CN111383222A - Intervertebral disc MRI image intelligent diagnosis system based on deep learning - Google Patents
Intervertebral disc MRI image intelligent diagnosis system based on deep learning Download PDFInfo
- Publication number
- CN111383222A CN111383222A CN202010190829.8A CN202010190829A CN111383222A CN 111383222 A CN111383222 A CN 111383222A CN 202010190829 A CN202010190829 A CN 202010190829A CN 111383222 A CN111383222 A CN 111383222A
- Authority
- CN
- China
- Prior art keywords
- data
- intervertebral disc
- module
- training
- subsystem
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000003745 diagnosis Methods 0.000 title claims abstract description 34
- 238000013135 deep learning Methods 0.000 title claims abstract description 13
- 238000012549 training Methods 0.000 claims abstract description 40
- 238000002372 labelling Methods 0.000 claims abstract description 19
- 238000007781 pre-processing Methods 0.000 claims abstract description 13
- 238000000034 method Methods 0.000 claims abstract description 12
- 238000013500 data storage Methods 0.000 claims abstract description 11
- 238000001514 detection method Methods 0.000 claims abstract description 11
- 238000012545 processing Methods 0.000 claims abstract description 11
- 238000013507 mapping Methods 0.000 claims abstract description 6
- 238000006243 chemical reaction Methods 0.000 claims description 16
- 238000012795 verification Methods 0.000 claims description 7
- 238000003062 neural network model Methods 0.000 claims description 5
- 238000003860 storage Methods 0.000 claims description 5
- 230000009466 transformation Effects 0.000 claims description 5
- 238000004140 cleaning Methods 0.000 claims description 4
- 230000010354 integration Effects 0.000 claims description 4
- 230000009467 reduction Effects 0.000 claims description 4
- 238000002595 magnetic resonance imaging Methods 0.000 description 33
- 208000021600 intervertebral disc degenerative disease Diseases 0.000 description 6
- 208000018650 Intervertebral disc disease Diseases 0.000 description 4
- 206010050296 Intervertebral disc protrusion Diseases 0.000 description 4
- 230000036285 pathological change Effects 0.000 description 4
- 231100000915 pathological change Toxicity 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 208000003618 Intervertebral Disc Displacement Diseases 0.000 description 3
- 230000007850 degeneration Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000009826 distribution Methods 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 208000008035 Back Pain Diseases 0.000 description 2
- 206010061246 Intervertebral disc degeneration Diseases 0.000 description 2
- 208000008930 Low Back Pain Diseases 0.000 description 2
- 208000012287 Prolapse Diseases 0.000 description 2
- 208000018180 degenerative disc disease Diseases 0.000 description 2
- 201000010099 disease Diseases 0.000 description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000001684 chronic effect Effects 0.000 description 1
- 238000005520 cutting process Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000005865 ionizing radiation Effects 0.000 description 1
- 230000003902 lesion Effects 0.000 description 1
- 210000004705 lumbosacral region Anatomy 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000005180 public health Effects 0.000 description 1
- 210000004872 soft tissue Anatomy 0.000 description 1
- 238000011179 visual inspection Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
-
- 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
- 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
- G06T2207/10088—Magnetic resonance imaging [MRI]
-
- 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
- G06T2207/30008—Bone
Abstract
The invention discloses an intervertebral disc MRI image intelligent diagnosis system based on deep learning. The method comprises the following steps: the data storage subsystem is used for storing data required by model training; the data processing subsystem is used for preprocessing data, expanding data, labeling data and converting data; a model training subsystem for training a network of modified YOLOv 3; and the real-time display subsystem is used for acquiring the MRI picture of the anterior vertebral disc to be detected uploaded by a user and displaying a diagnosis result. The invention adds an interconnection system structure on the basis of the original YOLOv3, improves the performance of multi-scale target detection by sharing the feature mapping information of different scales, and reduces the number of convolution layers in a detection module to reduce the number of network model parameters, thereby accelerating the training network and the network reasoning speed. The method can realize real-time, automatic and accurate detection and classification of the intervertebral disc MRI pictures, and assist doctors to make the next treatment scheme.
Description
Technical Field
The invention belongs to the technical field of deep learning, and particularly relates to an intervertebral disc MRI image intelligent diagnosis system based on deep learning.
Background
Degeneration of the intervertebral disc (IVD) is one of the major causes of chronic Low Back Pain (LBP), which has become a major public health problem in our society and a major cause of disability. Magnetic Resonance Imaging (MRI) is the first modality to assess degenerative disc disease in the lumbar spine, which provides good soft tissue contrast without the risk of ionizing radiation and is more acceptable to patients. In recent years, the research and application of multi-modality MRI has further improved the quality of diagnosis, treatment and follow-up of a variety of diseases, but these methods still have many problems:
(1) the visual inspection of the intervertebral disc image is time-consuming and labor-consuming, a doctor needs to accurately outline the outline of the intervertebral disc protrusion part and then adopts a corresponding treatment means according to the pathological change degree, however, the process is closely related to the personal experience and the service capability of the doctor, the risk of missed diagnosis or misdiagnosis exists, the process is complicated, the efficiency is low, and the accuracy is relatively low.
(2) Due to the different intensity distribution and deformation degree of the intervertebral disc images, the intervertebral disc images present a wider boundary range in different MRI scanners, and unnecessary data volume is increased, which brings burden to doctors for disease assessment.
In view of this, it is increasingly difficult for traditional research methods relying on the diagnosis of intervertebral disc MRI images by specialized physicians to meet the real-time and accuracy requirements. Therefore, it is a real need in the medical industry to find a reliable automatic diagnosis system for intervertebral disc images. The invention designs an intervertebral disc MRI image intelligent diagnosis system based on deep learning, which is characterized in that case intervertebral disc MRI image data are input and processed, the data comprise data preprocessing, data expansion, data labeling, data conversion and data classification, then the case intervertebral disc MRI image data are trained through an improved YOLOv3 network, and the trained models are stored. When the MRI picture of the intervertebral disc to be detected is transmitted, the pathological change position of the intervertebral disc can be accurately positioned in real time, and a doctor is assisted to formulate a next treatment scheme.
Disclosure of Invention
The invention aims to provide an intervertebral disc MRI image intelligent diagnosis system based on deep learning, which overcomes the defects of the traditional artificial diagnosis and assists doctors in carrying out classified diagnosis on intervertebral disc lesions.
The invention is realized by the following steps: an intervertebral disc MRI image intelligent diagnosis system based on deep learning comprises: the system comprises a data storage subsystem, a data processing subsystem, a model training subsystem and a real-time display subsystem; the data processing subsystem includes: the data conversion module is connected with the data marking module; the real-time display subsystem includes: RESTful API module and result display module, two modules link to each other in order. The subsystems and the modules are connected through a wired and/or WiFi wireless and/or 3G/4G/5G wireless network.
The storage subsystem stores data needed by model training, converted data, a trained model and an intervertebral disc MRI picture to be detected uploaded by a user.
The data processing subsystem comprises a data preprocessing module, a data expansion module, a data labeling module and a data conversion module.
The data preprocessing module is used for carrying out data cleaning, data integration, data reduction and data conversion on the intervertebral disc MRI data in the storage subsystem.
The data expansion module adopts methods including transformation, rotation, clipping, scaling and the like, and is characterized in that the original image is processed and processed by using the traditional mathematical form so as to obtain enough data to support the next training.
And the data labeling module is used for labeling the intervertebral disc area on the data set by using a common data labeling tool Labelme in a polygonal mode and marking a multi-classification label, and the labeled intervertebral disc area is stored as a json file.
And the data conversion module analyzes the json files after the label of the Labelme tool is generated into training available files in batch, wherein the training available files comprise intervertebral disc class files, training set label files, verification set label files and class index files.
The model training subsystem sends the processed intervertebral disc MRI samples to a neural network model with preset darknet53 pre-training weights provided by the darknet for training, improves the architecture of the YOLOv3, increases the interconnection architecture, shares feature mapping information of different scales, and improves the performance of multi-scale target detection.
The real-time display subsystem comprises a RESTful API module and a result display module.
The RESTful API module supports a user to send an HTTPS request to send a picture to be detected through an API provided by the intervertebral disc MRI image intelligent diagnosis system, the picture to be detected is classified in a trained model, meanwhile, the intervertebral disc MRI picture to be diagnosed submitted by the user is marked to form a new sample, the new sample is added into an original sample library, and data in the sample library is increased.
And after the diagnosis is finished, the result display module returns the diagnosis result to the browser or an application program calling the API.
Compared with the prior art, the intervertebral disc MRI image intelligent diagnosis system based on deep learning provided by the invention has the following beneficial effects:
1. the improved YOLOv3 network is trained through a large amount of intervertebral disc MRI picture data to obtain a stable network model, so that automatic high-precision intervertebral disc disease diagnosis is realized, and the burden of doctors on intervertebral disc disease diagnosis is reduced.
2. The improved YOLOv3 network increases an interconnection system structure, and improves the performance of multi-scale target detection by sharing the feature mapping information of different scales. Meanwhile, the number of the convolution layers in the candidate bounding box area is reduced, and the number of the network model parameters is correspondingly reduced, so that the training network and the network reasoning speed are accelerated.
3. By applying a series of data expansion methods, including the processes of intervertebral disc MRI picture transformation, rotation, cutting and scaling in the early stage, and adding the intervertebral disc MRI data to be detected submitted by a user to an original sample library, the data in the sample library is continuously increased, the problem that a more accurate network model is difficult to train due to less obtained intervertebral disc MRI data in the early stage is solved, and the stability of the network model and the accuracy of a diagnosis system are continuously improved.
Drawings
FIG. 1 is a schematic diagram of a system configuration according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a data expansion method of an intervertebral disc MRI picture based on an ROI (region of interest) area according to the invention;
fig. 3 is a network architecture diagram according to an embodiment of the present invention.
Detailed Description
Example (b):
as shown in fig. 1, an embodiment provides a deep learning based intervertebral disc MRI image intelligent diagnosis system, including: the system comprises a data storage subsystem 1, a data processing subsystem 6, a model training subsystem 7 and a real-time display subsystem 10; the data processing subsystem 6 comprises a data preprocessing module 2, a data expansion module 3, a data labeling module 4 and a data conversion module 5, wherein the data preprocessing module 2 is connected with the data expansion module 3, the data expansion module 3 is connected with the data labeling module 4, and the data labeling module 4 is connected with the data conversion module 5; the real-time display subsystem 10 comprises a RESTful API module 8 and a result display module 9, which are connected in sequence. The subsystems and the modules are connected through a wired and/or WiFi wireless and/or 3G/4G/5G wireless network.
The data storage subsystem 1 stores an intervertebral disc MRI picture data set, converted data, a trained model and an intervertebral disc MRI picture to be detected, which is subsequently uploaded by a user and is required to be used by model training, through a local disk database or a cloud disk database.
The data processing subsystem 6 comprises a data preprocessing module 2, a data expansion module 3, a data labeling module 4 and a data conversion module 5.
The data preprocessing module 2 performs data cleaning, data integration, data reduction and data conversion on data in the database, solves the problems of noise, inconsistency, numerical value loss and the like of original data, improves the overall quality of the data, and then stores the data subjected to a data preprocessing stage into the data storage subsystem 1.
The principle of the data expansion module 3 is as shown in fig. 2, and sample data is expanded through operations such as transformation, rotation, clipping, scaling and the like, so that the effective data expansion can expand the number of training samples and increase the diversity of the training samples; the trained model can avoid overfitting and improve the performance of the model.
The data labeling module 4 labels the pictures by using a Labelme tool, and adopts polygon labeling to replace default rectangular labeling, so as to increase the accuracy of later-stage identification and facilitate subsequent example segmentation. The annotation results are stored in the form of json files, and are divided into four categories, the first being normal disc (normal), the second being degeneration of the disc (degeneration), the third being Protrusion of the anterior disc (prolapse), and the fourth being herniated disc (prolapse).
The data conversion module 5 converts the json files generated in the previous step into corresponding data files in batch by performing data conversion on a terminal, wherein the data files comprise intervertebral disc class files, training set label files, verification set label files and class index files, and are stored in the data storage subsystem 1.
The model training subsystem 7 trains, verifies and evaluates a preset neural network model through a training set, a verification set and a test set, the data distribution ratio is 8: 1: 1, an improved YOLOv3 network architecture is shown in fig. 3, an interconnection architecture is added on the basis of an original YOLOv3, the performance of multi-scale target detection is improved by sharing feature mapping information of different scales, meanwhile, the number of convolutional layers in a detection module is reduced to reduce the number of network model parameters, thereby accelerating network training and network inference speed, a gray image with the size of 512 in the data set is input into a training network, a feature extraction layer in the network is responsible for extracting feature maps of three scales (128 ×, 64 ×, 32 2) in the image, the three feature maps are input into a right module for further processing, features under the three scales are extracted respectively, and finally, a semantic block3 and a semantic block 493 classification block are given by a boundary predictor and a semantic block3 for achieving a semantic classification of a semantic classification, a semantic classification rule, a regression learning rate of a Convolution learning rate of 512, a random learning rate, a convolutional learning rate, a Gradient.
The real-time display subsystem 10 comprises a RESTful API module 8 and a result display module 9.
The RESTful API module 8 supports a user to send an HTTPS request to send a picture to be detected through an API provided by the intervertebral disc MRI image intelligent diagnosis system, the picture to be detected is classified in the trained model, and is stored in the data storage subsystem 1 and added into the original intervertebral disc MRI picture sample library, so that data in the sample library is continuously increased, the trained neural network model is more and more stable, and the recognition rate is continuously increased.
And after the diagnosis is finished, the result display module 9 returns the diagnosis result to the browser or an application program calling the API, and if the pathological changes occur, the pathological change area is marked and an alarm is given, so that a doctor is assisted to make a next treatment scheme.
The intervertebral disc MRI image intelligent diagnosis system based on deep learning has the working process as follows:
(1) first, a disc MRI picture stored in the storage subsystem is acquired.
(2) And then preprocessing the picture data, including data cleaning, data integration, data reduction and data conversion, solving the problems of noise, inconsistency, numerical value loss and the like of the original data, and improving the overall quality of the data.
(3) And then, the data set subjected to data preprocessing is subjected to operations such as transformation, rotation, clipping, scaling and the like, and the number of samples is expanded.
(4) And labeling the picture by utilizing a Labelme tool, and replacing default rectangular labeling with polygonal labeling so as to hopefully increase the accuracy of later-stage identification and facilitate subsequent instance segmentation. The labeling results are stored in a json file form and are divided into four types, namely normal intervertebral disc, intervertebral disc degeneration, intervertebral disc protrusion and intervertebral disc prolapse.
(5) And analyzing the json file generated by the labeling in batches into training available files including an intervertebral disc class file, a training set label file, a verification set label file and a class index file, and storing the training available files after the analysis in batches into a data storage subsystem.
(6) And inputting the result processed samples in the data storage subsystem into a modified YOLOv3 network for training, wherein the distribution ratio of the training set, the verification set and the test set is 8: 1: 1, an interconnection architecture is added on the basis of the original YOLOv3, and the performance of multi-scale target detection is improved by sharing feature mapping information of different scales. Meanwhile, the number of convolution layers in the detection module is reduced so as to reduce the number of network model parameters, the training network and the network reasoning speed are accelerated, then the error of the network output result and the label data is calculated, an SGD optimization function is adopted to optimize the network in training until the threshold is met, if the threshold is met, the next step of verification and test is carried out, otherwise, the training is returned to continue, and the trained model is stored in the data storage subsystem.
(7) A user sends an HTTPS request to send a picture to be detected through an API provided by an intervertebral disc MRI image intelligent diagnosis system, the picture to be detected is classified in a trained model, the intervertebral disc MRI picture to be diagnosed submitted by the user is marked to form a new sample, the new sample is added into an original sample library, data in the sample library is continuously increased to deal with special cases which may appear when the model is actually applied, accordingly, the trained neural network model is more and more stable, the identification accuracy rate is continuously increased, and doctors are assisted to better make a next treatment scheme.
The intervertebral disc MRI image intelligent diagnosis system based on deep learning can realize real-time intervertebral disc disease diagnosis, reduce the burden of doctors on intervertebral disc disease diagnosis and improve the detection efficiency and accuracy.
While the foregoing is directed to the preferred embodiment of the present invention, it is understood that the foregoing is illustrative only and is not to be construed as limiting the scope of the invention, as numerous changes and modifications will become apparent to those skilled in the art in light of the foregoing description.
Claims (1)
1. An intervertebral disc MRI image intelligent diagnosis system based on deep learning is characterized in that: the system comprises a data storage subsystem, a data processing subsystem, a model training subsystem and a real-time display subsystem; the data processing subsystem includes: the data conversion module is connected with the data marking module; the real-time display subsystem includes: the RESTful API module and the result display module are sequentially connected, and the subsystems and the modules are connected through a wired and/or WiFi wireless and/or 3G/4G/5G wireless network;
the storage subsystem stores data needed by model training, converted data, a trained model and a to-be-detected intervertebral disc MRI picture uploaded by a user;
the data processing subsystem comprises a data preprocessing module, a data expansion module, a data labeling module and a data conversion module;
the data preprocessing module is used for carrying out data cleaning, data integration, data reduction and data conversion on the intervertebral disc MRI data in the storage subsystem;
the method adopted by the data expansion module comprises the steps of transformation, rotation, clipping and scaling treatment, and is characterized in that the original image is processed and processed by utilizing the traditional mathematical form so as to obtain enough data to support the next training;
the data marking module marks the intervertebral disc area on the data set in a polygon mode by using a common data marking tool Labelme, marks the intervertebral disc area on the data set with a multi-classification label, and saves the marked intervertebral disc area as a json file;
the data conversion module analyzes the json files after the label of the Labelme tool is generated into training available files in batch, wherein the training available files comprise intervertebral disc class files, training set label files, verification set label files and class index files;
the model training subsystem sends the processed intervertebral disc MRI samples into a neural network model with preset darknet53 pre-training weight provided by the darknet for training, improves the architecture of the YOLOv3, increases the interconnection architecture, shares the feature mapping information of different scales, and improves the performance of multi-scale target detection;
the real-time display subsystem comprises a RESTful API module and a result display module;
the RESTful API module supports a user to send an HTTPS request to send a picture to be detected through an API provided by the intervertebral disc MRI image intelligent diagnosis system, the picture to be detected is classified in a trained model, meanwhile, the intervertebral disc MRI picture to be diagnosed submitted by the user is marked to form a new sample, the new sample is added into an original sample library, and data in the sample library is increased;
and after the diagnosis is finished, the result display module returns the diagnosis result to the browser or an application program calling the API.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010190829.8A CN111383222A (en) | 2020-03-18 | 2020-03-18 | Intervertebral disc MRI image intelligent diagnosis system based on deep learning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010190829.8A CN111383222A (en) | 2020-03-18 | 2020-03-18 | Intervertebral disc MRI image intelligent diagnosis system based on deep learning |
Publications (1)
Publication Number | Publication Date |
---|---|
CN111383222A true CN111383222A (en) | 2020-07-07 |
Family
ID=71215429
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010190829.8A Pending CN111383222A (en) | 2020-03-18 | 2020-03-18 | Intervertebral disc MRI image intelligent diagnosis system based on deep learning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111383222A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112330644A (en) * | 2020-11-11 | 2021-02-05 | 复旦大学 | Medical image diagnosis system based on deep learning |
CN112635027A (en) * | 2020-10-29 | 2021-04-09 | 郑州大学第一附属医院 | System, method and terminal for processing MRI (magnetic resonance imaging) classified images of knee osteoarthritis |
TWI819936B (en) * | 2022-12-27 | 2023-10-21 | 財團法人工業技術研究院 | Vertebral canal abnormality determination system and determination device thereof |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107977971A (en) * | 2017-11-09 | 2018-05-01 | 哈尔滨理工大学 | The method of vertebra positioning based on convolutional neural networks |
CN108899075A (en) * | 2018-06-28 | 2018-11-27 | 众安信息技术服务有限公司 | A kind of DSA image detecting method, device and equipment based on deep learning |
CN109256212A (en) * | 2018-08-17 | 2019-01-22 | 上海米因医疗器械科技有限公司 | Bone health assessment models construction method, device, equipment, medium and appraisal procedure |
CN109886346A (en) * | 2019-02-26 | 2019-06-14 | 四川大学华西医院 | A kind of cardiac muscle MRI image categorizing system |
CN110400316A (en) * | 2019-04-19 | 2019-11-01 | 杭州健培科技有限公司 | A kind of orthopaedics image measuring method and device based on deep learning |
CN110533640A (en) * | 2019-08-15 | 2019-12-03 | 北京交通大学 | Based on the track circuit disease discrimination method for improving YOLOv3 network model |
-
2020
- 2020-03-18 CN CN202010190829.8A patent/CN111383222A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107977971A (en) * | 2017-11-09 | 2018-05-01 | 哈尔滨理工大学 | The method of vertebra positioning based on convolutional neural networks |
CN108899075A (en) * | 2018-06-28 | 2018-11-27 | 众安信息技术服务有限公司 | A kind of DSA image detecting method, device and equipment based on deep learning |
CN109256212A (en) * | 2018-08-17 | 2019-01-22 | 上海米因医疗器械科技有限公司 | Bone health assessment models construction method, device, equipment, medium and appraisal procedure |
CN109886346A (en) * | 2019-02-26 | 2019-06-14 | 四川大学华西医院 | A kind of cardiac muscle MRI image categorizing system |
CN110400316A (en) * | 2019-04-19 | 2019-11-01 | 杭州健培科技有限公司 | A kind of orthopaedics image measuring method and device based on deep learning |
CN110533640A (en) * | 2019-08-15 | 2019-12-03 | 北京交通大学 | Based on the track circuit disease discrimination method for improving YOLOv3 network model |
Non-Patent Citations (1)
Title |
---|
任嘉锋等: "基于改进YOLOv3的火灾检测与识别", 《计算机系统应用》, 5 December 2019 (2019-12-05), pages 2 - 3 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112635027A (en) * | 2020-10-29 | 2021-04-09 | 郑州大学第一附属医院 | System, method and terminal for processing MRI (magnetic resonance imaging) classified images of knee osteoarthritis |
CN112330644A (en) * | 2020-11-11 | 2021-02-05 | 复旦大学 | Medical image diagnosis system based on deep learning |
TWI819936B (en) * | 2022-12-27 | 2023-10-21 | 財團法人工業技術研究院 | Vertebral canal abnormality determination system and determination device thereof |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108268870B (en) | Multi-scale feature fusion ultrasonic image semantic segmentation method based on counterstudy | |
Jin et al. | DUNet: A deformable network for retinal vessel segmentation | |
Liu et al. | A framework of wound segmentation based on deep convolutional networks | |
CN111383222A (en) | Intervertebral disc MRI image intelligent diagnosis system based on deep learning | |
Kar et al. | Retinal vessel segmentation using multi-scale residual convolutional neural network (MSR-Net) combined with generative adversarial networks | |
CN110543912B (en) | Method for automatically acquiring cardiac cycle video in fetal key section ultrasonic video | |
CN110246579B (en) | Pathological diagnosis method and device | |
CN114758137B (en) | Ultrasonic image segmentation method and device and computer readable storage medium | |
CN112308822A (en) | Intervertebral disc CT image detection method based on deep convolutional neural network | |
CN113782184A (en) | Cerebral apoplexy auxiliary evaluation system based on facial key point and feature pre-learning | |
CN115147600A (en) | GBM multi-mode MR image segmentation method based on classifier weight converter | |
CN114494215A (en) | Transformer-based thyroid nodule detection method | |
CN114445356A (en) | Multi-resolution-based full-field pathological section image tumor rapid positioning method | |
CN116758336A (en) | Medical image intelligent analysis system based on artificial intelligence | |
CN114332910A (en) | Human body part segmentation method for similar feature calculation of far infrared image | |
Suganya et al. | Ultrasound ovary cyst image classification with deep learning neural network with Support vector machine | |
CN112785581A (en) | Training method and device for extracting and training large blood vessel CTA (computed tomography angiography) imaging based on deep learning | |
CN115359066B (en) | Focus detection method and device for endoscope, electronic device and storage medium | |
CN116486156A (en) | Full-view digital slice image classification method integrating multi-scale feature context | |
Lee et al. | Cardiac CT Image Segmentation for Deep Learning-Based Coronary Calcium Detection Using K-Means Clustering and Grabcut Algorithm. | |
CN111932523B (en) | Gender classifier based on brain imaging big data deep learning | |
Alharthi et al. | Do it the transformer way: A comprehensive review of brain and vision transformers for autism spectrum disorder diagnosis and classification | |
KR102566095B1 (en) | Deep learning apparatus and method for joint classification and segmentation of histopathology image | |
Fatema et al. | Development of an automated optimal distance feature-based decision system for diagnosing knee osteoarthritis using segmented X-ray images | |
CN113269778B (en) | Image weak supervision segmentation method based on iteration |
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 |