CN108852268A - A kind of digestive endoscopy image abnormal characteristic real-time mark system and method - Google Patents
A kind of digestive endoscopy image abnormal characteristic real-time mark system and method Download PDFInfo
- Publication number
- CN108852268A CN108852268A CN201810365426.5A CN201810365426A CN108852268A CN 108852268 A CN108852268 A CN 108852268A CN 201810365426 A CN201810365426 A CN 201810365426A CN 108852268 A CN108852268 A CN 108852268A
- Authority
- CN
- China
- Prior art keywords
- image
- model
- training
- module
- stomach
- 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
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B1/00—Instruments for performing medical examinations of the interior of cavities or tubes of the body by visual or photographical inspection, e.g. endoscopes; Illuminating arrangements therefor
- A61B1/273—Instruments for performing medical examinations of the interior of cavities or tubes of the body by visual or photographical inspection, e.g. endoscopes; Illuminating arrangements therefor for the upper alimentary canal, e.g. oesophagoscopes, gastroscopes
- A61B1/2736—Gastroscopes
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B1/00—Instruments for performing medical examinations of the interior of cavities or tubes of the body by visual or photographical inspection, e.g. endoscopes; Illuminating arrangements therefor
- A61B1/00002—Operational features of endoscopes
- A61B1/00004—Operational features of endoscopes characterised by electronic signal processing
- A61B1/00009—Operational features of endoscopes characterised by electronic signal processing of image signals during a use of endoscope
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B1/00—Instruments for performing medical examinations of the interior of cavities or tubes of the body by visual or photographical inspection, e.g. endoscopes; Illuminating arrangements therefor
- A61B1/00002—Operational features of endoscopes
- A61B1/00043—Operational features of endoscopes provided with output arrangements
Abstract
The invention discloses a kind of digestive endoscopy image abnormal characteristic real-time mark system and method, which includes image collection module, image pre-processing module, model training module, abnormality detection module and label display module；Model training module includes image data set, disaggregated model training unit and detection model training unit, using classification of diseases information before the suspicious stomach cancer of deep learning CNN disaggregated model acquisition, lesions position is quick and precisely obtained based on the target detection model of homing method using deep learning CNN.Utilize the present invention, help that stomach off-note under digestive endoscopy is effectively classified and detected, the rate of missed diagnosis based on doctor's long-time, subjective diagnosis can be reduced, and support doctor's analysis, real-time display Microendoscopic suspicious lesions in real time when implementing endoscopy, doctor's work load is reduced, the efficiency of medical diagnosis work is improved.
Description
Technical field
The invention belongs to medical data excavation applications, more particularly to a kind of digestive endoscopy image abnormal characteristic real-time mark
System and method.
Background technique
China's gastric cancer newly sends out gastric cancer 40.5 ten thousand, 32.5 ten thousand dead every year, accounts for 42.6% He of global total amount respectively
45.0%, it reduces China's incidence gastric cancer rate and the death rate is public health problem urgently to be resolved.Clinical research shows gastric cancer
Prognosis is closely related with therapeutic effect.For suffering from the patient of advanced gastric carcinoma (advanced gastric cancer, AGC),
Even if receiving the gastric cancer resection operation based on surgical operation, the postoperative five-year survival rate of patient is still below 30%, and patient is postoperative
Quality of life is low, brings great burden to family and society.If patient early gastric caacer receive in time endoscopy with
Treatment, five-year survival rate is up to 90%, it might even be possible under scope to early carcinoma of stomach (early gastric cancer,
EGC radical treatment) is carried out.Therefore, early discovery, early diagnosis, early treatment EGC are saved incidence gastric cancer rate and the death rate is reduced
Medical resource is of great significance.
Disease refers to the benign disease of stomach before the cancer of stomach, is the Major Risk Factors for causing gastric cancer comprising chronic to wither
Contracting gastritis, polyp of stomach, gastric ulcer, residual stomach and gastritis verrucosa etc..As main gastric precancerous lesion atrophic gastritis, cancer
Variability is 8.6~13.8%, and China is 1.2~7.1%.Existing result of study shows periodically to supervise gastric precancerous lesion
It surveys, can make the recall rate of early carcinoma of stomach is more than 50%.And gastric ulcer has the canceration rate of 1-2%.Therefore to the cancer for having gastric cancer risk
Preceding Disease should carry out facilitating cost-effective monitoring in early days, to be intervened, reduce the generation of gastric cancer.
Currently, the monitoring of gastric precancerous lesion canceration mainly uses ordinary optical endoscopic biopsy technique.In endoscopy,
The narrow problem in the scope visual field is usually made troubles to doctor：Such as due to the limitation in the visual field, doctor must be repeatedly in inspection
Ensure that all lesions are found to fail to pinpoint a disease in diagnosis in order to avoid bringing in target organ inner wall surface moving lens.Therefore, develop one kind
Reliably, quickly the exception based on digestion endoscopic images in face of big data quantity of auxiliary doctor progress endoscopy especially is special
It is very necessary to levy real-time mark system, which can be used for that doctor is assisted to carry out early carcinomatous change screening.
Deep learning is a kind of based on the method for carrying out representative learning to data, convolutional neural networks in machine learning
(Convolutional Neural Network, CNN) is a kind of deep learning feed forward-fuzzy control, is widely used in
Image classification, segmentation and target detection.CNN is widely used in the feature extraction and classification and Detection of picture, is based especially on candidate
The target detection CNN in region, accuracy of identification are being continuously improved always, but due to requiring model training parameter up to a million,
Time performance is lower, it is difficult to meet the needs of real-time detection gastroscope image under gastroscope video scene.
Summary of the invention
The present invention provides a kind of digestive endoscopy image abnormal characteristic real-time mark system and method.To stomach under digestive endoscopy
Disease is effectively classified and is detected before cancer, can reduce the rate of missed diagnosis based on doctor's long-time, subjective diagnosis, and support to cure
The raw analysis, real-time display Microendoscopic suspicious lesions in real time when implementing endoscopy, reduces doctor's work load, improves doctor
Treat the efficiency of diagnostic work.
A kind of digestive endoscopy image abnormal characteristic real-time mark system, including computer system, the computer system packet
Contain：
Image collection module obtains the stomach Conventional white endoscopic video of input by endoscopic images system in real time
Stream rejects wherein invalid frame, screens effective key frame；
Image pre-processing module, for carrying out image enhancement processing to the key frame filtered out；
Model training module, for carrying out model instruction to disease endoscope figure before early stage stomach cancer and lesions position information
Practice, obtains detection model；
Abnormality detection module, for by the image sequence for passing through image preprocessing be input to training completion detection model it
In, obtain classification of diseases information and lesion localization information before stomach cancer；
Display module is marked, location information is made a mark in sequence of pictures, and label is mapped to original input
In stomach Conventional white endoscopic video stream, and real-time display is carried out to the described stomach scope video flowing.
Digestive endoscopy image abnormal characteristic mainly includes：Stomach polyp, stomach ulcer, stomach erosion and stomach atrophy.
The image collection module is selected from data object when screening effective key frame using K-means algorithm at random
K object is selected as initial cluster center, is classified according to minimum distance criterion to data object, by the clustering algorithm,
Image in image library is divided into k class.
When making gastroscope, doctor has various operations in the stomach of patient, and such as inflation is deflated, clear water flushing, biopsy calibration
Deng, these operations can influence in varying degrees gastroscope picture quality, these images, which are rejected, or are ignored will effectively reduce calculation amount,
Key frame is screened using K-means algorithm, the efficiency of key-frame extraction can be greatly speeded up.
Image enhancement processing includes in described image preprocessing module：Image normalization, inactive pixels are cut, image is flat
Sliding, image sharpening and image scaling.
The model training module includes：
Image data set, for storing categorized data set and detection data collection based on digestive endoscopy；Categorized data set packet
Containing disease endoscope figure before a variety of early stage stomach cancers, detection data collection includes lesions position information；
Disaggregated model training unit, for carrying out the training of deep learning CNN disaggregated model to the categorized data set；
Detection model training unit, for using CNN disaggregated model as pre-training model, using the detection data collection as
Training set carries out the training of the deep learning CNN target detection model based on homing method.
Described image data set by translation transformation, mirror image switch or random cropping for being carried out before training CNN model
Data amplification.The step can further enhance the generalization ability of CNN, the training effect of lift scheme
The disaggregated model training unit introduces the side of intensified learning when carrying out deep learning CNN disaggregated model training
Method is alleviated using layer-by-layer dimension normalization layer to network regularization using Dropout method with the accuracy for improving network class
The problem of over-fitting, avoids gradient disappearance problem using ReLU activation primitive, by transfer learning mode to the instruction of disaggregated model
Practice efficiency to be promoted.
The detection model obtained after the model training module training includes disaggregated model, auxiliary network structure and target inspection
Survey model；The disaggregated model is based on VGG-16 frame, for obtaining classification information；The auxiliary network structure is for extracting figure
Piece feature；The target detection model is for obtaining lesions position information.
The present invention also provides a kind of digestive endoscopy image abnormal characteristic real-time mark methods, include the following steps：
(1) disease endoscope figure before early stage stomach cancer and lesions position information input model training module are trained,
Obtain detection model；
(2) subject's stomach Conventional white endoscopic video stream is obtained using endoscopic images system equipment；
(3) key-frame extraction will be carried out after the parsing of gastroscope video flowing, rejects invalid frame, it is effective that screening obtains stomach endoscope
Sequence of video images；
(4) effective video image sequence is input to image pre-processing module, carries out image preprocessing；
(5) image sequence after image preprocessing is input in detection model, before detection model exports suspicious stomach cancer
Classification of diseases result and classification confidence, while exporting lesion localization coordinate and positioning confidence level；
(6) it is marked in effective image sequence according to lesion localization coordinate, while label is mapped to former gastroscope and is regarded
In frequency stream；
(7) real-time display diagnoses labeled gastroscope video flowing over the display so that doctor observes confirmation.
The beneficial effects of the invention are as follows：
1, provide a kind of digestive endoscopy check in the technology of Computer-Aided Classification, detection is carried out to off-note, can
Suitable for carrying out effectively detecting the presence of suspicious lesions to disease before stomach cancer, find lesion classification, and to lesion specific location into
Row is accurately positioned, and helps to find minimal disease, lesion is avoided to omit.
2, convolutional neural networks of the method for the present invention based on deep learning, for the geometric transformation of picture to be detected, deformation,
Illumination has a degree of invariance, and the effect of tagsort is good, and it is to be detected to scan whole picture with lesser calculating cost
Image, for digestive endoscopy big data image procossing, this method can high efficiency complete clinical endoscopic image data end to end
Classification and detection.
3, present approach provides one kind to be able to achieve real-time adjuvant clinical detection means, and tradition is based on candidate region
Even if the deep learning model of CNN target detection possesses higher accuracy rate, it is also difficult to reach the calculating speed of real-time detection, it is past
Toward needing to shorten detection time as cost to sacrifice detection accuracy, the method for the present invention is to reduce under the premise of guaranteeing detection accuracy
Calculation amount, is extracted valid frame to scope video flowing, eliminates invalid frame, makes image sequence utmostly by image preprocessing
Prominent features information, and it is taken based on the deep learning object detection method of recurrence, it can adapt to handle in real time, show in real time
The demand shown, clinician can realize observed in real time during digestive endoscopy inspection classification, detection as a result, being not necessarily to
Fall into a long wait, can effectively accelerate diagnosis process, early discovery early treatment is realized to lesion.
4, digestive endoscopy image sequence of the technology based on image big data can assist alleviating doctor's high intensity, when long
Between diagosis work, avoid because of working strength and doctor's subjective judgement caused by the working time made mistakes, reduce doctor's work load,
Improve the efficiency of medical diagnosis work.
Detailed description of the invention
Fig. 1 is digestive endoscopy image abnormal characteristic real-time mark system structure diagram of the present invention；
Fig. 2 is digestive endoscopy image abnormal characteristic real-time mark method flow diagram of the present invention；
Fig. 3 is the valid frame image of stomach under digestive endoscopy of the embodiment of the present invention；
Fig. 4 is the invalid frame image of stomach under digestive endoscopy of the embodiment of the present invention；
Fig. 5 be under digestive endoscopy of the embodiment of the present invention before stomach cancer in disease stomach polyp original image；
Fig. 6 is stomach polyp map in disease before stomach cancer under digestive endoscopy of the embodiment of the present invention after image enhancement processing
Picture；
Fig. 7 be under digestive endoscopy of the embodiment of the present invention before stomach cancer in disease stomach polyp lesion positioning figure.
Specific embodiment
Technical solution of the present invention is described in further detail with reference to the accompanying drawing, but protection scope of the present invention is not limited to
It is as described below.
As shown in Figure 1, a kind of digestive endoscopy image abnormal characteristic real-time mark system, the system are transported on the computer systems
Row, including：
Image collection module obtains the stomach Conventional white endoscopic video of input by endoscopic images system in real time
Stream rejects wherein invalid frame, screens effective key frame.Key frame is screened using the method for K mean cluster, is calculated by the cluster
Image in image library can be divided into k class, complete prior classification processing, accelerate the efficiency of key-frame extraction by method.
Image pre-processing module, for carrying out image preprocessing to image sequence to realize that image enhancement, image sequence come
From image collection module, preprocessing process includes：Image normalization, inactive pixels cutting, image smoothing, image sharpening, image
Scaling.
Image data set, for storing the image data set based on digestive endoscopy, data set includes categorized data set and inspection
Measured data collection, categorized data set include disease endoscope figure before a variety of early stage stomach cancers, and detection data collection is believed comprising lesions position
Breath, image data set is when being used to train CNN model, through but not limited to modes such as translation transformation, mirror image switch, random croppings
Data amplification is carried out, the generalization ability of CNN, the training effect of lift scheme are further enhanced.
Model training module, including disaggregated model training unit and detection model training unit, for image data set
In early stage stomach cancer before disease endoscope figure and lesions position information carry out model training, obtain detection model.
Wherein, disaggregated model training unit, for carrying out the instruction of deep learning CNN disaggregated model to the categorized data set
Practice, the parameter of network model is adjusted in the method for introducing intensified learning, including but not limited to for specific gravity in network structure
The intensified learning mode that lesser parameter carries out zero setting and iterates trained, to promote the accuracy of network class；To network
Structure optimizes, including but not limited to using layer-by-layer dimension normalization layer (Batch Normalization, BN) to network just
Then change, enhances the problem of alleviating over-fitting using Dropout method, avoids gradient disappearance problem using ReLU activation primitive；It is logical
It crosses transfer learning mode to promote the training effectiveness of disaggregated model, i.e., before train classification models, VGG-16 will be based on
Taxonomy model (is different from VGG-16 master, the classification of complete entitled VGG_ILSVRC_16_layers_fc_reduced after modification
Frame) to the extensive visual identity challenge match of ImageNet (ImageNet Large Scale Visual Recognition
Challenge, ILSVRC) model after natural image collection training as pre-training model, with accelerate the modeling of disaggregated model into
Open up and improve its classification performance.
Detection model unit, for using the CNN disaggregated model that the disaggregated model unit is trained as pre-training model,
Using the detection data collection as training set, the training of the deep learning CNN target detection model based on homing method is carried out,
Addition auxiliary network structure and detection layers on the basis of disaggregated model, the network structure of auxiliary by CNN basic unit, that is, convolution
Layer, pond layer further extract the high-dimensional feature in picture as basic unit, and the network structure of auxiliary is 8 layers total, uses
Topological expansion mode with above-mentioned disaggregated model optimizes the network structure of auxiliary, and the detection layers of bottom are by two parts
On the one hand composition introduces different proportion rectangle frame (in order to adapt to the target of different shape ratio) to image in the feature of extraction
In off-note target that may be present carry out recurrence positioning, the ratio of rectangle frame adaptive targets under the action of regression algorithm
And in being selected in its frame, these rectangle frames are corrected by training dataset during training, on the other hand in fusion
The characteristics of image extracted in the convolutional layer in base categories model and the convolutional layer of auxiliary network structure is stated, to the choosing of rectangle circle
The classification confidence of image is predicted, is corrected by training dataset to classification results during training, with described
The training pictures that detection picture library includes carry out the training of the deep learning CNN target detection model based on homing method, with
The training that iterates of network model, finally enabling detection model unit to generate can correctly classify and detect containing abnormal
The picture frame of feature, detection model unit include but is not limited to be added be based on the pyramidal detection mode of feature, low resolution,
The high-level characteristic and high-resolution of high semantic information, the low-level feature of low semantic information are attached, so that under all scales
Feature has semantic information abundant, makes to detect comprising more semantic informations more accurate, including the full articulamentum of removal mentions significantly
High calculating speed, homing method are based on single detector (Single Shot MultiBox Detector) deep learning nerve
Network frame.
Abnormality detection module, for by the image sequence for passing through image preprocessing be input to training completion detection model it
In, obtain stomach off-note classification information and location information.
Display module is marked, is integrated for classification of diseases information before the stomach cancer of output and lesion localization information, it should
Information is exported by the lesion detection module, and label display unit makes a mark location information in sequence of pictures, will be positioned
Label is mapped to original input stomach Conventional white endoscope with the formal notation targeted site of rectangle frame by coordinate information
In video flowing, and real-time display is carried out to the described stomach scope video flowing.
As shown in Fig. 2, a kind of digestive endoscopy image abnormal characteristic real-time mark method, includes the following steps：
(1) subject's stomach Conventional white endoscopic video stream is obtained using endoscopic images system equipment；
(2) key-frame extraction will be carried out after the parsing of gastroscope video flowing, rejects invalid frame, it is effective that screening obtains stomach endoscope
Sequence of video images；
As shown in figure 3, for the valid frame image of stomach under digestive endoscopy；As shown in figure 4, being rinsed for clear water under digestive endoscopy
The invalid frame image of stomach.Under digestive endoscopy before stomach cancer disease original image by medical institutions endoscopic images system equipment
Endoscope probe capture, the early carcinomatous change for being passed to area of computer aided digestive endoscopy of the present invention in real time via interface identify screening system
In, valid frame therein is extracted by the key-frame extraction unit, the content information of video expression can be by its key frame
Expression,
K-means algorithm randomly chooses k object as initial cluster center, according to minimum from a large amount of data object
Distance criterion classifies to data object, by the clustering algorithm, the image in image library is divided into k class, is completed prior
Classification processing accelerates the efficiency of key-frame extraction, and the useless frame of stomach is rinsed for clear water under digestive endoscopy, this generic operation understands certain
Gastroscope picture quality is influenced in kind degree, if the calculation amount that will be much less is rejected or ignored to this kind of image.
(3) image preprocessing is carried out to gastroscope effective video image sequence.As shown in figure 5, for stomach cancer under digestive endoscopy
The original image of stomach polyp in preceding disease.Original image by image preprocessing polyp image have passed through image normalization,
Useless part letter is eliminated after inactive pixels cutting, image smoothing, image sharpening, the pretreatment of image scaling a series of images
Breath, the polyp image after image enhancement processing can obtain more accurately classifying and locating effect, image be scaled to
The size (300*300) that the model inspection lesion unit input image size matches, as shown in Figure 6.
(4) image sequence after image preprocessing is input in detection model, detection model is by disaggregated model plus auxiliary
Network structure and detection layers is helped to form, disaggregated model is based on VGG-16 frame, and auxiliary network structure further extracts picture feature,
Detection layers obtain the position of classification results and regressive object, are added and are based on the pyramidal detection mode of feature, low resolution, height
The high-level characteristic and high-resolution of semantic information, the low-level feature of low semantic information are attached.Detection model exports suspicious stomach
Classification of diseases result and classification confidence before portion's cancer, while exporting lesion localization coordinate and positioning confidence level；
(5) system marks in effective image sequence according to lesion localization coordinate, marks as shown in fig. 7, with eye-catching
Rectangle frame label the polyp lesion in image is accurately positioned, while label being mapped in former gastroscope video flowing.
(6) real-time display diagnoses labeled gastroscope video flowing over the display so that doctor observes confirmation.Display by
Frame shows the classification of gastroscope image sequence, detection information.
Claims (8)
1. a kind of digestive endoscopy image abnormal characteristic real-time mark system, including computer system, which is characterized in that the calculating
Machine system includes：
Image collection module is obtained the stomach Conventional white endoscopic video stream of input in real time by endoscopic images system, picked
Except wherein invalid frame, effective key frame is screened；
Image pre-processing module, for carrying out image enhancement processing to the key frame filtered out；
Model training module is obtained for carrying out model training to disease endoscope figure before early stage stomach cancer and lesions position information
To detection model；
Abnormality detection module, for the image sequence for passing through image preprocessing to be input among the detection model of training completion,
Obtain the classification information and location information of stomach off-note；
Display module is marked, location information is made a mark in sequence of pictures, and label is mapped to original input stomach
In Conventional white endoscopic video stream, and real-time display is carried out to the described stomach scope video flowing.
2. digestive endoscopy image abnormal characteristic real-time mark system according to claim 1, which is characterized in that the figure
K object is randomly choosed from data object using K-means algorithm when screening effective key frame as just as obtaining module
Beginning cluster centre classifies to data object according to minimum distance criterion, by the clustering algorithm, the image in image library
It is divided into k class.
3. digestive endoscopy image abnormal characteristic real-time mark system according to claim 1, which is characterized in that described image
Image enhancement processing includes in preprocessing module：Image normalization, inactive pixels cutting, image smoothing, image sharpening and image
Scaling.
4. digestive endoscopy image abnormal characteristic real-time mark system according to claim 1, which is characterized in that the mould
Type training module includes：
Image data set, for storing categorized data set and detection data collection based on digestive endoscopy；Categorized data set includes more
Disease endoscope figure before kind early stage stomach cancer, detection data collection include lesions position information；
Disaggregated model training unit, for carrying out the training of deep learning CNN disaggregated model to the categorized data set；
Detection model training unit is used for using CNN disaggregated model as pre-training model, using the detection data collection as training
Collection, carries out the training of the deep learning CNN target detection model based on homing method.
5. digestive endoscopy image abnormal characteristic real-time mark system according to claim 4, which is characterized in that described image
Data set is for carrying out data amplification by translation transformation, mirror image switch or random cropping before training CNN model.
6. digestive endoscopy image abnormal characteristic real-time mark system according to claim 4, which is characterized in that the classification
Model training unit introduces the method for intensified learning when carrying out deep learning CNN disaggregated model training to improve network class
Accuracy the problem of over-fitting is alleviated using Dropout method, adopted using layer-by-layer dimension normalization layer to network regularization
Gradient disappearance problem is avoided with ReLU activation primitive, is promoted by training effectiveness of the transfer learning mode to disaggregated model.
7. digestive endoscopy image abnormal characteristic real-time mark system according to claim 4, which is characterized in that the model
The detection model obtained after training module training includes disaggregated model, auxiliary network structure and target detection model；The classification
Model is based on VGG-16 frame, for obtaining classification information；The auxiliary network structure is for extracting picture feature；The target
Detection model is for obtaining location information.
8. a kind of side for carrying out digestive endoscopy image abnormal characteristic real-time mark using any one of the claim 1-7 system
Method, which is characterized in that include the following steps：
(1) disease endoscope figure before early stage stomach cancer and lesions position information input model training module are trained, are obtained
Detection model；
(2) subject's stomach Conventional white endoscopic video stream is obtained using endoscopic images system equipment；
(3) key-frame extraction will be carried out after the parsing of gastroscope video flowing, rejects invalid frame, screening obtains stomach endoscope effective video
Image sequence；
(4) effective video image sequence is input to image pre-processing module, carries out image preprocessing；
(5) image sequence after image preprocessing is input in detection model, detection model exports disease before suspicious stomach cancer
Lesion target positions coordinate, while exporting in positioning coordinate classification of diseases result and classification confidence before suspicious stomach cancer；
(6) it is marked in effective image sequence according to lesion localization coordinate, while label is mapped to former gastroscope video flowing
In；
(7) real-time display diagnoses labeled gastroscope video flowing over the display so that doctor observes confirmation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810365426.5A CN108852268A (en) | 2018-04-23 | 2018-04-23 | A kind of digestive endoscopy image abnormal characteristic real-time mark system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810365426.5A CN108852268A (en) | 2018-04-23 | 2018-04-23 | A kind of digestive endoscopy image abnormal characteristic real-time mark system and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108852268A true CN108852268A (en) | 2018-11-23 |
Family
ID=64327123
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810365426.5A Pending CN108852268A (en) | 2018-04-23 | 2018-04-23 | A kind of digestive endoscopy image abnormal characteristic real-time mark system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108852268A (en) |
Cited By (32)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109754389A (en) * | 2018-12-07 | 2019-05-14 | 北京市商汤科技开发有限公司 | A kind of lesion detection method, device and equipment |
| CN109919319A (en) * | 2018-12-31 | 2019-06-21 | 中国科学院软件研究所 | Deeply learning method and equipment based on multiple history best Q networks |
| CN110097080A (en) * | 2019-03-29 | 2019-08-06 | 广州思德医疗科技有限公司 | A kind of construction method and device of tag along sort |
| CN110232408A (en) * | 2019-05-30 | 2019-09-13 | 清华-伯克利深圳学院筹备办公室 | A kind of endoscopic image processing method and relevant device |
| CN110363768A (en) * | 2019-08-30 | 2019-10-22 | 重庆大学附属肿瘤医院 | A kind of early carcinoma lesion horizon prediction auxiliary system based on deep learning |
| CN110491502A (en) * | 2019-03-08 | 2019-11-22 | 腾讯科技（深圳）有限公司 | Microscope video stream processing method, system, computer equipment and storage medium |
| CN110477846A (en) * | 2019-08-23 | 2019-11-22 | 西安交通大学医学院第二附属医院 | A kind of digestive endoscopy off-note real-time mark method and system |
| CN110867233A (en) * | 2019-11-19 | 2020-03-06 | 西安邮电大学 | System and method for generating electronic laryngoscope medical test reports |
| CN111064934A (en) * | 2019-12-30 | 2020-04-24 | 元力(天津)科技有限公司 | Medical image processing system and method |
| CN111127412A (en) * | 2019-12-18 | 2020-05-08 | 广东工业大学 | Pathological image recognition device based on generation countermeasure network |
| CN111242921A (en) * | 2020-01-12 | 2020-06-05 | 刘涛 | Method and system for automatically updating medical ultrasonic image auxiliary diagnosis system |
| CN111275041A (en) * | 2020-01-20 | 2020-06-12 | 腾讯科技（深圳）有限公司 | Endoscope image display method and device, computer equipment and storage medium |
| CN111462082A (en) * | 2020-03-31 | 2020-07-28 | 重庆金山医疗技术研究院有限公司 | Focus picture recognition device, method and equipment and readable storage medium |
| CN111565303A (en) * | 2020-05-29 | 2020-08-21 | 深圳市易链信息技术有限公司 | Video monitoring method, system and readable storage medium based on fog calculation and deep learning |
| CN111899229A (en) * | 2020-07-14 | 2020-11-06 | 武汉楚精灵医疗科技有限公司 | Advanced gastric cancer auxiliary diagnosis method based on deep learning multi-model fusion technology |
| CN111915573A (en) * | 2020-07-14 | 2020-11-10 | 武汉楚精灵医疗科技有限公司 | Digestive endoscopy focus tracking method based on time sequence feature learning |
| CN111932507A (en) * | 2020-07-31 | 2020-11-13 | 苏州慧维智能医疗科技有限公司 | Method for identifying lesion in real time based on digestive endoscopy |
| CN112132772A (en) * | 2019-06-24 | 2020-12-25 | 杭州迪英加科技有限公司 | Pathological section real-time interpretation method, device and system |
| CN112651375A (en) * | 2021-01-05 | 2021-04-13 | 中国人民解放军陆军特色医学中心 | Helicobacter pylori stomach image recognition and classification system based on deep learning model |
| CN112801958A (en) * | 2021-01-18 | 2021-05-14 | 青岛大学附属医院 | Ultrasonic endoscope, artificial intelligence auxiliary identification method, system, terminal and medium |
| CN112837275A (en) * | 2021-01-14 | 2021-05-25 | 长春大学 | Capsule endoscope image organ classification method, device, equipment and storage medium |
| CN112949666A (en) * | 2019-11-26 | 2021-06-11 | 北矿机电科技有限责任公司 | Image identification method for fault cathode plate of zinc stripping machine |
| CN113052116A (en) * | 2021-04-06 | 2021-06-29 | 深圳华声医疗技术股份有限公司 | Ultrasonic video data processing method and device, ultrasonic equipment and storage medium |
| WO2021147500A1 (en) * | 2020-01-23 | 2021-07-29 | 腾讯科技（深圳）有限公司 | Endoscope image processing method and apparatus, and electronic device and storage medium |
| CN113256572A (en) * | 2021-05-12 | 2021-08-13 | 中国科学院自动化研究所 | Gastroscope image analysis system, method and equipment based on restoration and selective enhancement |
| CN113268623A (en) * | 2021-06-01 | 2021-08-17 | 上海市第一人民医院 | Artificial intelligence gastroscope image recognition processing system |
| CN113473899A (en) * | 2019-02-26 | 2021-10-01 | 奥林巴斯株式会社 | Endoscope device and program |
| CN114332025A (en) * | 2021-12-29 | 2022-04-12 | 长沙慧维智能医疗科技有限公司 | Digestive endoscopy oropharynx passing time automatic detection system and method |
| CN114464316A (en) * | 2022-04-11 | 2022-05-10 | 武汉大学 | Stomach abnormal risk grade prediction method, device, terminal and readable storage medium |
| CN114533148A (en) * | 2022-02-15 | 2022-05-27 | 佳木斯大学 | Sampling system for stomach cancer detection in digestive system department |
| CN115187591A (en) * | 2022-09-08 | 2022-10-14 | 深圳市资福医疗技术有限公司 | Focus detection method, system, equipment and storage medium |
| CN116473673A (en) * | 2023-06-20 | 2023-07-25 | 浙江华诺康科技有限公司 | Path planning method, device, system and storage medium for endoscope |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101584571A (en) * | 2009-06-15 | 2009-11-25 | 无锡骏聿科技有限公司 | Capsule endoscopy auxiliary film reading method |
| CN107368670A (en) * | 2017-06-07 | 2017-11-21 | 万香波 | Stomach cancer pathology diagnostic support system and method based on big data deep learning |
| CN107564580A (en) * | 2017-09-11 | 2018-01-09 | 合肥工业大学 | Gastroscope visual aids processing system and method based on integrated study |
| CN107730489A (en) * | 2017-10-09 | 2018-02-23 | 杭州电子科技大学 | Wireless capsule endoscope small intestine disease variant computer assisted detection system and detection method |
| CN107832802A (en) * | 2017-11-23 | 2018-03-23 | 北京智芯原动科技有限公司 | Quality of human face image evaluation method and device based on face alignment |
| CN107832838A (en) * | 2017-11-28 | 2018-03-23 | 北京羽医甘蓝信息技术有限公司 | The method and apparatus for evaluating cell smear sample satisfaction |
| CN107862241A (en) * | 2017-06-06 | 2018-03-30 | 哈尔滨工业大学深圳研究生院 | A kind of clothes fashion method for digging and visually-perceptible system based on star's identification |
| CN107895367A (en) * | 2017-11-14 | 2018-04-10 | 中国科学院深圳先进技术研究院 | A kind of stone age recognition methods, system and electronic equipment |
-
2018
- 2018-04-23 CN CN201810365426.5A patent/CN108852268A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101584571A (en) * | 2009-06-15 | 2009-11-25 | 无锡骏聿科技有限公司 | Capsule endoscopy auxiliary film reading method |
| CN107862241A (en) * | 2017-06-06 | 2018-03-30 | 哈尔滨工业大学深圳研究生院 | A kind of clothes fashion method for digging and visually-perceptible system based on star's identification |
| CN107368670A (en) * | 2017-06-07 | 2017-11-21 | 万香波 | Stomach cancer pathology diagnostic support system and method based on big data deep learning |
| CN107564580A (en) * | 2017-09-11 | 2018-01-09 | 合肥工业大学 | Gastroscope visual aids processing system and method based on integrated study |
| CN107730489A (en) * | 2017-10-09 | 2018-02-23 | 杭州电子科技大学 | Wireless capsule endoscope small intestine disease variant computer assisted detection system and detection method |
| CN107895367A (en) * | 2017-11-14 | 2018-04-10 | 中国科学院深圳先进技术研究院 | A kind of stone age recognition methods, system and electronic equipment |
| CN107832802A (en) * | 2017-11-23 | 2018-03-23 | 北京智芯原动科技有限公司 | Quality of human face image evaluation method and device based on face alignment |
| CN107832838A (en) * | 2017-11-28 | 2018-03-23 | 北京羽医甘蓝信息技术有限公司 | The method and apparatus for evaluating cell smear sample satisfaction |
Cited By (46)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109754389A (en) * | 2018-12-07 | 2019-05-14 | 北京市商汤科技开发有限公司 | A kind of lesion detection method, device and equipment |
| CN109754389B (en) * | 2018-12-07 | 2021-08-24 | 北京市商汤科技开发有限公司 | Image processing method, device and equipment |
| TWI724669B (en) * | 2018-12-07 | 2021-04-11 | 大陸商北京市商湯科技開發有限公司 | Lesion detection method and device, equipment and storage medium |
| CN109919319A (en) * | 2018-12-31 | 2019-06-21 | 中国科学院软件研究所 | Deeply learning method and equipment based on multiple history best Q networks |
| CN113473899A (en) * | 2019-02-26 | 2021-10-01 | 奥林巴斯株式会社 | Endoscope device and program |
| CN110491502A (en) * | 2019-03-08 | 2019-11-22 | 腾讯科技（深圳）有限公司 | Microscope video stream processing method, system, computer equipment and storage medium |
| CN110097080A (en) * | 2019-03-29 | 2019-08-06 | 广州思德医疗科技有限公司 | A kind of construction method and device of tag along sort |
| CN110097080B (en) * | 2019-03-29 | 2021-04-13 | 广州思德医疗科技有限公司 | Construction method and device of classification label |
| CN110232408A (en) * | 2019-05-30 | 2019-09-13 | 清华-伯克利深圳学院筹备办公室 | A kind of endoscopic image processing method and relevant device |
| CN110232408B (en) * | 2019-05-30 | 2021-09-10 | 清华-伯克利深圳学院筹备办公室 | Endoscope image processing method and related equipment |
| CN112132772A (en) * | 2019-06-24 | 2020-12-25 | 杭州迪英加科技有限公司 | Pathological section real-time interpretation method, device and system |
| CN112132772B (en) * | 2019-06-24 | 2024-02-23 | 杭州迪英加科技有限公司 | Pathological section real-time interpretation method, device and system |
| CN110477846A (en) * | 2019-08-23 | 2019-11-22 | 西安交通大学医学院第二附属医院 | A kind of digestive endoscopy off-note real-time mark method and system |
| CN110363768B (en) * | 2019-08-30 | 2021-08-17 | 重庆大学附属肿瘤医院 | Early cancer focus range prediction auxiliary system based on deep learning |
| CN110363768A (en) * | 2019-08-30 | 2019-10-22 | 重庆大学附属肿瘤医院 | A kind of early carcinoma lesion horizon prediction auxiliary system based on deep learning |
| CN110867233A (en) * | 2019-11-19 | 2020-03-06 | 西安邮电大学 | System and method for generating electronic laryngoscope medical test reports |
| CN112949666B (en) * | 2019-11-26 | 2024-04-26 | 北矿机电科技有限责任公司 | Image recognition method for fault cathode plate of zinc stripping machine |
| CN112949666A (en) * | 2019-11-26 | 2021-06-11 | 北矿机电科技有限责任公司 | Image identification method for fault cathode plate of zinc stripping machine |
| CN111127412A (en) * | 2019-12-18 | 2020-05-08 | 广东工业大学 | Pathological image recognition device based on generation countermeasure network |
| CN111127412B (en) * | 2019-12-18 | 2021-05-28 | 广东工业大学 | Pathological image recognition device based on generation countermeasure network |
| CN111064934A (en) * | 2019-12-30 | 2020-04-24 | 元力(天津)科技有限公司 | Medical image processing system and method |
| CN111242921A (en) * | 2020-01-12 | 2020-06-05 | 刘涛 | Method and system for automatically updating medical ultrasonic image auxiliary diagnosis system |
| CN111242921B (en) * | 2020-01-12 | 2023-12-12 | 刘涛 | Automatic updating method and system for medical ultrasonic image auxiliary diagnosis system |
| CN111275041A (en) * | 2020-01-20 | 2020-06-12 | 腾讯科技（深圳）有限公司 | Endoscope image display method and device, computer equipment and storage medium |
| WO2021147500A1 (en) * | 2020-01-23 | 2021-07-29 | 腾讯科技（深圳）有限公司 | Endoscope image processing method and apparatus, and electronic device and storage medium |
| CN111462082A (en) * | 2020-03-31 | 2020-07-28 | 重庆金山医疗技术研究院有限公司 | Focus picture recognition device, method and equipment and readable storage medium |
| CN111565303A (en) * | 2020-05-29 | 2020-08-21 | 深圳市易链信息技术有限公司 | Video monitoring method, system and readable storage medium based on fog calculation and deep learning |
| CN111899229A (en) * | 2020-07-14 | 2020-11-06 | 武汉楚精灵医疗科技有限公司 | Advanced gastric cancer auxiliary diagnosis method based on deep learning multi-model fusion technology |
| CN111915573A (en) * | 2020-07-14 | 2020-11-10 | 武汉楚精灵医疗科技有限公司 | Digestive endoscopy focus tracking method based on time sequence feature learning |
| CN111932507A (en) * | 2020-07-31 | 2020-11-13 | 苏州慧维智能医疗科技有限公司 | Method for identifying lesion in real time based on digestive endoscopy |
| CN112651375A (en) * | 2021-01-05 | 2021-04-13 | 中国人民解放军陆军特色医学中心 | Helicobacter pylori stomach image recognition and classification system based on deep learning model |
| CN112837275A (en) * | 2021-01-14 | 2021-05-25 | 长春大学 | Capsule endoscope image organ classification method, device, equipment and storage medium |
| CN112837275B (en) * | 2021-01-14 | 2023-10-24 | 长春大学 | Capsule endoscope image organ classification method, device, equipment and storage medium |
| CN112801958A (en) * | 2021-01-18 | 2021-05-14 | 青岛大学附属医院 | Ultrasonic endoscope, artificial intelligence auxiliary identification method, system, terminal and medium |
| CN113052116B (en) * | 2021-04-06 | 2022-02-22 | 深圳华声医疗技术股份有限公司 | Ultrasonic video data processing method and device, ultrasonic equipment and storage medium |
| CN113052116A (en) * | 2021-04-06 | 2021-06-29 | 深圳华声医疗技术股份有限公司 | Ultrasonic video data processing method and device, ultrasonic equipment and storage medium |
| CN113256572A (en) * | 2021-05-12 | 2021-08-13 | 中国科学院自动化研究所 | Gastroscope image analysis system, method and equipment based on restoration and selective enhancement |
| CN113256572B (en) * | 2021-05-12 | 2023-04-07 | 中国科学院自动化研究所 | Gastroscope image analysis system, method and equipment based on restoration and selective enhancement |
| CN113268623A (en) * | 2021-06-01 | 2021-08-17 | 上海市第一人民医院 | Artificial intelligence gastroscope image recognition processing system |
| CN113268623B (en) * | 2021-06-01 | 2022-07-19 | 上海市第一人民医院 | Artificial intelligence gastroscope image identification processing system |
| CN114332025A (en) * | 2021-12-29 | 2022-04-12 | 长沙慧维智能医疗科技有限公司 | Digestive endoscopy oropharynx passing time automatic detection system and method |
| CN114533148A (en) * | 2022-02-15 | 2022-05-27 | 佳木斯大学 | Sampling system for stomach cancer detection in digestive system department |
| CN114464316A (en) * | 2022-04-11 | 2022-05-10 | 武汉大学 | Stomach abnormal risk grade prediction method, device, terminal and readable storage medium |
| CN115187591A (en) * | 2022-09-08 | 2022-10-14 | 深圳市资福医疗技术有限公司 | Focus detection method, system, equipment and storage medium |
| CN116473673A (en) * | 2023-06-20 | 2023-07-25 | 浙江华诺康科技有限公司 | Path planning method, device, system and storage medium for endoscope |
| CN116473673B (en) * | 2023-06-20 | 2024-02-27 | 浙江华诺康科技有限公司 | Path planning method, device, system and storage medium for endoscope |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108852268A (en) | A kind of digestive endoscopy image abnormal characteristic real-time mark system and method | |
| Sun et al. | Computer-aided diagnosis in histopathological images of the endometrium using a convolutional neural network and attention mechanisms | |
| Jiang et al. | An automatic detection system of lung nodule based on multigroup patch-based deep learning network | |
| CN109858540B (en) | Medical image recognition system and method based on multi-mode fusion | |
| CN109727243A (en) | Breast ultrasound image recognition analysis method and system | |
| CN109191472A (en) | Based on the thymocyte image partition method for improving U-Net network | |
| CN101103924A (en) | Galactophore cancer computer auxiliary diagnosis method based on galactophore X-ray radiography and system thereof | |
| Chen et al. | Discriminative cervical lesion detection in colposcopic images with global class activation and local bin excitation | |
| Zhang et al. | Design of automatic lung nodule detection system based on multi-scene deep learning framework | |
| Sornapudi et al. | Automated cervical digitized histology whole-slide image analysis toolbox | |
| Yar et al. | Lung nodule detection and classification using 2D and 3D convolution neural networks (CNNs) | |
| Aslam et al. | Liver-tumor detection using CNN ResUNet | |
| Guo et al. | Coarse-to-fine airway segmentation using multi information fusion network and CNN-based region growing | |
| CN113539476A (en) | Stomach endoscopic biopsy Raman image auxiliary diagnosis method and system based on artificial intelligence | |
| Liu et al. | Weakly-supervised localization and classification of biomarkers in OCT images with integrated reconstruction and attention | |
| Xue et al. | Automatic detection of oral lesion measurement ruler toward computer-aided image-based oral cancer screening | |
| Luan et al. | R-FCN based laryngeal lesion detection | |
| Diao et al. | Computer aided cancer regions detection of hepatocellular carcinoma in whole-slide pathological images based on deep learning | |
| Vats et al. | Detecting oral cancer: the potential of artificial intelligence | |
| Xu et al. | A high-precision classification method of mammary cancer based on improved DenseNet driven by an attention mechanism | |
| Shao et al. | A segmentation method of airway from chest ct image based on vgg-unet neural network | |
| Guo et al. | Lightweight yolov4 with multiple receptive fields for detection of pulmonary tuberculosis | |
| Jiang et al. | Classifying Cervical Histopathological Whole Slide Images via Deep Multi-Instance Transfer Learning | |
| Xiong et al. | Deep Ensemble Learning Network for Kidney Lesion Detection | |
| TW202137233A (en) | Image-based identification method and system |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181123 |