CN107610773A - A kind of vascular dissection aided diagnosis method based on sustainer medical image - Google Patents
A kind of vascular dissection aided diagnosis method based on sustainer medical image Download PDFInfo
- Publication number
- CN107610773A CN107610773A CN201710816605.1A CN201710816605A CN107610773A CN 107610773 A CN107610773 A CN 107610773A CN 201710816605 A CN201710816605 A CN 201710816605A CN 107610773 A CN107610773 A CN 107610773A
- Authority
- CN
- China
- Prior art keywords
- medical image
- interested
- sustainer
- image
- roi
- 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
Landscapes
- Apparatus For Radiation Diagnosis (AREA)
- Magnetic Resonance Imaging Apparatus (AREA)
Abstract
The invention discloses a kind of vascular dissection aided diagnosis method based on sustainer medical image, comprise the following steps：S1：Extract the medical image information in primitive aorta region；S2：ROI detections interested are carried out to the medical image information in the primitive aorta region of extraction using deep learning theory, determine ROI region interested；S3：Enter the Classification and Identification of row aorta diagnosis to ROI region interested using deep learning theory, judge whether ROI region interested is suspicious lesions；S4：If 100% without suspicious lesions in image, it is normal images by the scope interpretation of ROI region interested, whether is focus image by doctor's decision-making if suspected abnormality in image be present, complete auxiliary diagnosis.It vascular dissection aided diagnosis method proposed by the present invention, can significantly reduce the diagnostic work amount of doctor, improve diagnosis efficiency, while doctor can be made to focus more on the diagnosis of suspicious lesions, reduce the phenomenon generation of Misdiagnosis.
Description
Technical field
The present invention relates to medical imaging diagnostic techniques field, more particularly to a kind of blood vessel clip based on sustainer medical image
Layer aided diagnosis method.
Background technology
Normal human body artery blood vessel is made up of inner membrance, middle film and outer membrane, and 3-tier architecture carries passing through for blood flow jointly.It is main
Artery dissection occurs, in master pulse endangium, to be influenceed by some external causes or internal cause, and inner membrance is progressively peeled off with other membrane structures,
And chamber forms true and false two chamber in the blood vessels, so as to form interlayer.
The maximum harm of dissection of aorta is dead.Sustainer is the major blood vessel of body, bears to jump directly from heart
Dynamic pressure, CBF is huge, theca interna tear occurs, if without appropriately and timely treating, the chance of rupture is very
Greatly, the death rate is also very high.Conventional reported literature, the death rate in 1 week are up to 50%, and the death rate in one month is in 60-
Between 70%.In addition, even if patient is survived, because of the expansion of false chamber and the increase of pressure, the CBF of true chamber blood vessel
Reduce, then can cause the internal organs ischemic in sustainer institute's blood supply region.
Making a definite diagnosis the main auxiliary examination methods of dissection of aorta is：CT angiograms (CTA), magnetic resonance examination (MRA) or
It is direct digital subtraction angiography (DSA).No matter which kind of imaging mode, be required to carry out patient three-dimensional scanning, doctor
It is raw to need to check interlayer lesion information from hundreds of tension fault images, because data volume is huge, have a strong impact on interlayer diagnosis
Efficiency, further, since the interlayer focus of early stage is not easy to observe, doctor also occurs that mistaken diagnosis in checking process, failed to pinpoint a disease in diagnosis, and causes
Malpractice.
The content of the invention
The invention aims to solve shortcoming present in prior art, and the one kind proposed is based on sustainer medical science
The vascular dissection aided diagnosis method of image.
A kind of vascular dissection aided diagnosis method based on sustainer medical image, comprises the following steps：
S1：Original medical image information is obtained by medical image instrument, then is carried in all medical image informations
Take the medical image information in primitive aorta region；
S2：ROI interested is carried out to the medical image information in the primitive aorta region of extraction using deep learning theory
Detection, normal medical image is filtered out, retain the aorta regions with prediction result strong correlation, be defined as ROI areas interested
Domain；
S3：The classification for being entered row aorta diagnosis to the ROI region interested that S2 steps determine using deep learning theory is known
Not, judge whether ROI region interested is suspicious lesions；
S4：If the result that S3 steps judge as in image 100% without suspicious lesions, by the image of ROI region interested
It is determined as normal images, and result of determination is shown by instrument, if the result that S3 steps judge is doubted to exist in image
Like focus, then image information is sent to doctor, whether is focus image by doctor's decision-making, complete vascular dissection auxiliary diagnosis.
Preferably, the method for the ROI detections interested is carried out using convolutional neural networks method.
Preferably, the concrete operations of the convolutional neural networks method are：It is tentative first on image to carry out ROI interested
Region selects frame to operate, and convolution and the operation scaled are constantly carried out by model parameter to the subgraph of inframe, recently enters
The characteristic vector data of one 1024 dimension, by the learning strategy of supervised, this feature data can feed back one and belong to sustainer
The confidence value in region, when the confidence value is more than certain limit, then it is determined as that sustainer ROI region positioning interested is accurate
Really, the then convolution operation processing of input picture, is finally utilized respectively known models to the characteristics of image of input according to multichannel
Parameter carry out process of convolution.
Preferably, the method for Classification and Identification is two classification in the S3 steps, and the concrete operations of two classification are：
By ROI region interested according to having interlayer or made a distinction without interlayer.
Vascular dissection aided diagnosis method proposed by the present invention is pressed from both sides using the technical assistance doctor of artificial intelligence field
The investigation of layer data and screening, by artificial intelligence technology before formal inspection, doctor is helped to filter out most of normal doctor
Image is learned, only retains the medical image with suspicious lesions, these final suspicious lesions images can be made final face by doctor
Bed diagnosis decision-making, in this way, can significantly reduce the diagnostic work amount of doctor, improve diagnosis efficiency, while can make
Doctor focuses more on the diagnosis of suspicious lesions, occurs so as to reduce the phenomenon of Misdiagnosis to a certain extent.
Brief description of the drawings
Fig. 1 is a kind of flow of the vascular dissection aided diagnosis method based on sustainer medical image proposed by the present invention
Figure；
Fig. 2 is sustainer in a kind of vascular dissection aided diagnosis method based on sustainer medical image proposed by the present invention
ROI region interested detection network；
Fig. 3 is using volume in a kind of vascular dissection aided diagnosis method based on sustainer medical image proposed by the present invention
Interested ROI region testing result of the product neural network to sustainer.
Embodiment
The present invention is made with reference to specific embodiment further to explain.
Embodiment
A kind of vascular dissection aided diagnosis method based on sustainer medical image proposed by the present invention, including following step
Suddenly：
S1：Original medical image information is obtained by medical image instrument, then is carried in all medical image informations
Take the medical image information in primitive aorta region；
S2：ROI interested is carried out to the medical image information in the primitive aorta region of extraction using deep learning theory
Detection, normal medical image is filtered out, retain the aorta regions with prediction result strong correlation, be defined as ROI areas interested
Domain, the method for the ROI detections interested use convolutional neural networks method, and concrete operations are：First on image it is tentative enter
Row ROI region interested selects frame to operate, and convolution and the behaviour scaled are constantly carried out by model parameter to the subgraph of inframe
Make, recently enter the characteristic vector data of one 1024 dimension, by the learning strategy of supervised, this feature data can feed back one
Belong to the confidence value of aorta regions, when the confidence value is more than certain limit, be then determined as sustainer ROI areas interested
The convolution operation processing of domain accurate positioning, then input picture, is finally utilized respectively to the characteristics of image of input according to multichannel
The parameter of known models carries out process of convolution, completes ROI detections interested；
S3：The classification for being entered row aorta diagnosis to the ROI region interested that S2 steps determine using deep learning theory is known
Not, the classification of sorting technique two, and ROI region interested is judged to feel emerging according to having interlayer or making a distinction without interlayer
Whether interesting ROI region is suspicious lesions；
S4：If the result that S3 steps judge as in image 100% without suspicious lesions, by the image of ROI region interested
It is determined as normal images, and result of determination is shown by instrument, if the result that S3 steps judge is doubted to exist in image
Like focus, then image information is sent to doctor, whether is focus image by doctor's decision-making, complete vascular dissection auxiliary diagnosis.
Vascular dissection aided diagnosis method proposed by the present invention is pressed from both sides using the technical assistance doctor of artificial intelligence field
The investigation of layer data and screening, the relevant feature detected according to sustainer, ROI detections are formulated and have been combined with ROI region identification
Technology path, most of normal medical image in influence is filtered out, only retained with can be supplied with the medical image of focus
Diagnosis, significantly reduce the diagnostic work amount of doctor, improve diagnosis efficiency, while doctor can be made to focus more on suspicious lesions
Diagnosis, so as to a certain extent reduce Misdiagnosis phenomenon occur, and vascular dissection proposed by the present invention auxiliary examine
The theoretical correlation technique of deep learning is incorporated into medical imaging diagnostic field by disconnected method, is provided for dissection of aorta detection new
Method.
The foregoing is only a preferred embodiment of the present invention, but protection scope of the present invention be not limited thereto,
Any one skilled in the art the invention discloses technical scope in, technique according to the invention scheme and its
Inventive concept is subject to equivalent substitution or change, should all be included within the scope of the present invention.
Claims (4)
1. a kind of vascular dissection aided diagnosis method based on sustainer medical image, it is characterised in that comprise the following steps：
S1：Original medical image information is obtained by medical image instrument, then original is extracted in all medical image informations
Beginning aorta regions medical image information；
S2：ROI detections interested are carried out to the medical image information in the primitive aorta region of extraction using deep learning theory,
Normal medical image is filtered out, retains the aorta regions with prediction result strong correlation, is defined as ROI region interested；
S3：Enter the Classification and Identification of row aorta diagnosis to the ROI region interested that S2 steps determine using deep learning theory, sentence
Whether ROI region interested of breaking is suspicious lesions；
S4：If the result that S3 steps judge as in image 100% without suspicious lesions, by the scope interpretation of ROI region interested
For normal images, and result of determination is shown by instrument, if the result that judges of S3 steps is has doubtful disease in image
Stove, then image information is sent to doctor, whether is focus image by doctor's decision-making, complete vascular dissection auxiliary diagnosis.
2. a kind of vascular dissection aided diagnosis method based on sustainer medical image according to claim 1, its feature
It is, the method for the ROI detections interested is carried out using convolutional neural networks method.
3. a kind of vascular dissection aided diagnosis method based on sustainer medical image according to claim 2, its feature
It is, the concrete operations of the convolutional neural networks method are：Tentative progress ROI region interested selects frame first on image
Operation, and convolution and the operation scaled are constantly carried out by model parameter to the subgraph of inframe, recently enter one 1024 dimension
Characteristic vector data, by the learning strategy of supervised, this feature data can feed back one and belong to the credible of aorta regions
Angle value, when the confidence value is more than certain limit, then it is determined as sustainer ROI region accurate positioning interested, then inputs
The convolution operation processing of image, the parameter for being finally utilized respectively known models according to multichannel to the characteristics of image of input are rolled up
Product processing.
4. a kind of vascular dissection aided diagnosis method based on sustainer medical image according to claim 1, its feature
It is, the method for Classification and Identification is two classification in the S3 steps, and the concrete operations of two classification are：Will be interested
ROI region is according to having interlayer or made a distinction without interlayer.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710816605.1A CN107610773A (en) | 2017-09-12 | 2017-09-12 | A kind of vascular dissection aided diagnosis method based on sustainer medical image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710816605.1A CN107610773A (en) | 2017-09-12 | 2017-09-12 | A kind of vascular dissection aided diagnosis method based on sustainer medical image |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107610773A true CN107610773A (en) | 2018-01-19 |
Family
ID=61063722
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710816605.1A Pending CN107610773A (en) | 2017-09-12 | 2017-09-12 | A kind of vascular dissection aided diagnosis method based on sustainer medical image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107610773A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109584229A (en) * | 2018-11-28 | 2019-04-05 | 武汉大学人民医院（湖北省人民医院） | A kind of real-time assistant diagnosis system of Endoscopic retrograde cholangio-pancreatiography art and method |
| CN110264465A (en) * | 2019-06-25 | 2019-09-20 | 中南林业科技大学 | A kind of dissection of aorta dynamic testing method based on morphology and deep learning |
| CN110348500A (en) * | 2019-06-30 | 2019-10-18 | 浙江大学 | Sleep disturbance aided diagnosis method based on deep learning and infrared thermal imagery |
| CN110706815A (en) * | 2019-11-26 | 2020-01-17 | 北京推想科技有限公司 | Evaluation method and device of image report and electronic equipment |
| CN111754485A (en) * | 2020-06-24 | 2020-10-09 | 成都市温江区人民医院 | Artificial intelligence ultrasonic auxiliary system for liver |
| CN113223704A (en) * | 2021-05-20 | 2021-08-06 | 吉林大学 | Auxiliary diagnosis method for computed tomography aortic aneurysm based on deep learning |
| CN113763337A (en) * | 2021-08-24 | 2021-12-07 | 慧影医疗科技（北京）有限公司 | Method and system for detecting blood supply of aortic dissection false cavity |
| CN115131508A (en) * | 2022-08-30 | 2022-09-30 | 南通佳茂霖智能科技有限公司 | DSA modeling point cloud data fusion processing method based on data processing |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102332162A (en) * | 2011-09-19 | 2012-01-25 | 西安百利信息科技有限公司 | Method for automatic recognition and stage compression of medical image regions of interest based on artificial neural network |
| US20150238148A1 (en) * | 2013-10-17 | 2015-08-27 | Siemens Aktiengesellschaft | Method and system for anatomical object detection using marginal space deep neural networks |
| CN105640577A (en) * | 2015-12-16 | 2016-06-08 | 深圳市智影医疗科技有限公司 | Method and system automatically detecting local lesion in radiographic image |
| CN106372390A (en) * | 2016-08-25 | 2017-02-01 | 姹ゅ钩 | Deep convolutional neural network-based lung cancer preventing self-service health cloud service system |
| CN106780460A (en) * | 2016-12-13 | 2017-05-31 | 杭州健培科技有限公司 | A kind of Lung neoplasm automatic checkout system for chest CT image |
| CN106897573A (en) * | 2016-08-01 | 2017-06-27 | 12西格玛控股有限公司 | Use the computer-aided diagnosis system for medical image of depth convolutional neural networks |
| CN107115111A (en) * | 2017-01-23 | 2017-09-01 | 上海联影医疗科技有限公司 | Blood flow state analysis system and method |
-
2017
- 2017-09-12 CN CN201710816605.1A patent/CN107610773A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102332162A (en) * | 2011-09-19 | 2012-01-25 | 西安百利信息科技有限公司 | Method for automatic recognition and stage compression of medical image regions of interest based on artificial neural network |
| US20150238148A1 (en) * | 2013-10-17 | 2015-08-27 | Siemens Aktiengesellschaft | Method and system for anatomical object detection using marginal space deep neural networks |
| CN105640577A (en) * | 2015-12-16 | 2016-06-08 | 深圳市智影医疗科技有限公司 | Method and system automatically detecting local lesion in radiographic image |
| CN106897573A (en) * | 2016-08-01 | 2017-06-27 | 12西格玛控股有限公司 | Use the computer-aided diagnosis system for medical image of depth convolutional neural networks |
| CN106372390A (en) * | 2016-08-25 | 2017-02-01 | 姹ゅ钩 | Deep convolutional neural network-based lung cancer preventing self-service health cloud service system |
| CN106780460A (en) * | 2016-12-13 | 2017-05-31 | 杭州健培科技有限公司 | A kind of Lung neoplasm automatic checkout system for chest CT image |
| CN107115111A (en) * | 2017-01-23 | 2017-09-01 | 上海联影医疗科技有限公司 | Blood flow state analysis system and method |
Non-Patent Citations (2)
| Title |
|---|
| 万艳丽，雷行云，王岩，胡红濮: "基于层次化深度学习的海量医学影像组织与检索研究", 《医学信息学杂志》 * |
| 李渊，骆志刚，管乃洋，尹晓尧，王兵，伯晓晨，李非: "生物医学数据分析中的深度学习方法应用", 《生物化学与生物物理进展》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109584229A (en) * | 2018-11-28 | 2019-04-05 | 武汉大学人民医院（湖北省人民医院） | A kind of real-time assistant diagnosis system of Endoscopic retrograde cholangio-pancreatiography art and method |
| CN110264465A (en) * | 2019-06-25 | 2019-09-20 | 中南林业科技大学 | A kind of dissection of aorta dynamic testing method based on morphology and deep learning |
| CN110348500A (en) * | 2019-06-30 | 2019-10-18 | 浙江大学 | Sleep disturbance aided diagnosis method based on deep learning and infrared thermal imagery |
| CN110706815A (en) * | 2019-11-26 | 2020-01-17 | 北京推想科技有限公司 | Evaluation method and device of image report and electronic equipment |
| CN111754485A (en) * | 2020-06-24 | 2020-10-09 | 成都市温江区人民医院 | Artificial intelligence ultrasonic auxiliary system for liver |
| CN113223704A (en) * | 2021-05-20 | 2021-08-06 | 吉林大学 | Auxiliary diagnosis method for computed tomography aortic aneurysm based on deep learning |
| CN113763337A (en) * | 2021-08-24 | 2021-12-07 | 慧影医疗科技（北京）有限公司 | Method and system for detecting blood supply of aortic dissection false cavity |
| CN113763337B (en) * | 2021-08-24 | 2024-05-03 | 慧影医疗科技(北京)股份有限公司 | Method and system for detecting blood supply of aortic dissection false cavity |
| CN115131508A (en) * | 2022-08-30 | 2022-09-30 | 南通佳茂霖智能科技有限公司 | DSA modeling point cloud data fusion processing method based on data processing |
| CN115131508B (en) * | 2022-08-30 | 2022-11-25 | 南通佳茂霖智能科技有限公司 | DSA modeling point cloud data fusion processing method based on data processing |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107610773A (en) | A kind of vascular dissection aided diagnosis method based on sustainer medical image | |
| Atwany et al. | Deep learning techniques for diabetic retinopathy classification: A survey | |
| Chan et al. | Texture-map-based branch-collaborative network for oral cancer detection | |
| CN109615636A (en) | Vascular tree building method, device in the lobe of the lung section segmentation of CT images | |
| CN113506310B (en) | Medical image processing method and device, electronic equipment and storage medium | |
| KR101953752B1 (en) | Method for classifying and localizing images using deep neural network and apparatus using the same | |
| Guo et al. | Dense residual network for retinal vessel segmentation | |
| KR102220109B1 (en) | Method for classifying images using deep neural network and apparatus using the same | |
| JP2005519685A (en) | Pixel coding method, image processing method, and image processing method for qualitative recognition of an object reproduced by one or more pixels | |
| CN111667456A (en) | Method and device for detecting vascular stenosis in coronary artery X-ray sequence radiography | |
| CN109685787A (en) | Output method, device in the lobe of the lung section segmentation of CT images | |
| Saeed et al. | Accuracy of using generative adversarial networks for glaucoma detection: systematic review and bibliometric analysis | |
| Li et al. | Multi-stage attention-unet for wireless capsule endoscopy image bleeding area segmentation | |
| Xie et al. | Vessel lumen segmentation in internal carotid artery ultrasounds with deep convolutional neural networks | |
| KR20200087427A (en) | The diagnostic method of lymph node metastasis in thyroid cancer using deep learning | |
| KR20220095342A (en) | The diagnostic method and system of lymph node metastasis in thyroid cancer using ct image | |
| Zhou et al. | TED: Two-stage expert-guided interpretable diagnosis framework for microvascular invasion in hepatocellular carcinoma | |
| Kumar et al. | A Novel Approach for Breast Cancer Detection by Mammograms | |
| Liu et al. | Dias: A comprehensive benchmark for dsa-sequence intracranial artery segmentation | |
| Yang et al. | Detection of microaneurysms and hemorrhages based on improved Hessian matrix | |
| US20240315655A1 (en) | A method for cerebral vessel calcification detection and quantification, using machine learning | |
| Zhang et al. | Blood vessel segmentation based on digital subtraction angiography sequence | |
| Kamolkunasiri et al. | A Comparative Study on Out of Scope Detection for Chest X-ray Images | |
| Shetty et al. | Self-Sequential Attention Layer based DenseNet for Thoracic Diseases Detection. | |
| Guo et al. | SEAM-STRESS: A weakly supervised framework for interstitial lung disease segmentation in chest CT |
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: 20180119 |