CN109448855A - A kind of diabetes glucose prediction technique based on CNN and Model Fusion - Google Patents
A kind of diabetes glucose prediction technique based on CNN and Model Fusion Download PDFInfo
- Publication number
- CN109448855A CN109448855A CN201811079861.8A CN201811079861A CN109448855A CN 109448855 A CN109448855 A CN 109448855A CN 201811079861 A CN201811079861 A CN 201811079861A CN 109448855 A CN109448855 A CN 109448855A
- Authority
- CN
- China
- Prior art keywords
- data
- layer
- model
- cnn
- linearregression
- 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
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/50—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for simulation or modelling of medical disorders
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Public Health (AREA)
- Medical Informatics (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Pathology (AREA)
- Epidemiology (AREA)
- Primary Health Care (AREA)
- Databases & Information Systems (AREA)
- Biophysics (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Evolutionary Computation (AREA)
- Computational Linguistics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Investigating Or Analysing Biological Materials (AREA)
Abstract
The diabetes glucose prediction technique based on CNN and Model Fusion that the present invention relates to a kind of.The following steps are included: being pre-processed first to data, the binaryzation and data conversion of processing, quantitative characteristic including vacancy value.Secondly, carrying out feature extraction to pretreated data using CNN.Model Fusion finally is carried out to xgboost, catboost and linearRegression using Stacking strategy, higher catboost, xgboost model of nicety of grading is used in first layer, second layer model uses the preferable linearRegression of robustness, the unification of precision of prediction and model robustness is realized, there is stronger generalization ability.The effective solution of the present invention forecasting problem of diabetes glucose, compared to traditional blood glucose prediction method has significant raising.
Description
Technical field
The present invention relates to machine learning fields, and in particular to carries out spy to structural data with convolutional neural networks (CNN)
Sign is extracted and is predicted using Model Fusion technology blood glucose.
Background technique
In recent years, the illness rate of diabetes is worldwide in rising trend, it has also become at present after tumour and heart and brain blood
Third position threatens the great non-communicable diseases of human health after pipe disease.Diabetes can not eradicate at present, can only pass through section
Learning effective prevention reduces disease incidence.Due to blood sugar concentration often with other physical examination indexs of people there are in certain contact,
Therefore, the illness of blood sugar concentration prediction model accurate evaluation people is established according to other physical examination indexs using machine learning techniques
Risk, and then high-risk individuals are intervened, help to realize effective prevention of disease.
Summary of the invention
In order to solve the above technical problems, the present invention provides a kind of based on convolutional neural networks (CNN) and Model Fusion
Diabetes glucose prediction technique, this method are to predict the blood sugar concentration of patient by indices in analysis physical examination data, assess
The risk of people, and then early warning is carried out to high-risk individuals, help to realize the prevention of diabetes.Specific step includes:
S1: pre-processing data, the binaryzation and data conversion of processing, quantitative characteristic including vacancy value；
S2: feature extraction is carried out to pretreated data using CNN；
S3: Model Fusion is carried out to xgboost, catboost and linearRegression using Stacking strategy.
Further, data prediction described in S1 specifically includes: (1) processing of vacancy value: a large amount of for there are data
The attribute of deficient phenomena does not consider these attributes present invention hair；For adopting there are the attribute of low volume data deficient phenomena
Averaging method is taken to carry out completion processing to the data of missing；(2) One Hot quantitative characteristic binaryzation: is carried out to nonumeric type data
Coding, is converted to numeric type data for categorical data；(3) data conversion: to treated, data do log1P transformation.
Further, it carries out feature extraction to pretreated data using CNN described in S2 to specifically include: the CNN
It altogether include 6 layers: 2 convolutional layers, 2 pond layers and a full articulamentum, in which: (1) first convolutional layer is by 64 convolution kernels
It constitutes, convolution kernel size is 1*5；Second convolutional layer is made of 32 convolution kernels, and convolution kernel size is 1*3；(2) 2 ponds
For layer respectively behind first and second convolutional layer, window size is 2, is carried out for realizing to time domain 1D signal
Maximum value pond；(3) the last layer is full articulamentum, and the output of 26 Batch-Normalized will be used as new feature in this layer
Building for eigenmatrix.
In an iterative process, use mean square deviation as loss function, using Adam optimizer to influence model training and mould
The network parameter of type output is updated.
Further, utilize Stacking strategy to xgboost, catboost and linearRegression described in S3
Model Fusion is carried out to specifically include；Catboost, xgboost are chosen as 1 layer of regression model, selection
Training set is first divided into 5 parts in first layer model training by regression model of the linearRegression as the second layer, according to
Secondary to be trained with 4 parts therein, 1 part is used to verify, and after 5 training, the prediction result of verifying collection is stitched together, is made
For the feature of second layer training set, meanwhile, the prediction result of test set is averaged, as the feature of second layer test set, most
Regression forecasting is carried out to the feature that first layer exports with linearRegression model in the second layer afterwards.
The present invention predicts the blood glucose of user, judges that user suffers from diabetes by the physical examination data of analysis user
Risk clearly recognizes so that user itself be made to have oneself potential illness probability, and it is auxiliary to treatment offer to be also beneficial to doctor
Foundation is helped, and then high-risk individuals are intervened, helps to realize effective prevention of diabetes.The Advantageous that the present invention obtains
Effect is as follows: (1) will extract feature capabilities CNN outstanding in high dimensional data and apply in the feature extraction of blood glucose level data, keep away
The shortcomings that having exempted from the manual intervention that traditional characteristic extracting mode workload is quite big and needs are more.(2) by xgboost,
Tri- regression models of catboost, linearRegression are fused together prediction blood sugar concentration.Respectively using different models
Different advantages predicts that fused model preferably to blood glucose.
Detailed description of the invention
Fig. 1 is based on Stacking Policy model fusion structure figure.
Fig. 2 CNN extracts feature extracting method structure chart.
Specific embodiment
In order to more specifically describe the present invention, with reference to the accompanying drawing and specifically embodiment to technology of the invention
Scheme is described in detail.
Implementation of the invention, using jupyter notebook as development platform, using python as development language.The number used
6642 patient's physical examination data are shared according to from Tianchi data platform, every data there are 40 physical examination indexs.Result is commented
Valence uses evaluation index of the mean square deviation (MSE) as prediction effect.
The present invention is based on the diabetes glucose prediction technique of CNN and Model Fusion, specific step is as follows:
Pretreatment of the S1 to data:
(1) processing of vacancy value:
1) for there are indexs such as hepatitis B e antibody, hepatitis B virus e antigen of a large amount of deficient phenomenas of data etc., due to the number of missing
According to excessive (more than 50%) is measured, completion can not be carried out by effective algorithm and therefore these indexs are directly carried out at deletion
Reason.
2) for there are the index of low volume data deficient phenomena, the present invention such as globulin, albumin take average with calculating
Several methods carries out completion processing to the data of missing.
(2) quantitative characteristic binaryzation: the present invention has carried out One_Hot coding to nonumeric type data, such as gender, will be non-
Numeric data is converted to numeric type data, in order to the analysis application of next step data.
(3) data conversion: the present invention does log1P transformation to treated data, makes it possible to preferably be returned
Return prediction.
S2 carries out feature extraction to pretreated data using CNN:
As shown in Fig. 2, the convolutional neural networks that the present invention uses include 6 layers: 2 convolutional layers altogether, 2 pond layers and one
Full articulamentum.Wherein:
(1) first convolutional layer is made of 64 convolution kernels, and convolution kernel size is 1*5；Second convolutional layer is rolled up by 32
Product core is constituted, and convolution kernel size is 1*3.
For (2) two pond layers respectively behind first and second convolutional layer, window size is 2, is used for
It realizes and maximum value pond is carried out to time domain 1D signal.
(3) the last layer is full articulamentum, and the output of 26 Batch-Normalized will be used as new feature in this layer
In the building of eigenmatrix.
In an iterative process, use mean square deviation (MSE) as loss function, using Adam optimizer to influence model training
It is updated with the network parameter of model output.
According to the above, the present invention extracts feature to utilization CNN and traditional utilization mutual information, singular value decomposition are extracted
Feature compares.Test result is as shown in table 1, finds these three moulds of catboost, xgboost and linearRegression
Type is that prediction effect is best in the feature extracted using CNN.Reason, which is CNN model not only, has stronger extensive energy
Power, and it is good at the excavation of data local feature and the extraction of global training characteristics, it is more advantageous to the prediction of blood sugar concentration.
Table 1 uses the forecast of regression model performance of different characteristic extracting method
S3 carries out Model Fusion to xgboost, catboost and linearRegression using Stacking strategy:
As shown in Figure 1, the present invention when carrying out Model Fusion using Stacking strategy, chooses catboost, xgboost
As 1 layer of regression model, regression model of the linearRegression as the second layer is selected.In first layer model training
When, training set is first divided into 5 parts by the present invention, is successively trained with 4 parts therein, and 1 part is used to verify.After 5 training,
The prediction result of verifying collection is stitched together, the feature as second layer training set.At the same time, to the prediction result of test set
It is averaged, the feature as second layer test set.It is finally defeated to first layer with linearRegression model in the second layer
Feature out carries out regression forecasting.
According to the above, the present invention uses the Model Fusion based on Stacking strategy, the mean square error of prediction result
It is 0.7885.The reason is that, the Model Fusion method based on Stacking strategy is higher using nicety of grading in first layer
Catboost, xgboost model, the second layer model use the preferable linearRegression of robustness, realize prediction
The unification of precision and model robustness has stronger generalization ability.
The foregoing is only a preferred embodiment of the present invention, but scope of protection of the present invention is not limited thereto,
Anyone skilled in the art within the technical scope of the present disclosure, according to the technique and scheme of the present invention and its
Inventive concept is subject to equivalent substitution or change, should be covered by the scope of protection of the present invention.
Claims (4)
1. a kind of diabetes glucose prediction technique based on CNN and Model Fusion, which is characterized in that this method includes following step
It is rapid:
S1: pre-processing data, the binaryzation and data conversion of processing, quantitative characteristic including vacancy value；
S2: feature extraction is carried out to pretreated data using CNN；
S3: Model Fusion is carried out to xgboost, catboost and linearRegression using Stacking strategy.
2. a kind of diabetes glucose prediction technique based on CNN and Model Fusion as described in claim 1, which is characterized in that
Data prediction described in S1 specifically includes:
(1) processing of vacancy value: for there are the attributes of a large amount of deficient phenomenas of data, these attributes are not considered；For
There are the attributes of low volume data deficient phenomena, and averaging method is taken to carry out completion processing to the data of missing；
(2) quantitative characteristic binaryzation: One Hot coding is carried out to nonumeric type data, categorical data is converted into numeric type
Data；
(3) data conversion: to treated, data do log1P transformation.
3. a kind of diabetes glucose prediction technique based on CNN and Model Fusion as described in claim 1, which is characterized in that
Feature extraction is carried out to pretreated data using CNN described in S2 to specifically include:
The CNN includes 6 layers: 2 convolutional layers, 2 pond layers and a full articulamentum altogether, in which:
(1) first convolutional layer is made of 64 convolution kernels, and convolution kernel size is 1*5；Second convolutional layer is by 32 convolution kernels
It constitutes, convolution kernel size is 1*3；
(2) 2 pond layers are respectively behind first and second convolutional layer, and window size is 2, for realizing right
Time domain 1D signal carries out maximum value pond；
(3) the last layer is full articulamentum, and the output of 26 Batch-Normalized will be used for spy as new feature in this layer
Levy the building of matrix；
In an iterative process, use mean square deviation as loss function, it is defeated to influence model training and model using Adam optimizer
Network parameter out is updated.
4. a kind of diabetes glucose prediction technique based on CNN and Model Fusion as described in claim 1, which is characterized in that
It is specific to xgboost, catboost and linearRegression progress Model Fusion using Stacking strategy described in S3
Including；
Catboost, xgboost are chosen as 1 layer of regression model, selects linearRegression returning as the second layer
Return model, in first layer model training, training set is first divided into 5 parts, is successively trained with 4 parts therein, 1 part is used to test
The prediction result of verifying collection is stitched together by card after 5 training, as the feature of second layer training set, meanwhile, to survey
The prediction result of examination collection is averaged, and as the feature of second layer test set, finally uses linearRegression in the second layer
Model carries out regression forecasting to the feature that first layer exports.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811079861.8A CN109448855A (en) | 2018-09-17 | 2018-09-17 | A kind of diabetes glucose prediction technique based on CNN and Model Fusion |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811079861.8A CN109448855A (en) | 2018-09-17 | 2018-09-17 | A kind of diabetes glucose prediction technique based on CNN and Model Fusion |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109448855A true CN109448855A (en) | 2019-03-08 |
Family
ID=65533238
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811079861.8A Pending CN109448855A (en) | 2018-09-17 | 2018-09-17 | A kind of diabetes glucose prediction technique based on CNN and Model Fusion |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109448855A (en) |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110120265A (en) * | 2019-04-29 | 2019-08-13 | 天津大学 | The method of raising prediction blood uric acid precision based on multidimensional characteristic and Model Fusion |
| CN110197706A (en) * | 2019-04-26 | 2019-09-03 | 深圳市宁远科技股份有限公司 | A kind of stratification feature selection approach, system and application based on SBS |
| CN110246577A (en) * | 2019-05-31 | 2019-09-17 | 深圳江行联加智能科技有限公司 | A method of based on artificial intelligence auxiliary gestational diabetes genetic risk prediction |
| CN110289097A (en) * | 2019-07-02 | 2019-09-27 | 重庆大学 | A kind of Pattern Recognition Diagnosis system stacking model based on Xgboost neural network |
| CN111145912A (en) * | 2019-12-23 | 2020-05-12 | 浙江大学 | Machine learning-based prediction device for personalized ovulation promotion scheme |
| CN111755122A (en) * | 2020-05-21 | 2020-10-09 | 甘肃卫生职业学院 | Diabetes blood sugar prediction system and method based on CNN and model fusion |
| CN111749675A (en) * | 2020-05-25 | 2020-10-09 | 中国地质大学（武汉） | Stratum drillability prediction method and system based on cascade model algorithm |
| CN112035582A (en) * | 2020-08-28 | 2020-12-04 | 光大科技有限公司 | Structured data classification method and device, storage medium and electronic device |
| CN112185555A (en) * | 2020-09-10 | 2021-01-05 | 北京工业大学 | Gestational diabetes prediction method based on stacking algorithm |
| CN113205111A (en) * | 2021-04-07 | 2021-08-03 | 零氪智慧医疗科技(天津)有限公司 | Identification method and device suitable for liver tumor and electronic equipment |
Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106446533A (en) * | 2016-09-12 | 2017-02-22 | 北京和信康科技有限公司 | Processing system of human body health data and method thereof |
| CN106446777A (en) * | 2016-08-29 | 2017-02-22 | 广东工业大学 | Noninvasive blood sugar data processing method and noninvasive blood sugar data processing system based on convolutional neural network |
| CN106682412A (en) * | 2016-12-22 | 2017-05-17 | 浙江大学 | Diabetes prediction method based on medical examination data |
| CN106934798A (en) * | 2017-02-20 | 2017-07-07 | 苏州体素信息科技有限公司 | Diabetic retinopathy classification stage division based on deep learning |
| CN106980746A (en) * | 2016-12-16 | 2017-07-25 | 清华大学 | A kind of general Woundless blood sugar Forecasting Methodology based on Time-Series analysis |
| CN107180155A (en) * | 2017-04-17 | 2017-09-19 | 中国科学院计算技术研究所 | A kind of disease forecasting method and system based on Manufacturing resource model |
| CN107358014A (en) * | 2016-11-02 | 2017-11-17 | 华南师范大学 | The clinical pre-treating method and system of a kind of physiological data |
| CN107358605A (en) * | 2017-05-04 | 2017-11-17 | 深圳硅基智能科技有限公司 | For identifying the deep neural network and system of diabetic retinopathy |
| CN107403072A (en) * | 2017-08-07 | 2017-11-28 | 北京工业大学 | A kind of diabetes B prediction and warning method based on machine 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 |
| CN107578822A (en) * | 2017-07-25 | 2018-01-12 | 广东工业大学 | A kind of pretreatment and feature extracting method for the multi-modal big data of medical treatment |
| CN107680676A (en) * | 2017-09-26 | 2018-02-09 | 电子科技大学 | A kind of gestational diabetes Forecasting Methodology based on electronic health record data-driven |
| CN108108757A (en) * | 2017-12-18 | 2018-06-01 | 深圳市唯特视科技有限公司 | A kind of diabetic foot ulcers sorting technique based on convolutional neural networks |
| CN108198625A (en) * | 2016-12-08 | 2018-06-22 | 北京推想科技有限公司 | A kind of deep learning method and apparatus for analyzing higher-dimension medical data |
-
2018
- 2018-09-17 CN CN201811079861.8A patent/CN109448855A/en active Pending
Patent Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106446777A (en) * | 2016-08-29 | 2017-02-22 | 广东工业大学 | Noninvasive blood sugar data processing method and noninvasive blood sugar data processing system based on convolutional neural network |
| CN106446533A (en) * | 2016-09-12 | 2017-02-22 | 北京和信康科技有限公司 | Processing system of human body health data and method thereof |
| CN107358014A (en) * | 2016-11-02 | 2017-11-17 | 华南师范大学 | The clinical pre-treating method and system of a kind of physiological data |
| CN108198625A (en) * | 2016-12-08 | 2018-06-22 | 北京推想科技有限公司 | A kind of deep learning method and apparatus for analyzing higher-dimension medical data |
| CN106980746A (en) * | 2016-12-16 | 2017-07-25 | 清华大学 | A kind of general Woundless blood sugar Forecasting Methodology based on Time-Series analysis |
| CN106682412A (en) * | 2016-12-22 | 2017-05-17 | 浙江大学 | Diabetes prediction method based on medical examination data |
| CN106934798A (en) * | 2017-02-20 | 2017-07-07 | 苏州体素信息科技有限公司 | Diabetic retinopathy classification stage division based on deep learning |
| CN107180155A (en) * | 2017-04-17 | 2017-09-19 | 中国科学院计算技术研究所 | A kind of disease forecasting method and system based on Manufacturing resource model |
| CN107358605A (en) * | 2017-05-04 | 2017-11-17 | 深圳硅基智能科技有限公司 | For identifying the deep neural network and system of diabetic retinopathy |
| CN107578822A (en) * | 2017-07-25 | 2018-01-12 | 广东工业大学 | A kind of pretreatment and feature extracting method for the multi-modal big data of medical treatment |
| CN107403072A (en) * | 2017-08-07 | 2017-11-28 | 北京工业大学 | A kind of diabetes B prediction and warning method based on machine 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 |
| CN107680676A (en) * | 2017-09-26 | 2018-02-09 | 电子科技大学 | A kind of gestational diabetes Forecasting Methodology based on electronic health record data-driven |
| CN108108757A (en) * | 2017-12-18 | 2018-06-01 | 深圳市唯特视科技有限公司 | A kind of diabetic foot ulcers sorting technique based on convolutional neural networks |
Non-Patent Citations (4)
| Title |
|---|
| BO JIN 等: "Predicting the Risk of Heart Failure With EHR Sequential Data Modeling", 《SPECIAL SECTION ON RECENT COMPUTATIONAL METHODS IN KNOWLEDGE ENGINEERING AND INTELLIGENCE COMPUTATION》 * |
| MIN CHEN等: "Disease Prediction by Machine Learning Over Big Data From Healthcare Communities", 《SPECIAL SECTION ON HEALTHCARE BIG DATA》 * |
| SABA BASHIR 等: "An Efficient Rule-based Classification of Diabetes Using ID3, C4.5 & CART Ensembles", 《2014 12TH INTERNATIONAL CONFERENCE ON FRONTIERS OF INFORMATION TECHNOLOGY》 * |
| 周志华: "《机器学习》", 31 January 2016, 清华大学出版社 * |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110197706A (en) * | 2019-04-26 | 2019-09-03 | 深圳市宁远科技股份有限公司 | A kind of stratification feature selection approach, system and application based on SBS |
| CN110197706B (en) * | 2019-04-26 | 2021-08-27 | 深圳市宁远科技股份有限公司 | Hierarchical feature selection method, system and application based on SBS |
| CN110120265A (en) * | 2019-04-29 | 2019-08-13 | 天津大学 | The method of raising prediction blood uric acid precision based on multidimensional characteristic and Model Fusion |
| CN110120265B (en) * | 2019-04-29 | 2023-03-31 | 天津大学 | Method for improving blood uric acid prediction precision based on multi-dimensional feature and model fusion |
| CN110246577B (en) * | 2019-05-31 | 2021-04-30 | 深圳江行联加智能科技有限公司 | Method for assisting gestational diabetes genetic risk prediction based on artificial intelligence |
| CN110246577A (en) * | 2019-05-31 | 2019-09-17 | 深圳江行联加智能科技有限公司 | A method of based on artificial intelligence auxiliary gestational diabetes genetic risk prediction |
| CN110289097A (en) * | 2019-07-02 | 2019-09-27 | 重庆大学 | A kind of Pattern Recognition Diagnosis system stacking model based on Xgboost neural network |
| CN111145912A (en) * | 2019-12-23 | 2020-05-12 | 浙江大学 | Machine learning-based prediction device for personalized ovulation promotion scheme |
| CN111145912B (en) * | 2019-12-23 | 2023-04-18 | 浙江大学 | Machine learning-based prediction device for personalized ovulation promotion scheme |
| CN111755122A (en) * | 2020-05-21 | 2020-10-09 | 甘肃卫生职业学院 | Diabetes blood sugar prediction system and method based on CNN and model fusion |
| CN111749675A (en) * | 2020-05-25 | 2020-10-09 | 中国地质大学（武汉） | Stratum drillability prediction method and system based on cascade model algorithm |
| CN112035582A (en) * | 2020-08-28 | 2020-12-04 | 光大科技有限公司 | Structured data classification method and device, storage medium and electronic device |
| CN112185555A (en) * | 2020-09-10 | 2021-01-05 | 北京工业大学 | Gestational diabetes prediction method based on stacking algorithm |
| CN113205111A (en) * | 2021-04-07 | 2021-08-03 | 零氪智慧医疗科技(天津)有限公司 | Identification method and device suitable for liver tumor and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109448855A (en) | A kind of diabetes glucose prediction technique based on CNN and Model Fusion | |
| CN109920547A (en) | A kind of diabetes prediction model construction method based on electronic health record data mining | |
| CN106202955A (en) | Diagnosis associated packets method and system based on intellectual coded adaptation | |
| Chen et al. | Generative consistency for semi-supervised cerebrovascular segmentation from TOF-MRA | |
| CN116051574A (en) | Semi-supervised segmentation model construction and image analysis method, device and system | |
| Chen et al. | SDFNet: Automatic segmentation of kidney ultrasound images using multi-scale low-level structural feature | |
| Rajput et al. | An accurate and noninvasive skin cancer screening based on imaging technique | |
| CN114881968A (en) | OCTA image vessel segmentation method, device and medium based on deep convolutional neural network | |
| CN116386860A (en) | Diabetes and complications intelligent auxiliary prediction and diagnosis platform based on multiple modes | |
| CN116013543A (en) | TACE curative effect prediction method based on deep learning | |
| ZongRen et al. | DenseTrans: multimodal brain tumor segmentation using swin transformer | |
| CN114822874A (en) | Prescription efficacy classification method based on characteristic deviation alignment | |
| Palaniswamy et al. | Internet of Things and Deep Learning Enabled Diabetic Retinopathy Diagnosis Using Retinal Fundus Images | |
| Ding et al. | Efficient Unet with depth-aware gated fusion for automatic skin lesion segmentation | |
| Huang et al. | ADDNS: An asymmetric dual deep network with sharing mechanism for medical image fusion of CT and MR-T2 | |
| Yu et al. | M3U-CDVAE: Lightweight retinal vessel segmentation and refinement network | |
| Chowdary et al. | Multiple Disease Prediction by Applying Machine Learning and Deep Learning Algorithms | |
| Rajan et al. | Comparative Analysis of Liver diseases by using Machine Learning Techniques | |
| CN115619810B (en) | Prostate partition segmentation method, system and equipment | |
| Yan et al. | MRSNet: Joint consistent optic disc and cup segmentation based on large kernel residual convolutional attention and self-attention | |
| Ma et al. | ConvLSTM coordinated longitudinal transformer under spatio-temporal features for tumor growth prediction | |
| Liu et al. | A gastric cancer recognition algorithm on gastric pathological sections based on multistage attention‐DenseNet | |
| Kaushal et al. | Eye Disease Detection Through Image Classification Using Federated Learning | |
| CN108961171A (en) | A kind of mammary gland DTI image de-noising method | |
| Feng et al. | A drug information embedding method based on graph convolution neural network |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20190308 |
|
| WD01 | Invention patent application deemed withdrawn after publication |