CN112001877B - Thyroid malignant nodule detection method based on deep learning - Google Patents
Thyroid malignant nodule detection method based on deep learning Download PDFInfo
- Publication number
- CN112001877B CN112001877B CN202010417483.0A CN202010417483A CN112001877B CN 112001877 B CN112001877 B CN 112001877B CN 202010417483 A CN202010417483 A CN 202010417483A CN 112001877 B CN112001877 B CN 112001877B
- Authority
- CN
- China
- Prior art keywords
- model
- detection
- training
- target
- deep learning
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 53
- 230000003211 malignant effect Effects 0.000 title claims abstract description 27
- 210000001685 thyroid gland Anatomy 0.000 title claims abstract description 27
- 238000013135 deep learning Methods 0.000 title claims abstract description 22
- 238000012549 training Methods 0.000 claims abstract description 31
- 238000000034 method Methods 0.000 claims abstract description 15
- 238000002372 labelling Methods 0.000 claims abstract description 6
- 230000000694 effects Effects 0.000 claims description 10
- 238000012360 testing method Methods 0.000 claims description 7
- 238000011156 evaluation Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 5
- 238000005520 cutting process Methods 0.000 claims description 4
- 238000007781 pre-processing Methods 0.000 claims description 4
- 238000012795 verification Methods 0.000 claims description 4
- 238000012935 Averaging Methods 0.000 claims description 3
- 238000004458 analytical method Methods 0.000 claims description 2
- 238000004519 manufacturing process Methods 0.000 claims description 2
- 206010054107 Nodule Diseases 0.000 abstract description 9
- 238000003745 diagnosis Methods 0.000 abstract description 8
- 238000005516 engineering process Methods 0.000 description 3
- 238000000605 extraction Methods 0.000 description 3
- 208000009453 Thyroid Nodule Diseases 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 210000001525 retina Anatomy 0.000 description 2
- 238000002604 ultrasonography Methods 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000036770 blood supply Effects 0.000 description 1
- 230000002308 calcification Effects 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 230000003902 lesion Effects 0.000 description 1
- 238000010801 machine learning Methods 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
- 238000012545 processing Methods 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
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
- 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
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H30/00—ICT specially adapted for the handling or processing of medical images
- G16H30/20—ICT specially adapted for the handling or processing of medical images for handling medical images, e.g. DICOM, HL7 or PACS
-
- G—PHYSICS
- 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
- G06T2207/10132—Ultrasound image
-
- 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)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Radiology & Medical Imaging (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Medical Informatics (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Biophysics (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Epidemiology (AREA)
- Primary Health Care (AREA)
- Public Health (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
- Ultra Sonic Daignosis Equipment (AREA)
Abstract
The invention relates to a thyroid malignant nodule detection method based on deep learning, which is used for auxiliary diagnosis of thyroid malignant nodules and can realize automatic labeling of a region of interest after training of a large amount of data under the characteristics of low resolution, low precision and low target and background identification of medical images. Errors caused by subjective factors can be effectively reduced, and a radiologist is helped to diagnose rapidly and accurately. The method achieves considerable height in detection precision and speed and has potential of clinical application.
Description
Technical Field
The invention belongs to the field of image processing, relates to a deep learning technology and a target detection technology, and particularly relates to a thyroid malignant nodule detection method based on deep learning.
Background
The ultrasonic diagnosis technology is widely applied to clinic due to the advantages of convenient examination, low cost and the like. Studies have shown that ultrasound is advantageous in identifying benign and malignant thyroid nodules because the benign and malignant thyroid nodules in ultrasound images differ significantly in size, morphology, number, cystic changes, calcification, blood supply, etc., and thus clinicians can determine the nature of the nodules based on these characteristics. However, ultrasonic diagnosis requires a doctor to manually mark a lesion area, and is heavy in workload, and doctors with different experiences and levels often have subjectivity on diagnosis of the result. Therefore, new techniques are needed to extract the nodule region quickly and accurately, and then extract a series of effective features from the extracted nodule region for discriminating the benign and malignant nodules.
Deep learning is one of the technical and research fields of machine learning, and artificial intelligence is implemented in a computing system by establishing an artificial neural network having a hierarchical structure. Because hierarchical ANN can extract and screen input information layer by layer, deep learning has characteristic learning ability, and end-to-end supervised learning and unsupervised learning can be realized. The hierarchical ANN used for deep learning has various forms, and the complexity of the hierarchy is called "depth". The deep learning forms include multi-layer perceptrons, convolutional neural networks, recurrent neural networks, deep belief networks, and other hybrid architectures, depending on the type of architecture. Deep learning uses data to update parameters in its construction to achieve training goals, a process known as "learning". Common methods of learning are gradient descent algorithms and variants thereof, and some statistical learning theory is used for optimization of the learning process.
The object detection, also called object extraction, is an image segmentation based on the geometric and statistical characteristics of the object, which combines the segmentation and recognition of the object into one, and the accuracy and the real-time performance are an important capability of the whole system. Especially in complex scenes, when multiple targets need to be processed in real time, automatic extraction and recognition of the targets are particularly important. The main current target detection algorithms mainly have two directions: one-stage and two-stage.
Disclosure of Invention
The invention aims to overcome the defects of the prior art, and provides a thyroid malignant nodule detection method based on deep learning, which is used for auxiliary diagnosis of thyroid malignant nodules and can realize automatic labeling of a region of interest after training of a large amount of data under the characteristics of low resolution, low precision and low target and background recognition of medical images. Errors caused by subjective factors can be effectively reduced, and a radiologist is helped to diagnose rapidly and accurately. The method achieves considerable height in detection precision and speed and has potential of clinical application.
The invention solves the technical problems by the following technical proposal:
a thyroid malignant nodule detection method based on deep learning is characterized by comprising the following steps: the method comprises the following steps:
1) Reading ultrasonic image data by using a python programming program, and preprocessing;
2) Constructing a target detection network by adopting a Faster-RCNN algorithm, and adjusting and training a model;
3) Constructing a target detection network by adopting RETINANET algorithm, and adjusting and training a model;
4) And detecting and evaluating target detection networks constructed by the two algorithms, and comparing the effects and the analysis results.
Moreover, the specific operation of preprocessing the ultrasonic image data in the step 1) is as follows:
a. Cutting out the picture, namely cutting out additional marks such as equipment name models, patient privacy information and the like contained in the input ultrasonic image;
b. storing pictures into JPEGImages files and naming the pictures in a unified format to generate a txt file under a Main folder, wherein the txt file comprises a verification set, a training set and a test set picture number;
c. Labeling the picture by using labelImg tools, establishing a box frame for the target object, and storing and generating an xml file which comprises the class name of the target and the coordinates of the predicted frame.
Moreover, said step 2) adjusts and trains the specific operation of the model as:
a. Training on the basis of pretrained VGG16 weight on an ImageNet by adopting BackBone network as VGG16, and manufacturing an image classification dataset aiming at thyroid ultrasonic nodules by taking the format of the VOC2007 dataset as a standard;
b. Changing the target categories in pascal _voc. Py and demo. Py into backgroups and jiejie, changing the category number in demo. Py into 2, calculating the pixel average value of the data set, and updating the pix_mean list in config. Py;
c. setting initial parameters LEARNINGRATE, BATCHSIZE, maximum iteration times and stepsize of the model, adjusting a data set, the iteration times and the size of an anchor according to a detection result, and formally training the detection model after the intersection ratio of a predicted target area and a real target area is close to 100 percent and the predicted frame of the model and the real target area are close to each other.
Moreover, said step 3) adjusts and trains the specific operation of the model as:
a. The adopted BackBone network is ResNet, a model of ResNet50 pre-trained on the ImageNet is downloaded, and the model is stored in a corresponding folder;
b. Setting initial parameters Scorethreshold and IoUthreshold of the model, adjusting the data set, the iteration times and the size of the anchor according to the detection result, and formally training the detection model after multiple adjustments.
In addition, in the step 4), mAP is used as an evaluation index in the detection evaluation to evaluate and verify the effect of the algorithm, and meanwhile, the detection accuracy and recall rate of the model and the intersection ratio (IoU) of the predicted target frame and the real target frame are required to be calculated, and the calculation formula is as follows:
wherein: TP means that the positive sample is correctly identified as a positive sample;
FP is the negative sample that is erroneously identified as a positive sample;
FN is positive samples that are erroneously identified as negative samples;
AP is the area enclosed by the curve formed by Precision in the formula (1) and Recall in the formula (2);
k represents k categories;
mAP is the average of k classes of APs;
m represents mean, which is the value of mAP obtained by re-averaging the APs of each class.
The invention has the advantages and beneficial effects that:
1. According to the thyroid malignant nodule detection method based on deep learning, the thyroid malignant nodule is assisted to be detected by training a model through training set data and adjusting model parameters through verification set data.
2. According to the thyroid malignant nodule detection method based on deep learning, experimental effects show that the fast-RCNN algorithm and the RETINANET algorithm have the feasibility of clinical application for assisting thyroid malignant nodule detection, wherein the RETINANET algorithm is adopted, and after the number of training sets and the iterative training times are adjusted, a target detection network based on deep learning for thyroid malignant nodule detection has higher detection accuracy; by adopting the Faster-RCNN algorithm, the fitting phenomenon is not easy to occur, and the robustness of the trained model is good.
3. The thyroid malignant nodule detection method based on deep learning is used for auxiliary diagnosis of thyroid malignant nodules, and can realize automatic labeling of a region of interest after training of a large amount of data under the characteristics of low resolution, low precision and low target and background identification of medical images; errors caused by subjective factors can be effectively reduced, and a radiologist is helped to diagnose rapidly and accurately; the method achieves considerable height in detection precision and speed and has potential of clinical application.
Drawings
Fig. 1 is a diagram of a network architecture of the present invention.
Detailed Description
The invention is further illustrated by the following examples, which are intended to be illustrative only and not limiting in any way.
A thyroid malignant nodule detection method based on deep learning is characterized by comprising the following steps: the method comprises the following steps:
s0101: inputting ultrasonic medical images for model training and detection, wherein the program removes some extra marks contained in the image data, including information such as the size of a nodule, the name and model of equipment, patient privacy and the like;
S0102: storing pictures into JPEGImages files and naming the pictures in a unified format;
S0103: dividing a data set to generate a txt file under a Main folder, wherein the txt file comprises a verification set, a training set and a test set picture number, labeling the picture by using labelImg tools, establishing a box frame for a target object, and then storing and generating an xml file, wherein the xml file comprises a class name of the target and a coordinate of a predicted frame;
s0201: adjustment of the model using the fast-RCNN algorithm was started. The adopted BackBone network is VGG16, training is carried out on the basis of pretrained VGG16 weight on the ImageNet, an image classification dataset aiming at thyroid ultrasonic nodules is manufactured by taking the format of a VOC2007 dataset as a standard, target categories in pascal _voc.py and demo.py are changed into background and jiejie, the category number in demo.py is changed into 2, the pixel average value of the dataset is calculated, and a pix_mean table in config.py is updated;
first, LEARNINGRATE is set to 0.001, the batch size is set to 100, the maximum iteration number is set to 500, the stepsize is set to 400, at this time, the object recognition degree of the model is extremely low, any feature of the object is hardly learned, the overall characteristic of the object cannot be grasped, so the data set dividing ratio is adjusted, and the data set is enlarged. If the effect is still poor after adjustment, iteration is added, the anchor size is changed, and along with the step of parameter adjustment, the predicted frame of the model and the real target area are more and more close;
S0202: gradually increasing iterative training times, 1000 times and 2000 times until 10000 times of the model of the adjusted Faster-RCNN algorithm, knowing the training process by acquiring training information printed and output by a control console, saving the control console information as a txt format file, writing a program to search a matching character string in the txt file, for example 'totalloss', extracting the subsequent numerical value and storing the numerical value into an array to obtain a total loss value, drawing the loss value array through drawing software to obtain a loss curve, and testing the model by utilizing testing set data;
S0301: adjustment of the model using RETINANET algorithm is started. The BackBone network adopted is ResNet, and a model of ResNet50 pre-trained on the ImageNet is downloaded first and stored in a corresponding folder. Setting initial model parameters Scorethreshold to 0.05 and iouthreshold to 0.5, wherein the two parameters relate to the calculation of the mAP, and a prediction target frame with a confidence score threshold lower than 0.05 is discarded and does not participate in the calculation of the mAP;
S0302: and adding iteration training times to the model of the RETINANET algorithm after adjustment, and training 10000 iterations, 20000 iterations, 30000 iterations and 40000 iterations. The training process is known by acquiring training information printed and output by a control console, the control console information is stored as a txt format file, a program is written to search matching character strings in the txt file, such as 'totalloss', the subsequent numerical values are extracted and stored into an array to be the total loss value, the loss value array is drawn through drawing software to obtain a loss curve, and a testing set data testing model is utilized;
S0401: by calculating and comparing the mAP index, the effect of the target detection network constructed by the two algorithms can be evaluated and verified, a proper model is selected,
Meanwhile, the detection accuracy, recall rate and the intersection ratio (IoU) of the predicted target frame and the real target frame of the model are required to be calculated, and the calculation formula is as follows:
wherein: TP means that the positive sample is correctly identified as a positive sample;
FP is the negative sample that is erroneously identified as a positive sample;
FN is positive samples that are erroneously identified as negative samples;
AP is the area enclosed by the curve formed by Precision in the formula (1) and Recall in the formula (2);
k represents k categories;
mAP is the average of k classes of APs;
m represents mean, which is the value of mAP obtained by re-averaging the APs of each class.
The experimental effects of the comparison experiment are shown in the table 1 and the table 2, and the mAP (maximum performance) detection effect is steadily improved along with the increase of iteration times by a target detection network adopting the fast-RCNN algorithm, and the target detection network reaches 0.765 after 10000 times of iteration training, so that the target detection network has feasibility; the detection effect of the target detection network adopting RETINA NET algorithm is obviously better, but the fitting phenomenon occurs after the iteration times exceed 10000, which can cause poor model robustness, poor adaptation to new data and influence the detection performance.
Table 1 Faster-RCNN Experimental evaluation index Table
| Number of iterations | 1000 | 2000 | 3000 | 4000 | 5000 | 6000 | 7000 | 8000 | 9000 | 10000 |
| mAP | 0.091 | 0.104 | 0.359 | 0.474 | 0.488 | 0.549 | 0.689 | 0.661 | 0.723 | 0.765 |
Table 2 Retina Net Experimental evaluation index Table
| Number of iterations | 10000 | 20000 | 30000 | 40000 |
| mAP | 0.9750 | 0.9796 | 0.9770 | 0.9740 |
With reference to tables 1 and 2, the two algorithms have advantages when applied to constructing a deep learning-based target detection network for thyroid malignant nodule detection, and both have feasibility of assisting in clinical application of thyroid malignant nodule detection.
The thyroid malignant nodule detection method based on deep learning detects thyroid malignant nodule ultrasonic images, processes and learns ultrasonic image information, adopts a deep learning method, automatically marks an area of interest, identifies thyroid malignant nodules and assists diagnosis; two algorithms with advantages are adopted to construct a model, so that feature extraction can be automatically realized, the value of a convolution kernel can be adaptively adjusted, the error between a model predicted value and a true value can be minimized, and a proper model can be selected according to the actual requirement; ultrasonic image data of thyroid malignant nodules are learned in a large quantity through a deep learning method, thyroid malignant nodule characteristics are captured more comprehensively and in detail, and diagnosis accuracy and speed are improved.
Although the embodiments of the present invention and the accompanying drawings have been disclosed for illustrative purposes, those skilled in the art will appreciate that: various substitutions, changes and modifications are possible without departing from the spirit and scope of the invention and the appended claims, and therefore the scope of the invention is not limited to the embodiments and the disclosure of the drawings.
Claims (1)
1. A thyroid malignant nodule detection method based on deep learning is characterized by comprising the following steps: the method comprises the following steps:
1) Reading ultrasonic image data by using a python programming program, and preprocessing;
2) Constructing a target detection network by adopting a Faster-RCNN algorithm, and adjusting and training a model;
3) Constructing a target detection network by adopting RETINANET algorithm, and adjusting and training a model;
4) Detecting and evaluating a target detection network constructed by two algorithms, and comparing the effect and the analysis result;
The specific operation of preprocessing the ultrasonic image data in the step 1) is as follows:
a. Cutting out the picture, namely cutting out additional marks such as equipment name models, patient privacy information and the like contained in the input ultrasonic image;
b. storing pictures into JPEGImages files and naming the pictures in a unified format to generate a txt file under a Main folder, wherein the txt file comprises a verification set, a training set and a test set picture number;
c. Labeling the picture by using labelImg tools, establishing a box frame for the target object, and then storing and generating an xml file which comprises the class name of the target and the coordinates of the predicted frame;
The step 2) adjusts and trains the specific operation of the model as follows:
a. Training on the basis of pretrained VGG16 weight on an ImageNet by adopting BackBone network as VGG16, and manufacturing an image classification dataset aiming at thyroid ultrasonic nodules by taking the format of the VOC2007 dataset as a standard;
b. changing the target categories in pascal _voc. Py and demo. Py into backgroups and jiejie, changing the category number in demo. Py into 2, calculating the pixel average value of the data set, and updating the pix_mean list in config. Py;
c. Setting initial parameters LEARNINGRATE, BATCHSIZE, maximum iteration times and stepsize of a model, adjusting a data set, the iteration times and the size of an anchor according to a detection result, and formally training a detection model after the intersection ratio of a predicted target area and a real target area is close to 100 percent and the predicted frame of the model and the real target area are close to each other;
said step 3) adjusts and trains the specific operation of the model as:
a. the adopted BackBone network is ResNet, a model of ResNet50 pre-trained on the ImageNet is downloaded, and the model is stored in a corresponding folder;
b. setting initial parameters Scorethreshold and IoUthreshold of the model, adjusting a data set, iteration times and the size of an anchor according to the detection result, and formally training the detection model after multiple adjustments;
in the step 4), mAP is used as an evaluation index in the detection evaluation to evaluate and verify the effect of the algorithm, and meanwhile, the detection accuracy and recall rate of the model and the intersection ratio (IoU) of the predicted target frame and the real target frame are required to be calculated, and the calculation formula is as follows:
wherein: TP means that the positive sample is correctly identified as a positive sample;
FP is the negative sample that is erroneously identified as a positive sample;
FN is positive samples that are erroneously identified as negative samples;
AP is the area enclosed by the curve formed by Precision in the formula (1) and Recall in the formula (2);
k represents k categories;
mAP is the average of k classes of APs;
m represents mean, which is the value of mAP obtained by re-averaging the APs of each class.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010417483.0A CN112001877B (en) | 2020-05-18 | 2020-05-18 | Thyroid malignant nodule detection method based on deep learning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010417483.0A CN112001877B (en) | 2020-05-18 | 2020-05-18 | Thyroid malignant nodule detection method based on deep learning |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112001877A CN112001877A (en) | 2020-11-27 |
| CN112001877B true CN112001877B (en) | 2024-08-16 |
Family
ID=73461765
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010417483.0A Active CN112001877B (en) | 2020-05-18 | 2020-05-18 | Thyroid malignant nodule detection method based on deep learning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112001877B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112927808A (en) * | 2021-03-01 | 2021-06-08 | 北京小白世纪网络科技有限公司 | Thyroid ultrasound image-based nodule grading system and method |
| CN114820455A (en) * | 2022-03-31 | 2022-07-29 | 华中科技大学同济医学院附属同济医院 | Ovarian cancer ultrasonic image processing method, system, medium, equipment and terminal |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110390691A (en) * | 2019-06-12 | 2019-10-29 | 合肥合工安驰智能科技有限公司 | A kind of ore scale measurement method and application system based on deep learning |
| CN110490892A (en) * | 2019-07-03 | 2019-11-22 | 中山大学 | A kind of Thyroid ultrasound image tubercle automatic positioning recognition methods based on USFaster R-CNN |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109447940B (en) * | 2018-08-28 | 2021-09-28 | 天津医科大学肿瘤医院 | Convolutional neural network training method, ultrasonic image identification and positioning method and system |
| CN109726658B (en) * | 2018-12-21 | 2022-10-04 | 上海科技大学 | Crowd counting and positioning method and system, electronic terminal and storage medium |
| CN109830303A (en) * | 2019-02-01 | 2019-05-31 | 上海众恒信息产业股份有限公司 | Clinical data mining analysis and aid decision-making method based on internet integration medical platform |
| CN110223287A (en) * | 2019-06-13 | 2019-09-10 | 首都医科大学北京友谊医院 | A method of early diagnosing mammary cancer rate can be improved |
| CN110931112B (en) * | 2019-12-05 | 2023-04-18 | 电子科技大学中山学院 | Brain medical image analysis method based on multi-dimensional information fusion and deep learning |
-
2020
- 2020-05-18 CN CN202010417483.0A patent/CN112001877B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110390691A (en) * | 2019-06-12 | 2019-10-29 | 合肥合工安驰智能科技有限公司 | A kind of ore scale measurement method and application system based on deep learning |
| CN110490892A (en) * | 2019-07-03 | 2019-11-22 | 中山大学 | A kind of Thyroid ultrasound image tubercle automatic positioning recognition methods based on USFaster R-CNN |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112001877A (en) | 2020-11-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3876190B1 (en) | Endoscopic image processing method and system and computer device | |
| CN111476292B (en) | Small sample element learning training method for medical image classification processing artificial intelligence | |
| US11288550B2 (en) | Data processing apparatus and method, recognition apparatus, learning data storage apparatus, machine learning apparatus, and program | |
| CN107464250B (en) | Automatic breast tumor segmentation method based on three-dimensional MRI (magnetic resonance imaging) image | |
| CN108389201B (en) | Lung nodule benign and malignant classification method based on 3D convolutional neural network and deep learning | |
| CN107610087B (en) | Tongue coating automatic segmentation method based on deep learning | |
| CN107862694A (en) | A kind of hand-foot-and-mouth disease detecting system based on deep learning | |
| CN112101451A (en) | Breast cancer histopathology type classification method based on generation of confrontation network screening image blocks | |
| CN110738247B (en) | Fine-grained image classification method based on selective sparse sampling | |
| CN112149717A (en) | Confidence weighting-based graph neural network training method and device | |
| CN112001877B (en) | Thyroid malignant nodule detection method based on deep learning | |
| Lan et al. | Run: Residual u-net for computer-aided detection of pulmonary nodules without candidate selection | |
| CN111882538A (en) | Processing system and information processing method for distinguishing pulmonary tuberculosis information and tumor information | |
| CN110634132A (en) | Method for automatically generating tuberculosis quantitative diagnosis report based on deep learning 3D CT image | |
| CN113139928B (en) | Training method of lung nodule detection model and lung nodule detection method | |
| CN116091490A (en) | Lung nodule detection method based on YOLOv4-CA-CBAM-K-means++ -SIOU | |
| Ramkumar et al. | Multi res U-Net based image segmentation of pulmonary tuberculosis using CT images | |
| CN118469950A (en) | Lung cancer intelligent detection system based on image processing and machine learning | |
| CN117975170A (en) | Medical information processing method and system based on big data | |
| CN108681731A (en) | A kind of thyroid cancer ultrasound picture automatic marking method and system | |
| CN109948706B (en) | Micro-calcification cluster detection method combining deep learning and feature multi-scale fusion | |
| WO2020199692A1 (en) | Method and apparatus for screening predictive image features for cancer metastasis, and storage medium | |
| CN110837844A (en) | Pancreatic cystic tumor benign and malignant classification method based on CT image dissimilarity characteristics | |
| CN113436027B (en) | Medical insurance reimbursement abnormal data detection method and system | |
| TWI781000B (en) | Machine learning device and method |
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 |