CN109978852B - Deep learning-based radiotherapy image target region delineation method and system for micro tissue organ - Google Patents
Deep learning-based radiotherapy image target region delineation method and system for micro tissue organ Download PDFInfo
- Publication number
- CN109978852B CN109978852B CN201910221344.8A CN201910221344A CN109978852B CN 109978852 B CN109978852 B CN 109978852B CN 201910221344 A CN201910221344 A CN 201910221344A CN 109978852 B CN109978852 B CN 109978852B
- Authority
- CN
- China
- Prior art keywords
- module
- image
- training
- delineation
- defining
- 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
- 238000000034 method Methods 0.000 title claims abstract description 49
- 210000000056 organ Anatomy 0.000 title claims abstract description 43
- 238000001959 radiotherapy Methods 0.000 title claims abstract description 21
- 238000013135 deep learning Methods 0.000 title claims abstract description 18
- 230000006870 function Effects 0.000 claims abstract description 23
- 238000012549 training Methods 0.000 claims description 28
- 230000008569 process Effects 0.000 claims description 15
- 238000012360 testing method Methods 0.000 claims description 12
- 238000007781 pre-processing Methods 0.000 claims description 10
- 238000012545 processing Methods 0.000 claims description 8
- 238000005457 optimization Methods 0.000 claims description 7
- 230000008676 import Effects 0.000 claims description 4
- 238000000605 extraction Methods 0.000 claims description 3
- FFBHFFJDDLITSX-UHFFFAOYSA-N benzyl N-[2-hydroxy-4-(3-oxomorpholin-4-yl)phenyl]carbamate Chemical compound OC1=C(NC(=O)OCC2=CC=CC=C2)C=CC(=C1)N1CCOCC1=O FFBHFFJDDLITSX-UHFFFAOYSA-N 0.000 claims 1
- 238000013507 mapping Methods 0.000 claims 1
- 230000003439 radiotherapeutic effect Effects 0.000 claims 1
- 238000013461 design Methods 0.000 abstract description 5
- 238000013528 artificial neural network Methods 0.000 abstract description 3
- 238000003709 image segmentation Methods 0.000 abstract description 3
- 210000001519 tissue Anatomy 0.000 description 24
- 210000001508 eye Anatomy 0.000 description 7
- 210000001331 nose Anatomy 0.000 description 7
- 238000005516 engineering process Methods 0.000 description 4
- 210000001328 optic nerve Anatomy 0.000 description 4
- 150000001875 compounds Chemical class 0.000 description 2
- 201000010099 disease Diseases 0.000 description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 2
- 238000007499 fusion processing Methods 0.000 description 2
- 210000003128 head Anatomy 0.000 description 2
- 210000003739 neck Anatomy 0.000 description 2
- 206010025323 Lymphomas Diseases 0.000 description 1
- 208000002454 Nasopharyngeal Carcinoma Diseases 0.000 description 1
- 206010061306 Nasopharyngeal cancer Diseases 0.000 description 1
- 206010028980 Neoplasm Diseases 0.000 description 1
- 206010068771 Soft tissue neoplasm Diseases 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 201000011510 cancer Diseases 0.000 description 1
- 238000007408 cone-beam computed tomography Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 210000001165 lymph node Anatomy 0.000 description 1
- 210000005075 mammary gland Anatomy 0.000 description 1
- 206010027191 meningioma Diseases 0.000 description 1
- 201000011216 nasopharynx carcinoma Diseases 0.000 description 1
- 210000003977 optic chiasm Anatomy 0.000 description 1
- 210000002990 parathyroid gland Anatomy 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 210000003079 salivary gland Anatomy 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 210000001685 thyroid gland Anatomy 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30096—Tumor; Lesion
Landscapes
- Engineering & Computer Science (AREA)
- Quality & Reliability (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Radiology & Medical Imaging (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Radiation-Therapy Devices (AREA)
- Image Analysis (AREA)
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The invention provides a deep learning-based radiotherapy image target region delineation method for micro tissues and organs. A deep learning neural network adopted in the model design of image segmentation provides a new loss function design, and the target organs are sketched. The method has the advantages that the organ delineation accuracy of the target region is high, the delineation of large organs and small organs is considered, and the delineation accuracy of the small organs is also high.
Description
Technical Field
The invention belongs to the technical field of medical image processing, and particularly relates to a deep learning-based radiotherapy image target region delineation method and system for micro tissues and organs.
Background
The treatment of the disease is often run at follow-up time, and a long time may be required for delineating the target area for treatment. In the prior art, the adoption of automatic software is reported to greatly shorten the drawing time. For example, automated radiotherapy target delineation software developed by a university of Toronto researcher utilizes artificial intelligence to learn historical radiotherapy data, and can be used for developing treatment plans.
However, in the field of medical image processing technology, most of the related technologies for target region delineation are directed to delineation of clinical target regions of large organs (such as common malignant tumors), such as target regions of lymphoma, meningioma, nasopharyngeal carcinoma, etc., while there are few reports on target regions of small organs, and small tissues.
The process of treatment plan optimization is a process of continuous improvement of the treatment plan, including determination of the target volume throughout the entire treatment plan involvement and execution. With the development of stereotactic radiotherapy, it has been a goal to accurately delineate the boundaries of a target region in a treatment technology system. Developing a precise target region delineation technology for small organs, micro organs and micro tissues is a significant problem and challenge for researchers. For example, the optic nerve is a very tiny tissue organ, which occupies a very small proportion of the whole human body image, about 1/10000 degrees, and the prior art method can not accurately and precisely delineate such tiny tissue organ.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a method and a system for delineating a target region of a radiotherapy image aiming at a micro tissue organ based on deep learning. The images include, but are not limited to, medical images of tumors such as CT, CBCT, MRI, PET, etc.
In the present invention, the small tissue organ and the minute tissue organ refer to organs less than 10% of the total amount in one image, including optic nerves of head and neck, eyes, nose, salivary glands, maxillofacial region and neck, thyroid gland, parathyroid gland, mammary gland and superficial lymph node, body and soft tissue tumor of limbs, etc. Preferably, the optic nerve of the head and neck.
In the present invention, the target region refers to a region that is a target for disease treatment.
In order to realize the purpose, the invention adopts the following technical scheme:
the invention provides a deep learning-based radiotherapy image target region delineation method of a micro tissue organ, which comprises the following steps:
(I) image training procedure
1) Importing and preprocessing training image data;
the preprocessing comprises three-dimensional reconstruction, denoising, enhancement, registration and fusion processing of training image data.
2) Defining a network structure;
3) defining a loss function
Wherein the loss function
As follows:
in the formula (I), the compound is shown in the specification,
TP p (c) the method comprises the following steps Indicates that it is determined to be a positive sample and is also a positive sample in fact;
FN p (c) the method comprises the following steps Indicates that it is judged to be a negative sample, but is in fact a positive sample;
FP p (c) the method comprises the following steps Indicates a positive sample, but is in fact a negative sample;
p n (c) the method comprises the following steps Representing the probability that the nth voxel is predicted as class c;
c represents the number of the types to be predicted +1, and +1 represents that one type is background; if the number of types to be predicted in a medical image is 2, the number of types with prediction in the algorithm is 3.
For example, the following steps are carried out: the total number of 2500 pixels in a medical image is 50 x 50, eyes and noses are needed to be predicted, the eyes are 300 pixels, the noses are 200 pixels, and the remaining 2000 pixels are the background. In the algorithm, all pixels are classified into 3 types, eye (300 points), nose (200 points), and background (2000 points).
λ: represents L Dice And L Focal Lambda is in [0,1]]；
α, β: indicating for FN p (c) And FP p (c) α, β ∈ [0,1] of]；
N: representing the number of voxels in the medical image;
4) defining an optimization algorithm and adjusting a learning rate;
5) training a model;
the image training process of the invention is to obtain a training model. Wherein designing a suitable model is the key to the overall process. The invention belongs to one of the key steps of the training process through the proposed loss function.
(II) reasoning Process
6) Acquiring a sequence image (such as a CT sequence image) of a patient;
7) test data import and pretreatment;
8) importing a trained model;
9) processing the test data by using the trained model, and predicting the target area;
10) and performing edge extraction through the predicted target area to obtain a sketched target area.
The invention also provides a radiotherapy image target area delineation device of the micro tissue organ based on deep learning, which is suitable for the target area delineation method and comprises the following parts:
image training module
1) A training image data importing and preprocessing module;
2) defining a network structure module;
3) defining a loss function
(Loss function) module;
4) defining an optimization algorithm and adjusting a learning rate module;
5) a model training module;
(II) reasoning module
6) Acquiring a sequence image module (such as CT sequence image) of a user;
7) the test data importing and preprocessing module;
8) leading in a trained model module;
9) processing the test data by using the trained model, and performing a prediction module on the target area;
10) a target region delineation module.
The invention also provides the application of the method in the delineation of the target area of the radiotherapy image of large and small tissues and organs.
As shown in figure 1As shown, the background accounts for 98.18% of all voxels for the voxel frequency of each class on the MICCAI 2015challenge dataset. The optic chiasm accounts for 0.35% of the foreground, which means that it accounts for the entire CT only 365 Around 1/100,000 of the image. The large imbalance of the voxels of the micro-organ of tissue causes difficulties in delineation of the micro-organ of tissue.
Compared with the prior art, the method for delineating the target area of the radiotherapy image of the micro tissue organ based on the deep learning comprises a training process and an inference process of a model, wherein a deep learning neural network (adopting a deep learning segmentation algorithm) is adopted in the model design of image segmentation, and a new very key Loss function design-Loss function (Loss function) is provided at the same time, so that the delineation of the target area of the organ image is realized. The degree of disagreement of the predicted value f (x) of the model with the true value Y is evaluated by a loss function, which is a non-negative real-valued function, generally denoted L (Y, f (x)). The smaller the loss function, the better the robustness of the trained model. The loss function is the core of the empirical risk function and is also an important component of the structural risk function. How the loss function is designed also has a very important influence on the result of the algorithm. According to the radiotherapy image target region delineation method of the micro tissue organ based on the deep learning, the target region delineation problem of the small tissue organ and the micro tissue organ can be well solved by adopting the Loss function (Loss function); the method is used for sketching the target area of the medical image of the large tissue organ, and has remarkable beneficial effects of higher accuracy and higher precision. The method can simultaneously give consideration to the delineation of the target area of the radiotherapy image of a large tissue organ and a small tissue organ.
Drawings
Fig. 1 is a diagram illustrating a ratio of optic nerves in a human body image.
Fig. 2 is a flowchart of a method for delineating a target region of a radiotherapy image of a micro tissue organ based on deep learning according to the present invention.
Detailed Description
The invention is further described in detail with reference to the following specific examples and the accompanying drawings. The procedures, conditions, experimental methods and the like for carrying out the present invention are general knowledge and common general knowledge in the art except for the contents specifically mentioned below, and the present invention is not particularly limited.
As shown in fig. 2, the method for delineating a target region of a radiotherapy image of a micro tissue organ based on deep learning in this embodiment includes the following steps:
(I) image training procedure
1) Importing and preprocessing training image data;
the preprocessing comprises three-dimensional reconstruction, denoising, enhancement, registration and fusion processing of training image data.
2) Defining a network structure;
3) defining a loss function
Wherein the loss function
As follows:
in the formula (I), the compound is shown in the specification,
TP p (c) the method comprises the following steps Indicates that it is determined to be a positive sample and is also a positive sample in fact;
FN p (c) the method comprises the following steps Indicating a negative sample, but in fact a positive sample.
FP p (c) The method comprises the following steps Indicating a positive sample, but in fact a negative sample.
p n (c) The method comprises the following steps Representing the probability that the nth voxel is predicted as class c;
c represents the number of the types to be predicted +1, and +1 represents that one type is background; if the number of types to be predicted in a medical image is 2, the number of types with prediction in the algorithm is 3.
For example, the following steps are carried out: the total number of 2500 pixels in a medical image is 50 x 50, eyes and noses are needed to be predicted, the eyes are 300 pixels, the noses are 200 pixels, and the remaining 2000 pixels are the background. In the algorithm, all pixels are classified into 3 types, eye (300 points), nose (200 points), and background (2000 points).
λ: represents L Dice And L Focal λ ∈ [0,1] in the control parameter of]；
α, β: represents the regulatory parameters for FN and FP, α, β ∈ [0,1 ];
n: representing the number of voxels in the medical image;
4) defining an optimization algorithm and adjusting a learning rate;
5) training a model;
the image training process of the invention is to obtain a training model. Wherein designing a suitable model is the key to the overall process. The invention belongs to one of the key steps of the training process through the proposed loss function.
(II) reasoning Process
6) Acquiring a sequence image (such as a CT sequence image) of a patient;
7) test data import and pretreatment;
8) importing a trained model;
9) processing the test data by using the trained model, and predicting the target area;
10) and performing edge extraction through the predicted target area to obtain a sketched target area.
The radiotherapy image target area delineation device of the tiny tissue organ based on the deep learning in the embodiment is suitable for the target area delineation method, and comprises the following parts:
image training module
1) A training image data importing and preprocessing module;
2) defining a network structure module;
3) defining a loss function
(Loss function) module;
4) defining an optimization algorithm and adjusting a learning rate module;
5) a model training module;
(II) reasoning module
6) Acquiring a sequence image module (such as CT sequence image) of a user;
7) the test data importing and preprocessing module;
8) leading in a trained model module;
9) processing the test data by using the trained model, and performing a prediction module on the target area;
10) a target region delineation module.
The method comprises a training process and an inference process of the model, and a deep learning neural network adopted in the model design of image segmentation, so as to provide a new loss function design and realize the delineation of target organs. The method has the advantages that the organ delineation accuracy of the target region is high, the delineation of large organs and small organs is considered, and the delineation accuracy of the small organs is also high.
The protection of the present invention is not limited to the above embodiments. Variations and advantages that may occur to those skilled in the art may be incorporated into the invention without departing from the spirit and scope of the inventive concept, and the scope of the appended claims is intended to be protected.
Claims (4)
1. A radiotherapy image target region delineation method of micro tissue organs based on deep learning is characterized by comprising the following steps:
image training process
1) Training radiotherapy image data import and pretreatment;
2) defining a network structure;
3) defining a loss function
Wherein the loss function
As follows:
in the formula, TP p (c) Indicates that it is determined to be a positive sample and is also a positive sample in fact; FN (FN) p (c) Indicates that it is judged to be a negative sample, but is in fact a positive sample; FP p (c) Indicates a positive sample, but is in fact a negative sample; p is a radical of n (c) Representing the probability that the nth voxel is predicted as class c; the number of the types c is 1 more than the number of the types needing to be predicted, and the 1 more types are backgrounds; λ represents L Dice And L Focal Lambda is in [0,1]](ii) a Alpha, beta for FN p (c) And FP p (c) α, β ∈ [0,1] of](ii) a N represents the number of voxels in the radiotherapy image;
4) defining an optimization algorithm and adjusting a learning rate;
5) training a model;
second, reasoning process
6) Acquiring a sequence of images of a patient;
7) test data import and pretreatment;
8) importing a trained model;
9) processing the test data by using the trained model, and predicting the target area;
10) and performing edge extraction through the predicted target area to obtain a sketched target area.
2. The method of claim 1, wherein said small tissue organ is less than 10% of the total organ in an image.
3. A deep learning based radiotherapeutic image target volume delineation device of tiny tissue organs, the delineation device being adapted to the method of claim 1 or 2, comprising the following parts:
image training module
1) A training image data importing and preprocessing module;
2) defining a network structure module;
3) defining a loss function
A module;
4) defining an optimization algorithm and adjusting a learning rate module;
5) a model training module;
second, reasoning module
6) Acquiring a sequence image module of a user;
7) the test data importing and preprocessing module;
8) leading in a trained model module;
9) processing the test data by using the trained model, and performing a prediction module on the target area;
10) a target region delineation module.
4. Use of the method according to claim 1 or 2 for mapping a target region of a large or small tissue organ with radiotherapy images.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910221344.8A CN109978852B (en) | 2019-03-22 | 2019-03-22 | Deep learning-based radiotherapy image target region delineation method and system for micro tissue organ |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910221344.8A CN109978852B (en) | 2019-03-22 | 2019-03-22 | Deep learning-based radiotherapy image target region delineation method and system for micro tissue organ |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109978852A CN109978852A (en) | 2019-07-05 |
| CN109978852B true CN109978852B (en) | 2022-08-16 |
Family
ID=67080015
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910221344.8A Active CN109978852B (en) | 2019-03-22 | 2019-03-22 | Deep learning-based radiotherapy image target region delineation method and system for micro tissue organ |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109978852B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110570483B (en) * | 2019-08-08 | 2023-12-22 | 上海联影智能医疗科技有限公司 | Scanning method, scanning device, computer equipment and storage medium |
| CN111738989B (en) * | 2020-06-02 | 2023-10-24 | 北京全域医疗技术集团有限公司 | Organ sketching method and device |
| CN113421234B (en) * | 2021-06-17 | 2024-06-21 | 韩从辉 | Microscopic bladder endoscope imaging system of mathematical algorithm |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107403201A (en) * | 2017-08-11 | 2017-11-28 | 强深智能医疗科技（昆山）有限公司 | Tumour radiotherapy target area and jeopardize that organ is intelligent, automation delineation method |
| US20180061059A1 (en) * | 2016-08-26 | 2018-03-01 | Elekta, Inc. | System and methods for image segmentation using convolutional neural network |
| CN108257134A (en) * | 2017-12-21 | 2018-07-06 | 深圳大学 | Nasopharyngeal Carcinoma Lesions automatic division method and system based on deep learning |
-
2019
- 2019-03-22 CN CN201910221344.8A patent/CN109978852B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180061059A1 (en) * | 2016-08-26 | 2018-03-01 | Elekta, Inc. | System and methods for image segmentation using convolutional neural network |
| CN107403201A (en) * | 2017-08-11 | 2017-11-28 | 强深智能医疗科技（昆山）有限公司 | Tumour radiotherapy target area and jeopardize that organ is intelligent, automation delineation method |
| CN108257134A (en) * | 2017-12-21 | 2018-07-06 | 深圳大学 | Nasopharyngeal Carcinoma Lesions automatic division method and system based on deep learning |
Non-Patent Citations (1)
| Title |
|---|
| Brain tissue segmentation based on spatial information fusion by Dempster-Shafer theory;Jamal GHASEMI等;《Journal of Zhejiang University-SCIENCE C (Computers & Electronics)》;20120731;第13卷(第7期);第520-533页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109978852A (en) | 2019-07-05 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11386557B2 (en) | Systems and methods for segmentation of intra-patient medical images | |
| Fu et al. | A review of deep learning based methods for medical image multi-organ segmentation | |
| US11710241B2 (en) | Atlas-based segmentation using deep-learning | |
| CN109978852B (en) | Deep learning-based radiotherapy image target region delineation method and system for micro tissue organ | |
| Hou et al. | Unsupervised histopathology image synthesis | |
| Harlow et al. | The analysis of radiographic images | |
| US11315254B2 (en) | Method and device for stratified image segmentation | |
| CN113793304A (en) | Intelligent segmentation method for lung cancer target area and organs at risk | |
| Sahoo et al. | A comparative analysis of PGGAN with other data augmentation technique for brain tumor classification | |
| Vu et al. | A data-adaptive loss function for incomplete data and incremental learning in semantic image segmentation | |
| Mukherjee et al. | Domain adapted multitask learning for segmenting amoeboid cells in microscopy | |
| CN109829885A (en) | A kind of automatic identification nasopharyngeal carcinoma primary tumo(u)r method based on deep semantic segmentation network | |
| Fang et al. | Multi-organ segmentation network with adversarial performance validator | |
| CN116580814A (en) | Deep learning-based radiotherapy plan automatic generation system and method | |
| Uka et al. | FASTER R-CNN for cell counting in low contrast microscopic images | |
| Fallahdizcheh et al. | Sequential active contour based on morphological-driven thresholding for ultrasound image segmentation of ascites | |
| Ip et al. | Current landscape and potential future applications of artificial intelligence in medical physics and radiotherapy | |
| Lu et al. | Better Rough than Scarce: Proximal Femur Fracture Segmentation with Rough Annotations | |
| Yu et al. | Multiple organ segmentation framework for brain metastasis radiotherapy | |
| Biswas et al. | Adaptive Smooth Activation for Improved Disease Diagnosis and Organ Segmentation from Radiology Scans | |
| Chen et al. | Adaptive Region-Specific Loss for Improved Medical Image Segmentation | |
| Xu | Automated Segmentation of Organs at Risk for Nasopharyngeal Carcinoma by Dropout U-Net | |
| Dolz | Towards automatic segmentation of the organs at risk in brain cancer context via a deep learning classification scheme | |
| Yuan et al. | A deep regression model for seed identification in prostate brachytherapy | |
| Lankton | Localized statistical models in computer vision |
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: 20231227 Address after: 200949 Room 922, Zone D, 1st Floor, Building 1, No. 58 Dijie Road, Baoshan District, Shanghai Patentee after: Shanghai Ludong Biotechnology Co.,Ltd. Address before: Room 901, Building 2, No. 111 Xiangke Road, China (Shanghai) Pilot Free Trade Zone, Pudong New Area, Shanghai, 201210 Patentee before: SUILAN INTELLIGENT TECHNOLOGY (SHANGHAI) Co.,Ltd. |