[go: up one dir, main page]

CN109919931A - Coronary stenosis degree evaluation model training method and evaluation system - Google Patents

Coronary stenosis degree evaluation model training method and evaluation system Download PDF

Info

Publication number
CN109919931A
CN109919931A CN201910176706.6A CN201910176706A CN109919931A CN 109919931 A CN109919931 A CN 109919931A CN 201910176706 A CN201910176706 A CN 201910176706A CN 109919931 A CN109919931 A CN 109919931A
Authority
CN
China
Prior art keywords
sample
cloud
specific
sample set
samples
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.)
Granted
Application number
CN201910176706.6A
Other languages
Chinese (zh)
Other versions
CN109919931B (en
Inventor
郑超
王振常
杨正汉
韩丹
肖月庭
阳光
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shukun Shenzhen Intelligent Network Technology Co ltd
Original Assignee
Digital Kun (beijing) Network Technology Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Digital Kun (beijing) Network Technology Co Ltd filed Critical Digital Kun (beijing) Network Technology Co Ltd
Priority to CN201910176706.6A priority Critical patent/CN109919931B/en
Publication of CN109919931A publication Critical patent/CN109919931A/en
Application granted granted Critical
Publication of CN109919931B publication Critical patent/CN109919931B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Landscapes

  • Image Analysis (AREA)
  • Apparatus For Radiation Diagnosis (AREA)

Abstract

The invention discloses a kind of coronary stenosis degree evaluation model training methods, comprising: S1, sample is collected from each hospital by cloud method, to form cloud sample set；S2, the cloud scoring for classifying and obtaining classification samples is manually labeled to the sample extracted from cloud sample set；S3, the classification samples training image disaggregated model based on mark；S4, specific sample is collected from certain particular hospital, forms specific sample collection；Specific sample collection is inputted in disaggregated model and is classified；S5, the difficulty value for determining specific sample: S6, complexity is confirmed to specific sample collection；S7, it is distributed the sample for extracting corresponding ratio from cloud sample set according to the complexity of specific sample collection, generates sample set；S8, the evaluation model training of coronary stenosis degree is carried out using sample set.Meanwhile the invention also discloses a kind of coronary stenosis degree evaluation systems.The present invention fully considers the otherness of the scoring tactics of different experts, make its export result be more in line with particular doctor scoring habit and more practicability.

Description

Coronary stenosis degree evaluation model training method and evaluation system
Technical field
The present invention relates to coronary artery art of image analysis, and in particular to a kind of coronary stenosis degree evaluation model training method and comments Valence system.
Background technique
AI can obtain prediction model according to great amount of samples training in the features such as monitoring calcification, soft spot automatically, therefore, Its sample is important input.Sample includes that raw scanning data and mark (i.e. answer), raw scanning data can pass through instrument Device quick obtaining, and mark and then need a large amount of artificial a large amount of participations.
One important intermediate result of CAG & PCI is narrowness, and the narrowness of quantization has in conjunction with raw scanning data Patient is diagnosed conducive to expert, and the calculating difficult point of narrowness is in particular, in that the identification of soft spot and mixing for part punctate clacification The identification for closing patch, needs certain experiences for the identification of the part, usually provides evaluation by expert.But due to difference Hospital or doctor have different Evaluation Strategies, and the evaluation generated to same initial data disagrees, therefore in practical application, The limitation of AI is very big, and analysis result has significant limitation, cannot give the corresponding reference of specific specialists, may draw instead It rises and misleads.
Summary of the invention
The purpose of the present invention is to provide a kind of coronary stenosis evaluation model training methods, are met as far as possible with acquisition specific The coronary stenosis evaluation model of expert opinion strategy.
To achieve the above object, the invention adopts the following technical scheme:
Coronary stenosis degree evaluation model training method, comprising:
S1, obtain cloud sample set: sample collected from each hospital by cloud method, the sample include raw scanning data with Narrowness scoring, to form cloud sample set；
S2, the cloud scoring for classifying and obtaining classification samples is manually labeled to the sample extracted from cloud sample set；
S3, the classification samples training image disaggregated model based on mark；
S4, it obtains specific sample collection and classifies: being collected from certain particular hospital specific corresponding to certain in the hospital or the hospital The specific sample of expert, to form specific sample collection；Specific sample collection is inputted in disaggregated model and is classified；
S5, the difficulty value for determining specific sample: being based on classification results, according to class where the practical scoring of specific sample and its Cloud scoring otherness determine its difficulty value；
S6, complexity is confirmed to specific sample collection based on the difficulty value of specific sample；
The complexity distribution situation that S7, corresponding specific sample are concentrated, which corresponds to extract in classification sample from cloud sample set, to be corresponded to The sample of ratio generates sample set；
S8, the evaluation model training of coronary stenosis degree is carried out using sample set.
Further, in S1, cloud sample set is based on same model data and generates, and the sample standard deviation in subsequent step is based on phase Same type machine data pick-up.
Further, difficulty value is calculate by the following formula:In formula, M1 is class where certain specific sample Cloud scoring, M2 are the practical scoring of certain specific sample.
The invention also discloses a kind of coronary stenosis degree evaluation systems, comprising:
Cloud sample set obtains module, and the cloud sample set obtains module and collects sample from each hospital by cloud method, described Sample includes that raw scanning data and narrowness score, to form cloud sample set；
Human-computer interaction module, cloud expert by the human-computer interaction module, manually go forward side by side from cloud sample set by sample drawn Rower note classification, while obtaining the cloud scoring of each classification samples；
Sample classification module, the sample classification module are built-in with image classification model, and described image disaggregated model is based on The classification samples of mark are trained；
Specific sample collection acquisition module, the specific sample collection acquisition module is collected from certain particular hospital corresponds to the hospital Or in the hospital certain specific specialists specific sample, to form specific sample collection, and specific sample collection is input to sample classification Classify in module；
Specific sample evaluation module, the specific sample evaluation module is based on classification results, according to the reality of specific sample The otherness of scoring and the cloud scoring of class where it determines its difficulty value；And based on the difficulty value of specific sample to specific sample collection Confirm complexity；
Sampling module, the complexity distribution situation that the sampling module is concentrated based on specific sample is from cloud sample The sample that corresponding proportion is extracted in the corresponding classification sample of this collection, generates sample set；
Coronary stenosis degree evaluation module, the coronary stenosis degree evaluation module are built-in with coronary stenosis degree evaluation model, institute Coronary stenosis degree evaluation model is stated to be trained based on the sample set.
After adopting the above technical scheme, compared with the background technology, the present invention, having the advantages that
The present invention fully considers the otherness of Different hospital or the Evaluation Strategy of different experts, evaluates sample difficulty Classification carries out model training according to complexity distribution situation to obtain the complexity distribution situation of sample in particular hospital The extraction of sample is allowed to meet the scoring habit of specific specialists in particular hospital or particular hospital, joins its appraisal result more The property examined promotes the efficiency of doctor.
Detailed description of the invention
Fig. 1 is coronary stenosis degree evaluation model training method flow chart of the present invention；
Fig. 2 is coronary stenosis degree evaluation system block diagram of the present invention.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to the accompanying drawings and embodiments, right The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and It is not used in the restriction present invention.
In the present invention it should be noted that term " on " "lower" " left side " " right side " "vertical" "horizontal" "inner" "outside" etc. is Be based on the orientation or positional relationship shown in the drawings, it is only for convenient for description the present invention and simplify description, rather than indicate or It implies that the device of the invention or element must have a particular orientation, therefore is not considered as limiting the invention.
Embodiment 1
It please refers to shown in Fig. 1, it mainly to include 8 that the invention discloses a kind of coronary stenosis degree evaluation model training methods A core procedure.
S1, obtain cloud sample set: sample collected from each hospital by cloud method, the sample include raw scanning data with Narrowness scoring, to form cloud sample set.
Pair consider the otherness that different type of machines data generate, in S1, cloud sample set is based on same model data and generates, i.e., In same model data, the data of Different hospital can be mixed as big data acquisition system, the sample in subsequent step It is based on same model data pick-up.
S2, the cloud scoring for classifying and obtaining classification samples is manually labeled to the sample extracted from cloud sample set.
Sample is collected, manual sort is carried out to sample based on picture quality and lesion appearance position, by lesion appearance position And picture quality is similar is classified as one kind, assigns tag along sort, obtains training sample.Meanwhile each classification samples can be produced Raw cloud scoring (i.e. similar scoring).Cloud scoring can be being averaged for the narrowness of each sample of classification samples concentration Point, it is also possible to cloud expert and is directed to the comprehensive score that such sample provides.
S3, the classification samples training image disaggregated model based on mark.
By the processing of S2, training sample is all manually to annotate image image whether similar with certain class image, is obtained Such " stacking " mark, which can train, obtains image classification model.
S4, it obtains specific sample collection and classifies: being collected from certain particular hospital specific corresponding to certain in the hospital or the hospital The specific sample of expert, to form specific sample collection；Specific sample collection is inputted in disaggregated model and is classified.
The acquisition of specific sample collection can be randomly selects 100-200 corresponding to the hospital or the doctor as unit of year The specific sample of institute's specific specialists, it is opposite in this way to meet NATURAL DISTRIBUTION.
S5, the difficulty value for determining specific sample: being based on classification results, according to class where the practical scoring of specific sample and its Cloud scoring otherness determine its difficulty value.
The concept of difficulty value is obtained by the practical marking of specific specialists and the cloud scoring of such sample.Score is lower, generation The marking habit of the table specific specialists is close with cloud, and different experts reach unanimity to the marking of the sample, represents the sample and gets over Simply；Score is higher, and marking habit and the cloud otherness for representing the specific specialists are larger, special in the marking result of the sample It is different to determine the strategy that expert is taken, represents the sample and is more difficult to.
Difficulty value is calculate by the following formula:In formula, M1 is the cloud scoring of class where certain specific sample, M2 For the practical scoring of certain specific sample.
S6, complexity is confirmed to specific sample collection based on the difficulty value of specific sample.
Complexity can be divided into several grades by sequence from the easier to the more advanced.
Such as, there are 8 samples.Each sample difficulty value be respectively 32%, 35%, 48%, 60%, 62%, 69%, 70% and 78%, complexity is divided into 4 grades, then is exactly to divide a class from 0~100% by every 25% interval.Then 32%, 35%, 48% belongs to the second class；60%, 62%, 69%, 70% belongs to third class；78% belongs to the 4th class.
The complexity distribution situation that S7, corresponding specific sample are concentrated, which corresponds to extract in classification sample from cloud sample set, to be corresponded to The sample of ratio generates sample set.
Such as to 4 classes of grade of difficulty point, concentrating distribution situation from the easier to the more advanced in specific sample is 3:3:2:2, then pressing According to this kind proportion, sample drawn generates sample set from cloud sample set.
S8, the evaluation model training of coronary stenosis degree is carried out using sample set.
Embodiment 2
The invention also discloses a kind of coronary stenosis degree evaluation systems, comprising:
Cloud sample set obtains module, and the cloud sample set obtains module and collects sample from each hospital by cloud method, described Sample includes that raw scanning data and narrowness score, to form cloud sample set；
Human-computer interaction module, cloud expert by the human-computer interaction module, manually go forward side by side from cloud sample set by sample drawn Rower note classification, while obtaining the cloud scoring of each classification samples；
Sample classification module, the sample classification module are built-in with image classification model, and described image disaggregated model is based on The classification samples of mark are trained；
Specific sample collection acquisition module, the specific sample collection acquisition module is collected from certain particular hospital corresponds to the hospital Or in the hospital certain specific specialists specific sample, to form specific sample collection, and specific sample collection is input to sample classification Classify in module；
Specific sample evaluation module, the specific sample evaluation module is based on classification results, according to the reality of specific sample The otherness of scoring and the cloud scoring of class where it determines its difficulty value；And based on the difficulty value of specific sample to specific sample collection Confirm complexity；
Sampling module, the complexity distribution situation that the sampling module is concentrated based on specific sample is from cloud sample The sample that corresponding proportion is extracted in the corresponding classification sample of this collection, generates sample set；
Coronary stenosis degree evaluation module, the coronary stenosis degree evaluation module are built-in with coronary stenosis degree evaluation model, institute Coronary stenosis degree evaluation model is stated to be trained based on the sample set.
The foregoing is only a preferred embodiment of the present invention, but scope of protection of the present invention is not limited thereto, In the technical scope disclosed by the present invention, any changes or substitutions that can be easily thought of by anyone skilled in the art, It should be covered by the protection scope of the present invention.Therefore, protection scope of the present invention should be with scope of protection of the claims Subject to.

Claims (4)

1.冠脉狭窄度评价模型训练方法，其特征在于，包括：1. a method for training a coronary artery stenosis degree evaluation model, characterized in that it comprises: S1、获取云样本集：通过云方法从各医院收集样本，所述样本包括原始扫描数据与狭窄度评分，以形成云样本集；S1. Obtain a cloud sample set: collect samples from various hospitals through a cloud method, and the samples include raw scan data and stenosis scores to form a cloud sample set; S2、人工对从云样本集中抽取的样本进行标注分类并获得分类样本的云评分；S2, manually label and classify the samples extracted from the cloud sample set and obtain the cloud score of the classified samples; S3、基于标注的分类样本训练图像分类模型；S3. An image classification model is trained based on the labeled classification samples; S4、获取特定样本集并分类：从某特定医院收集对应于该医院或该医院中某特定专家的特定样本，以形成特定样本集；将特定样本集输入分类模型中分类；S4. Obtain and classify a specific sample set: collect specific samples corresponding to a specific hospital or a specific expert in the hospital from a specific hospital to form a specific sample set; input the specific sample set into a classification model for classification; S5、确定特定样本的难度值：基于分类结果，根据特定样本的实际评分与其所在类的云评分的差异性确定其难度值；S5. Determine the difficulty value of a specific sample: Based on the classification result, determine the difficulty value of the specific sample according to the difference between the actual score of the specific sample and the cloud score of the class it belongs to; S6、基于特定样本的难度值对特定样本集确认难易程度；S6. Confirm the difficulty level of a specific sample set based on the difficulty value of the specific sample; S7、对应特定样本集中的难易程度分布情况从云样本集对应类别样本中抽取对应比例的样本，生成样本子集；S7, corresponding to the difficulty degree distribution in a specific sample set, extract a corresponding proportion of samples from the corresponding category samples of the cloud sample set, and generate a sample subset; S8、使用样本子集进行冠脉狭窄度评价模型训练。S8. Use the sample subset to train the coronary artery stenosis degree evaluation model. 2.如权利要求1所述的冠脉狭窄度评价模型训练方法，其特征在于：S1中，云样本集基于相同机型数据而生成，后续步骤中的样本均基于相同机型数据抽取。2 . The method for training a coronary artery stenosis degree evaluation model according to claim 1 , wherein in S1 , the cloud sample set is generated based on the same model data, and the samples in the subsequent steps are all extracted based on the same model data. 3 . 3.如权利要求1所述的冠脉狭窄度评价模型训练方法，其特征在于，难度值通过下式计算：式中，M1为某特定样本所在类的云评分，M2为某特定样本的实际评分。3. coronary artery stenosis degree evaluation model training method as claimed in claim 1, is characterized in that, difficulty value is calculated by following formula: In the formula, M1 is the cloud score of the class of a specific sample, and M2 is the actual score of a specific sample. 4.冠脉狭窄度评价系统，其特征在于，包括：4. Coronary stenosis evaluation system, characterized in that it includes: 云样本集获取模块，所述云样本集获取模块通过云方法从各医院收集样本，所述样本包括原始扫描数据与狭窄度评分，以形成云样本集；A cloud sample set acquisition module, the cloud sample set acquisition module collects samples from various hospitals through a cloud method, and the samples include raw scan data and stenosis scores to form a cloud sample set; 人机交互模块，云端专家通过所述人机交互模块人工从云样本集中抽取样本并进行标注分类，同时获得各分类样本的云评分；a human-computer interaction module, through which cloud experts manually extract samples from the cloud sample set, label and classify them, and obtain cloud scores for each classified sample; 样本分类模块，所述样本分类模块内置有图像分类模型，所述图像分类模型基于标注的分类样本进行训练；a sample classification module, the sample classification module has a built-in image classification model, and the image classification model is trained based on the labeled classified samples; 特定样本集采集模块，所述特定样本集采集模块从某特定医院收集对应于该医院或该医院中某特定专家的特定样本，以形成特定样本集，并将特定样本集输入到样本分类模块中分类；A specific sample set collection module that collects specific samples from a specific hospital corresponding to the hospital or a specific expert in the hospital to form a specific sample set, and inputs the specific sample set into the sample classification module Classification; 特定样本评价模块，所述特定样本评价模块基于分类结果，根据特定样本的实际评分与其所在类的云评分的差异性确定其难度值；并基于特定样本的难度值对特定样本集确认难易程度；A specific sample evaluation module, the specific sample evaluation module determines the difficulty value of the specific sample based on the difference between the actual score of the specific sample and the cloud score of its class based on the classification result; and confirms the difficulty level of the specific sample set based on the difficulty value of the specific sample ; 样本抽取模块，所述样本抽取模块基于特定样本集中的难易程度分布情况从云样本集对应类别样本中抽取相应比例的样本，生成样本子集；a sample extraction module, the sample extraction module extracts a corresponding proportion of samples from the corresponding category samples of the cloud sample set based on the difficulty degree distribution in a specific sample set, and generates a sample subset; 冠脉狭窄度评价模块，所述冠脉狭窄度评价模块内置有冠脉狭窄度评价模型，所述冠脉狭窄度评价模型基于所述样本子集进行训练。A coronary artery stenosis degree evaluation module, the coronary artery stenosis degree evaluation module has a built-in coronary artery stenosis degree evaluation model, and the coronary artery stenosis degree evaluation model is trained based on the sample subset.
CN201910176706.6A 2019-03-08 2019-03-08 Coronary stenosis degree evaluation model training method and evaluation system Active CN109919931B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910176706.6A CN109919931B (en) 2019-03-08 2019-03-08 Coronary stenosis degree evaluation model training method and evaluation system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910176706.6A CN109919931B (en) 2019-03-08 2019-03-08 Coronary stenosis degree evaluation model training method and evaluation system

Publications (2)

Publication Number Publication Date
CN109919931A true CN109919931A (en) 2019-06-21
CN109919931B CN109919931B (en) 2020-12-25

Family

ID=66963936

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910176706.6A Active CN109919931B (en) 2019-03-08 2019-03-08 Coronary stenosis degree evaluation model training method and evaluation system

Country Status (1)

Country Link
CN (1) CN109919931B (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110568483A (en) * 2019-07-22 2019-12-13 中国石油化工股份有限公司 Automatic evaluation method of seismic linear noise suppression effect based on convolutional neural network
CN119324064A (en) * 2024-12-19 2025-01-17 北京清华长庚医院 Coronary artery stenosis severity determination method, device, electronic equipment and storage medium

Citations (23)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101071445A (en) * 2007-06-22 2007-11-14 腾讯科技（深圳）有限公司 Classified sample set optimizing method and content-related advertising server
US8515777B1 (en) * 2010-10-13 2013-08-20 ProcessProxy Corporation System and method for efficient provision of healthcare
CN104572820A (en) * 2014-12-03 2015-04-29 百度在线网络技术（北京）有限公司 Method and device for generating model and method and device for acquiring importance degree
CN104679860A (en) * 2015-02-27 2015-06-03 北京航空航天大学 Classifying method for unbalanced data
CN104751642A (en) * 2015-03-11 2015-07-01 同济大学 Real-time estimating method for high-grade road traffic flow running risks
CN105320957A (en) * 2014-07-10 2016-02-10 腾讯科技（深圳）有限公司 Classifier training method and device
CN105550747A (en) * 2015-12-09 2016-05-04 四川长虹电器股份有限公司 Sample training method for novel convolutional neural network
CN107133628A (en) * 2016-02-26 2017-09-05 阿里巴巴集团控股有限公司 A kind of method and device for setting up data identification model
CN107392259A (en) * 2017-08-16 2017-11-24 北京京东尚科信息技术有限公司 The method and apparatus for building unbalanced sample classification model
CN108135003A (en) * 2017-12-25 2018-06-08 广东海格怡创科技有限公司 The construction method and system of interference type identification model
CN108171175A (en) * 2017-12-29 2018-06-15 苏州科达科技股份有限公司 A kind of deep learning sample enhancing system and its operation method
CN108197545A (en) * 2017-12-25 2018-06-22 广东海格怡创科技有限公司 The recognition methods of interference type and system
CN108280462A (en) * 2017-12-11 2018-07-13 北京三快在线科技有限公司 A kind of model training method and device, electronic equipment
CN108389201A (en) * 2018-03-16 2018-08-10 北京推想科技有限公司 The good pernicious sorting technique of Lung neoplasm based on 3D convolutional neural networks and deep learning
CN108573040A (en) * 2018-04-08 2018-09-25 西北工业大学 A kind of sample set optimization algorithm based on target distribution
CN108830155A (en) * 2018-05-10 2018-11-16 北京红云智胜科技有限公司 A kind of heart coronary artery segmentation and knowledge method for distinguishing based on deep learning
CN108985369A (en) * 2018-07-06 2018-12-11 太原理工大学 A kind of same distribution for unbalanced dataset classification integrates prediction technique and system
CN109036551A (en) * 2018-07-10 2018-12-18 北京心世纪医疗科技有限公司 A kind of coronary artery physiologic index relationship is established and application method, device
CN109146667A (en) * 2018-08-20 2019-01-04 众安在线财产保险股份有限公司 A kind of construction method of the external interface integrated application model based on quantitative statistics
US20190017119A1 (en) * 2017-07-12 2019-01-17 The General Hospital Corporation Genetic Risk Predictor
CN109272514A (en) * 2018-10-05 2019-01-25 数坤（北京）网络科技有限公司 The sample evaluation method and model training method of coronary artery parted pattern
CN109327464A (en) * 2018-11-15 2019-02-12 中国人民解放军战略支援部队信息工程大学 A method and device for processing class imbalance in network intrusion detection
CN109346179A (en) * 2018-12-10 2019-02-15 陈阳 Coronary heart disease recurrence prediction model after interventional operation and its modeling method and device

Patent Citations (23)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101071445A (en) * 2007-06-22 2007-11-14 腾讯科技（深圳）有限公司 Classified sample set optimizing method and content-related advertising server
US8515777B1 (en) * 2010-10-13 2013-08-20 ProcessProxy Corporation System and method for efficient provision of healthcare
CN105320957A (en) * 2014-07-10 2016-02-10 腾讯科技（深圳）有限公司 Classifier training method and device
CN104572820A (en) * 2014-12-03 2015-04-29 百度在线网络技术（北京）有限公司 Method and device for generating model and method and device for acquiring importance degree
CN104679860A (en) * 2015-02-27 2015-06-03 北京航空航天大学 Classifying method for unbalanced data
CN104751642A (en) * 2015-03-11 2015-07-01 同济大学 Real-time estimating method for high-grade road traffic flow running risks
CN105550747A (en) * 2015-12-09 2016-05-04 四川长虹电器股份有限公司 Sample training method for novel convolutional neural network
CN107133628A (en) * 2016-02-26 2017-09-05 阿里巴巴集团控股有限公司 A kind of method and device for setting up data identification model
US20190017119A1 (en) * 2017-07-12 2019-01-17 The General Hospital Corporation Genetic Risk Predictor
CN107392259A (en) * 2017-08-16 2017-11-24 北京京东尚科信息技术有限公司 The method and apparatus for building unbalanced sample classification model
CN108280462A (en) * 2017-12-11 2018-07-13 北京三快在线科技有限公司 A kind of model training method and device, electronic equipment
CN108135003A (en) * 2017-12-25 2018-06-08 广东海格怡创科技有限公司 The construction method and system of interference type identification model
CN108197545A (en) * 2017-12-25 2018-06-22 广东海格怡创科技有限公司 The recognition methods of interference type and system
CN108171175A (en) * 2017-12-29 2018-06-15 苏州科达科技股份有限公司 A kind of deep learning sample enhancing system and its operation method
CN108389201A (en) * 2018-03-16 2018-08-10 北京推想科技有限公司 The good pernicious sorting technique of Lung neoplasm based on 3D convolutional neural networks and deep learning
CN108573040A (en) * 2018-04-08 2018-09-25 西北工业大学 A kind of sample set optimization algorithm based on target distribution
CN108830155A (en) * 2018-05-10 2018-11-16 北京红云智胜科技有限公司 A kind of heart coronary artery segmentation and knowledge method for distinguishing based on deep learning
CN108985369A (en) * 2018-07-06 2018-12-11 太原理工大学 A kind of same distribution for unbalanced dataset classification integrates prediction technique and system
CN109036551A (en) * 2018-07-10 2018-12-18 北京心世纪医疗科技有限公司 A kind of coronary artery physiologic index relationship is established and application method, device
CN109146667A (en) * 2018-08-20 2019-01-04 众安在线财产保险股份有限公司 A kind of construction method of the external interface integrated application model based on quantitative statistics
CN109272514A (en) * 2018-10-05 2019-01-25 数坤（北京）网络科技有限公司 The sample evaluation method and model training method of coronary artery parted pattern
CN109327464A (en) * 2018-11-15 2019-02-12 中国人民解放军战略支援部队信息工程大学 A method and device for processing class imbalance in network intrusion detection
CN109346179A (en) * 2018-12-10 2019-02-15 陈阳 Coronary heart disease recurrence prediction model after interventional operation and its modeling method and device

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
于明川 等: "不同图像后处理技术评价冠状动脉狭窄静态模型中血管狭窄程度", 《中国医学影像技术》 *
于重重 等: "非平衡样本分类的集成迁移学习算法", 《电子学报》 *
钱政 等: "组合神经网络模型中典型训练样本集的选取", 《高电压技术》 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110568483A (en) * 2019-07-22 2019-12-13 中国石油化工股份有限公司 Automatic evaluation method of seismic linear noise suppression effect based on convolutional neural network
CN119324064A (en) * 2024-12-19 2025-01-17 北京清华长庚医院 Coronary artery stenosis severity determination method, device, electronic equipment and storage medium

Also Published As

Publication number Publication date
CN109919931B (en) 2020-12-25

Similar Documents

Publication Publication Date Title
CN107767376B (en) X-ray bone age prediction method and system based on deep learning
CN112184617B (en) Spine MRI image key point detection method based on deep learning
Ogiela et al. Modern computational intelligence methods for the interpretation of medical images
CN109858540B (en) Medical image recognition system and method based on multi-mode fusion
CN111524580A (en) Virus screening and epidemic prevention system and method based on neural network
CN109727243A (en) Breast ultrasound image recognition and analysis method and system
CN109272048A (en) A Pattern Recognition Method Based on Deep Convolutional Neural Network
CN106203488B (en) A Breast Image Feature Fusion Method Based on Restricted Boltzmann Machine
CN110335665A (en) A method and system for image search by image applied to aided diagnosis and analysis of medical images
CN110097969A (en) A kind of analysis method of diagnosis report, device and equipment
CN110543912B (en) A method for automatically obtaining the central dynamic cycle video of the ultrasound video of the fetal key section
CN110009605A (en) A method and system for bone age prediction based on deep learning
CN102429679A (en) Computer-assisted emphysema analysis system based on chest CT (Computerized Tomography) image
CN111179227A (en) Evaluation method of breast ultrasound image quality based on auxiliary diagnosis and subjective aesthetics
CN114358194B (en) Abnormal limb behavior detection method for autism spectrum disorder based on posture tracking
CN108596174A (en) A kind of lesion localization method of skin disease image
Hu et al. Object detection in hospital facilities: A comprehensive dataset and performance evaluation
CN105139390A (en) Image processing method for detecting pulmonary tuberculosis focus in chest X-ray DR film
Shen et al. Guest editorial: special issue on imaging-based diagnosis of COVID-19
CN109817333A (en) A kind of syndrome Differentiation System of Traditional Chinese Medicine and its method of four methods of diagnosis information weight analysis
CN113397485A (en) Scoliosis screening method based on deep learning
CN112690815A (en) System and method for assisting in diagnosing lesion grade based on lung image report
CN109919931A (en) Coronary stenosis degree evaluation model training method and evaluation system
CN117237351A (en) Ultrasonic image analysis method and related device
CN115349828A (en) Neonate pain assessment system based on computer deep learning

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
CP03 Change of name, title or address
CP03 Change of name, title or address

Address after: 100120 rooms 303, 304, 305, 321 and 322, building 3, No. 11, Chuangxin Road, science and Technology Park, Changping District, Beijing

Patentee after: Shukun (Beijing) Network Technology Co.,Ltd.

Address before: 100102 No. 501 No. 12, 5th floor, No. 6, Wangjing Dongyuan District 4, Chaoyang District, Beijing

Patentee before: SHUKUN (BEIJING) NETWORK TECHNOLOGY Co.,Ltd.

TR01 Transfer of patent right
TR01 Transfer of patent right

Effective date of registration: 20230117

Address after: 518026 Rongchao Economic and Trade Center A308-D9, No. 4028, Jintian Road, Fuzhong Community, Lianhua Street, Futian District, Shenzhen, Guangdong Province

Patentee after: Shukun (Shenzhen) Intelligent Network Technology Co.,Ltd.

Address before: 100120 rooms 303, 304, 305, 321 and 322, building 3, No. 11, Chuangxin Road, science and Technology Park, Changping District, Beijing

Patentee before: Shukun (Beijing) Network Technology Co.,Ltd.