 1 and 2, respectively, to obtain a center line ac = a1+ (a2-a1)/2 of the left lung and a center line bc = b1+ (b2-b1)/2 of the right lung;
s3, lung non-uniform pooling operation:
firstly, pooling the lungs according to the center lines of the two lungs obtained in step S2, sequentially using pooling windows with step sizes of 1, 2 and 3 from the center line to both sides in the left and right lung areas, respectively, to compress the image areas with sizes of 1 x 1, 2 x 2 and 3 x 3 into a value during the pooling operation, i.e. taking the maximum value of the coverage area of the pooling windows as the output of the pooling windows, and defining for further determining the number of each pooling window: a) the image size after lung non-uniform pooling is half of the input image, b) of the three pooling windows, the pooling window with step size of 2 accounts for half of the number of all pooling windows, the number n1 of pooling windows with step size of 1, the number n2 of pooling windows with step size of 2 and the number n3 of pooling windows with step size of 3 are calculated by the following formula:
where [ ] denotes a rounding-down operation and I denotes the size of the input image, which will result in a remainder since I cannot necessarily be divided by 8 or 4, resulting in the three pooling windows not completely covering the input image, the remainders r1 and r2 are introduced and are calculated as follows:
then, the following lookup table L is constructed:
finally, the final number of pooling windows N1 of step size 1, the final number of pooling windows N2 of step size 2 and the final number of pooling windows N3 of step size 3 are calculated by the following equations, respectively:
after the final number of the pooling windows is obtained, arranging the three pooling windows on the input image, and when the pooling windows are arranged, preferentially arranging the pooling windows with the step length of 1 near the center line of the lung, then arranging the pooling windows with the step length of 2, finally arranging the pooling windows with the step length of 3 at the position farthest from the center line of the lung, and after the lung in the column direction is subjected to non-uniform pooling, uniformly pooling the maximum value in the row direction of the image to obtain a final pooling result image;
s4, convolutional neural network based on lung non-uniform pooling:
and constructing a convolutional neural network based on the lung pooling operation proposed in the step S3, and realizing the processing of the new coronary pneumonia CT image.
The technical solution of the present invention is further described in detail by the accompanying drawings and embodiments.
Drawings
FIG. 1 is a general flowchart of a CT image processing method of new coronary pneumonia based on non-uniform pooling of lungs according to an embodiment of the present invention;
FIG. 2 is a flowchart of a fully automatic lung segmentation algorithm of a CT image processing method for new coronary pneumonia based on non-uniform pooling of lungs in an embodiment of the present invention;
FIG. 3 is a dual lung centerline identification process of a new coronary pneumonia CT image processing method based on non-uniform pooling of lungs in an embodiment of the present invention;
FIG. 4 is a flowchart of a lung non-uniform pooling algorithm of a new coronary pneumonia CT image processing method based on lung non-uniform pooling according to an embodiment of the present invention;
FIG. 5 is a block diagram of a new coronary pneumonia CT image processing method based on non-uniform lung pooling according to an embodiment of the present invention.
Detailed Description
The present invention will be further described with reference to the accompanying drawings, and it should be noted that the present embodiment is based on the technical solution, and the detailed implementation and the specific operation process are provided, but the protection scope of the present invention is not limited to the present embodiment.
Fig. 1 is an overall flowchart of a CT image processing method for treating new coronary pneumonia based on non-uniform pooling of lungs in an embodiment of the present invention, which comprises the following steps:
s1, a full-automatic lung segmentation algorithm based on the feature pyramid full convolution neural network FPN:
as shown in fig. 2, a characteristic pyramid full convolution neural network based on DenseNet121 is constructed, and lung regions are segmented from CT images in a full-automatic manner. The characteristic pyramid full convolution neural network FPN uses a DenseNet network with weights pre-trained in ImageNet as a basic network, then extracts the output of the last layer of convolution layer from each Dense block in the DenseNet as multi-scale characteristics in a characteristic pyramid mode, then samples and splices the characteristics of different scales up step by step, and finally obtains a segmented lung region, namely a lung target region ROI, in a full convolution network mode.
S2, identifying the center lines of the two lungs:
after the lung region is fully automatically segmented from the CT image in step S1, as shown in fig. 3, first, connected component screening is performed on the segmented lung region image to obtain connected components 1 and 2 corresponding to the left lung and the right lung, respectively. Then, the minimum abscissas a1 and b1, and the maximum abscissas a2 and b2 of the connected domain 1 and the connected domain 2, respectively, are detected; the center line ac = a1+ (a2-a1)/2 of the left lung and the center line bc = b1+ (b2-b1)/2 of the right lung are obtained.
S3, lung pooling operation:
after obtaining the center lines of the left lung and the right lung, in order to focus more on image information in the lung and eliminate redundant irrelevant information outside the lung when images are analyzed and processed subsequently, a novel lung pooling algorithm is provided, the algorithm adopts a non-uniform pooling mode, a smaller pooling window is used near the center lines of the two lungs to retain more detailed information, and a larger pooling window is used far away from the center lines of the lungs to remove the redundant irrelevant information outside the lungs. Finally, the intra-pulmonary information is amplified and the extrapulmonary irrelevant information is compressed in the non-uniform pooling mode.
In specific implementation, firstly, the number of pooling windows with different sizes is determined, the invention uses three pooling windows with step lengths of 1, 2 and 3, which respectively represent that image areas with sizes of 1 x 1, 2 x 2 and 3 x 3 are compressed into a value during pooling operation, namely, the maximum value of the coverage area of the pooling windows is taken as the output of the pooling windows. To further determine the number of each pooling window, the present invention defines: a) the image size after the lung non-uniform pooling is half of the input image; b) of the three kinds of pooling windows, the pooling window having the step length of 2 occupies half of the number of all the pooling windows. Therefore, the number n1 of pooling windows with step size 1, the number n2 of pooling windows with step size 2, and the number n3 of pooling windows with step size 3 are respectively calculated by the following formula:
where [ ] denotes a rounding-down operation, and I denotes the size of the input image. Since I is not necessarily evenly divisible by 8 or 4, the above operation will yield a remainder, resulting in the three pooling windows not completely covering the input image. The invention introduces remainders r1 and r2, and the calculation mode is as follows:
then, the following lookup table L is constructed:
finally, the final number of pooling windows N1 of step size 1, the final number of pooling windows N2 of step size 2 and the final number of pooling windows N3 of step size 3 are calculated by the following equations, respectively:
after the final number of pooling windows is obtained, the three pooling windows are arranged on the input image as shown in FIG. 4. When arranging the pooling windows, the pooling windows with step size 1 are preferentially arranged near the center line of the lung, then the pooling windows with step size 2 are arranged, and finally the pooling windows with step size 3 are arranged at the position farthest from the center line of the lung. After the lung is non-uniformly pooled in the column direction, the maximum value in the row direction is uniformly pooled to obtain a final pooled result image.
S4, convolutional neural network based on lung non-uniform pooling:
in order to realize the new coronary pneumonia CT image processing, the invention constructs an image classification network shown in figure 5, which uses the lung target region ROI obtained in the step 1 as input and uses a convolutional neural network of a DenseNet structure for classification. Different from the traditional DenseNet network, the classification network shown in FIG. 5 inserts the lung non-uniform pooling layer proposed in step 3 between every two Dense blocks for lung feature amplification and extrapulmonary feature compression, so that image information in the lung can be gradually amplified in the step-by-step non-uniform pooling operation process, and simultaneously redundant irrelevant information outside the lung is compressed, the classification network focuses on CT images in the lung, so that the classification accuracy is improved, and a full-automatic and high-accuracy new coronary pneumonia CT image processing method is realized.
Finally, it should be noted that: the above embodiments are only for illustrating the technical solutions of the present invention and not for limiting the same, and although the present invention is described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that: modifications and equivalents may be made to the invention without departing from the spirit and scope of the invention.
Claims (1)
1. A new coronary pneumonia CT image processing method is characterized in that: the method comprises the following steps:
s1, a full-automatic lung segmentation algorithm based on a feature pyramid full convolution network FPN:
constructing a characteristic pyramid full convolution network FPN based on DenseNet, and fully automatically segmenting lung regions from the new coronary pneumonia CT image;
s2, identifying the center lines of the two lungs:
after segmenting the lung region from the new coronary pneumonia CT image fully automatically by step S1, first, performing connected component screening on the segmented lung region image to obtain connected components 1 and 2 corresponding to the left and right lungs, respectively, and then detecting minimum abscissas a1 and b1 and maximum abscissas a2 and b2 of the connected components 1 and 2, respectively, to obtain a center line ac = a1+ (a2-a1)/2 of the left lung and a center line bc = b1+ (b2-b1)/2 of the right lung;
s3, lung non-uniform pooling operation:
firstly, pooling the lungs according to the center lines of the two lungs obtained in step S2, sequentially using pooling windows with step sizes of 1, 2 and 3 from the center line to both sides in the left and right lung areas, respectively, to compress the image areas with sizes of 1 x 1, 2 x 2 and 3 x 3 into a value during the pooling operation, i.e. taking the maximum value of the coverage area of the pooling windows as the output of the pooling windows, and defining for further determining the number of each pooling window: a) the image size after lung non-uniform pooling is half of the input image, b) of the three pooling windows, the pooling window with step size of 2 accounts for half of the number of all pooling windows, the number n1 of pooling windows with step size of 1, the number n2 of pooling windows with step size of 2 and the number n3 of pooling windows with step size of 3 are calculated by the following formula:
where [ ] denotes a rounding-down operation and I denotes the size of the input image, which will result in a remainder since I cannot necessarily be divided by 8 or 4, resulting in the three pooling windows not completely covering the input image, the remainders r1 and r2 are introduced and are calculated as follows:
then, the following lookup table L is constructed:
finally, the final number of pooling windows N1 of step size 1, the final number of pooling windows N2 of step size 2 and the final number of pooling windows N3 of step size 3 are calculated by the following equations, respectively:
after the final number of the pooling windows is obtained, arranging the three pooling windows on the input image, and when the pooling windows are arranged, preferentially arranging the pooling windows with the step length of 1 near the center line of the lung, then arranging the pooling windows with the step length of 2, finally arranging the pooling windows with the step length of 3 at the position farthest from the center line of the lung, and after the lung in the column direction is subjected to non-uniform pooling, uniformly pooling the maximum value in the row direction of the image to obtain a final pooling result image;
s4, convolutional neural network based on lung non-uniform pooling:
and constructing a convolutional neural network based on the lung pooling operation proposed in the step S3, and realizing the processing of the new coronary pneumonia CT image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011184333.6A CN112017184B (en) | 2020-10-30 | 2020-10-30 | New coronary pneumonia CT image processing method based on lung non-uniform pooling |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011184333.6A CN112017184B (en) | 2020-10-30 | 2020-10-30 | New coronary pneumonia CT image processing method based on lung non-uniform pooling |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112017184A CN112017184A (en) | 2020-12-01 |
| CN112017184B true CN112017184B (en) | 2021-01-26 |
Family
ID=73527728
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011184333.6A Active CN112017184B (en) | 2020-10-30 | 2020-10-30 | New coronary pneumonia CT image processing method based on lung non-uniform pooling |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112017184B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111209917A (en) * | 2020-01-03 | 2020-05-29 | 天津大学 | Pneumonia detection device |
| CN111612764A (en) * | 2020-05-21 | 2020-09-01 | 佛山市普世医学科技有限责任公司 | New crown pneumonochemical contrast analysis method and system for glass lesions and storage medium |
-
2020
- 2020-10-30 CN CN202011184333.6A patent/CN112017184B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111209917A (en) * | 2020-01-03 | 2020-05-29 | 天津大学 | Pneumonia detection device |
| CN111612764A (en) * | 2020-05-21 | 2020-09-01 | 佛山市普世医学科技有限责任公司 | New crown pneumonochemical contrast analysis method and system for glass lesions and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112017184A (en) | 2020-12-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111507990B (en) | Tunnel surface defect segmentation method based on deep learning | |
| CN106950276B (en) | Pipeline defect depth inversion method based on convolutional neural network | |
| CN113838054B (en) | Mechanical part surface damage detection method based on artificial intelligence | |
| CN111815564B (en) | Method and device for detecting silk ingots and silk ingot sorting system | |
| CN107122735B (en) | Multi-target tracking method based on deep learning and conditional random field | |
| CN109903282B (en) | Cell counting method, system, device and storage medium | |
| CN115994907B (en) | Intelligent processing system and method for comprehensive information of food detection mechanism | |
| CN114487129B (en) | Flexible material damage identification method based on acoustic emission technology | |
| CN113139928B (en) | Training method of lung nodule detection model and lung nodule detection method | |
| CN115908142A (en) | Contact net tiny part damage testing method based on visual recognition | |
| CN108764343B (en) | Method for positioning tracking target frame in tracking algorithm | |
| CN112017184B (en) | New coronary pneumonia CT image processing method based on lung non-uniform pooling | |
| CN110349119B (en) | Pavement disease detection method and device based on edge detection neural network | |
| CN116109621B (en) | Defect detection method and system based on depth template | |
| JP6246978B2 (en) | Method for detecting and quantifying fibrosis | |
| CN115406359A (en) | Image-based concrete crack measuring method | |
| CN107123105A (en) | Images match defect inspection method based on FAST algorithms | |
| CN113780488A (en) | Scratching and scratching defect detection method based on position polymerization degree | |
| CN110376436B (en) | Multi-scale noise power spectral line spectrum detection method | |
| CN111967424A (en) | Buckwheat disease identification method based on convolutional neural network | |
| CN111881922A (en) | Insulator image identification method and system based on significance characteristics | |
| CN116958128B (en) | Medical image automatic positioning method based on deep learning | |
| CN113870265B (en) | Industrial part surface defect detection method | |
| CN118411368B (en) | Intelligent detection method, medium and system for baked tobacco leaves | |
| CN115953384B (en) | Online detection and prediction method for morphological parameters of tobacco leaves |
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 | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20211118 Address after: 300050 No. 1, Dali Road, Heping District, Tianjin Patentee after: ENVIRONMENTAL MEDICINE AND OPERATIONAL MEDICINE Research Institute ACADEMY OF MILITARY MEDICAL SCIENCES Address before: 1502, 12 / F, building 1, yard 1, Jiuqiao Road, Daxing District, Beijing 100163 Patentee before: Beijing Xinnuo Weikang Technology Co.,Ltd. |