CN109659031A - Lung function index prediction meanss and determining method - Google Patents
Lung function index prediction meanss and determining method Download PDFInfo
- Publication number
- CN109659031A CN109659031A CN201811413177.9A CN201811413177A CN109659031A CN 109659031 A CN109659031 A CN 109659031A CN 201811413177 A CN201811413177 A CN 201811413177A CN 109659031 A CN109659031 A CN 109659031A
- Authority
- CN
- China
- Prior art keywords
- lung function
- network
- function index
- prediction
- prediction model
- 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
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
- G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
Landscapes
- Health & Medical Sciences (AREA)
- Engineering & Computer Science (AREA)
- Medical Informatics (AREA)
- Public Health (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Pathology (AREA)
- Epidemiology (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The invention discloses a kind of lung function index prediction meanss and determine method, wherein the device includes: a lung function instrument, acquires incomplete breath signal；One processing system determines lung function index Prediction Parameters according to breath signal, and according to prediction model.A prediction model built in the processing system, the model are a kind of learning machine network structure that transfinites of integrated form, including n preposition sub-networks and a postposition export network.The method of determination includes: that lung function instrument acquires incomplete breath signal, as original signal imput process system；Original signal characteristic is converted to form new data characteristics as input feature vector value；Establish input feature vector matrix；It trains and establishes prediction model；External data input prediction model simultaneously exports lung function index through operation.Lung function index prediction meanss proposed by the present invention and determining method, based on transfiniting, learning machine network structure is realized, can utilize incomplete breath signal Accurate Prediction lung function index parameter value.
Description
Technical field
The present invention relates to Medical Devices and bio-signal acquisition fields, and in particular to a kind of lung function index prediction meanss and
Determine method.
Background technique
Chronic Obstructive Pulmonary Disease (COPD) is common pulmonary disease.According to statistics, the global cause of death in the lethal residence COPD
4th.Research shows that the diagnosing and treating ahead of time of COPD is to reduce the important means of the death rate, forced vital capacity test is had become
For the important way for assessing and diagnosing COPD, key index therein includes forced vital capacity (FVC), forced vital capacity of the first second
(FEV1) and one second rate (FEV1/FVC).If FEV1/FVC is less than normal, FEV1/FVC < 0.7 under normal conditions, and FVC is just
Often, it is determined as obstructive dysfunction of pulmonary ventilation.As it can be seen that FVC is a very crucial index in assessment pulmonary ventilation function.In order to
Accurate FVC is obtained, ATS/ERS forced vital capacity tests quality control and receives the suction that standard suggests that tester must first slowly
Then sufficient gas is exerted oneself, quick exhaled gas, and breathe out the sufficiently long time and put down on capacity time graph with reaching
Platform, it is believed that tester has reached the standard that forced vital capacity test expiration terminates, and the key of the test, which is that, to need to test
Person's expiratory duration long enough to guarantee breathe out all gas, but this for many people come it is extremely difficult, such as it is some on
The old man at age or the patient being damaged with severe lung, especially serious chronic obstructive pulmonary disease patient are to be unable to complete entire lung function to survey
Examination, in addition for some without trained testers, not due to forced vital capacity test mode and usual expiration mode
Together, many testers, which are difficult to, meets testing standard.Above-mentioned reason causes test result that can not react the true patient's condition, is easy
The mistaken diagnosis for causing tuberculosis, is failed to pinpoint a disease in diagnosis.Therefore incomplete data Accurate Prediction FVC value is utilized, for examining for the related diseases such as chronic obstructive pulmonary disease
Disconnected screening is most important.
Currently, the lung function parameter prediction technique developed not yet based on portable turbine lung function instrument, turbine type lung
Function instrument pushes turbine rotation by exhaling, and the revolving speed of different expiration phase turbines is different, and therefore, breath signal has very strong
Timing.
Summary of the invention
The present invention provides one kind to be based on portable turbine lung function instrument, utilizes the prediction of the FVC of incomplete measurement data
Model and method propose a kind of lung function index prediction meanss and determining method, especially to solve incomplete expiratory gas flow
The problem of measurement data inaccuracy.
In order to achieve the above objectives, the present invention provides a kind of lung function index prediction meanss and determining methods.
Wherein, which includes:
One lung function instrument acquires incomplete breath signal；
One processing system determines lung function index Prediction Parameters according to the breath signal, and according to prediction model.
Further, processing system prestores a prediction model, which is a kind of learning machine net that transfinites of integrated form
Network structure, including n preposition sub-networks and a postposition export network, realize the prediction of incomplete breath signal.
Further, n preposition sub-networks are that the n independent learning machine networks that transfinite work side by side, a postposition output
Network is the learning machine network that transfinites.
Further, the quantity n of preposition sub-network is according to each data volume for testing acquisition and combination and the data volume phase
Depending on the function of pass.
Further, input feature vector of the output of preposition sub-network as postposition output network, it is defeated which exports network
Lung function Prediction Parameters out.
The present invention also provides a kind of lung function index to determine method, comprising:
Lung function instrument acquires incomplete breath signal, as original signal imput process system；
Original signal characteristic is converted to form new data characteristics as input feature vector value；
Establish input feature vector matrix；
It trains and establishes prediction model；
External data input prediction model simultaneously exports lung function index through operation.
Further, original signal characteristic is converted and forms new data characteristics and includes:
It is proposed a kind of airway resistance function based on gaussian probability distribution: Wherein a0, μ, σ are Constant Model parameter, are acquired by practical calibration, and V (i) indicates non-complete
I moment turbine speed in whole breath signal collection process；
Driving pressure: P (i)=E (FVC-kV (i)), wherein E indicates that driving constant, k indicate conversion constant, pass through calibration
It obtains, FVC indicates the incomplete breath signal parameter measured；
F (i), which is acquired, using f (i)=E (FVC-kV (i))/R (i) establishes input feature vector as input feature vector value, and with this
Matrix.
Further, input feature vector matrix is established are as follows:
Wherein, n indicates characteristic, m representative sample number, and each sample is that there are two the two-dimensional array (f of element for a toolm
(n), n).
Further, it trains and establishes prediction model and include:
Extract matrix FmnIn each column construct the input feature vector matrix of each preposition sub-network of the prediction model, input value
For fm(i) and i, wherein 1≤i≤n, m=1,2 ..., m acquire the defeated of each preposition sub-network using the learning machine principle that transfinites
Matrix out, related coefficient are generated by computer random；
Input feature vector by the output of preposition sub-network as postposition output network, and in conjunction with lung function index, again with
The learning machine principle that transfinites acquires the output weight of postposition output network, exports weight with this and establishes prediction model.
Further, it trains and establishes prediction model further include:
The model of different performance is obtained by adjusting the number for the learning machine network hidden node that individually transfinites in prediction model.
Lung function index prediction meanss proposed by the present invention and determining method, based on transfiniting, learning machine network structure is realized,
It can be realized and utilize incomplete breath signal Accurate Prediction lung function index parameter value.
Detailed description of the invention
Fig. 1 is that the integrated form of the embodiment of the present invention transfinites learning machine network structure；
Fig. 2 is the ELM network of the embodiment of the present invention.
Specific embodiment
To make the objectives, technical solutions, and advantages of the present invention clearer, below in conjunction with specific embodiment, and reference
Attached drawing, the present invention is described in further detail.
One embodiment of the invention provides a kind of lung function index prediction meanss, which includes:
One lung function instrument acquires complete or incomplete breath signal；
One processing system determines lung function index Prediction Parameters according to the breath signal, and according to prediction model.
In some embodiments, processing system prestores a prediction model, which is a kind of study of transfiniting of integrated form
Machine network structure, including n preposition sub-networks and a postposition export network, realize the prediction of incomplete breath signal.
In some embodiments, n preposition sub-networks are that the n independent learning machine networks that transfinite work side by side, a postposition
Output network is the learning machine network that transfinites.
In some embodiments, the quantity n of preposition sub-network is according to each data volume for testing acquisition and combination and the data
Depending on measuring relevant function.
In some embodiments, input feature vector of the output of preposition sub-network as postposition output network, the postposition exports net
Network exports lung function Prediction Parameters.
In the present embodiment, a kind of ELM network structure of integrated form is established based on the portable pulmonary function detection device of turbine type,
It is divided into preposition sub-network and postposition output network, preposition sub-network quantity is depending on the data volume that each test obtains.It is all
The output of preposition sub-network exports the input feature vector of network as postposition, the lung function parameter to be predicted, such as FVC value, as
The output of postposition network, by using instrument acquisition incomplete breath signal, and correctly lung function index to model into
The certain training of row, solving model parameter can set up prediction model, and then may be implemented real using incomplete breath signal
Existing lung function parameter prediction.
Another embodiment of the present invention additionally provides a kind of lung function index and determines method, comprising:
Lung function instrument acquires complete or incomplete breath signal, as original signal imput process system；
Original signal characteristic is converted to form new data characteristics as input feature vector value；
Establish input feature vector matrix；
It trains and establishes prediction model；
External data input prediction model simultaneously exports lung function index through operation.
In some embodiments, original signal characteristic is converted to form new data characteristics and include:
It is proposed a kind of airway resistance function based on gaussian probability distribution: Wherein a0, μ, σ are Constant Model parameter, are acquired by practical calibration, and V (i) indicates non-complete
I moment turbine speed in whole breath signal collection process；
Driving pressure: P (i)=E (FVC-kV (i)), wherein E indicates that driving constant, k indicate conversion constant, pass through calibration
It obtains, FVC indicates the incomplete breath signal parameter measured；
F (i), which is acquired, using f (i)=E (FVC-kV (i))/R (i) establishes input feature vector as input feature vector value, and with this
Matrix.
In the present embodiment, the integrated form lung function parameter based on the portable pulmonary function detection instrument of turbine type predicts network, to lung
The collected original signal of function instrument carries out feature pretreatment first, according to airway resistance Variation Features in exhalation process, proposes
A kind of airway resistance function based on gaussian probability distribution, it is specific to quantify the variation of the airway resistance in exhalation process
Expression formula are as follows:
Wherein a0, μ, σ are Constant Model parameter, are acquired by practical calibration.V (i) indicates i moment turbine speed.
Driving pressure P (i),
P (i)=E (FVC-kV (i)) (2)
E indicates that driving constant, k indicate conversion constant, can be obtained by calibration.
F (i)=E (FVC-kV (i))/R (i) (3)
It regard obtained f (i) as input feature vector value, establishes input feature vector matrix.
In some embodiments, input feature vector matrix is established are as follows:
Wherein, n indicates characteristic, m representative sample number, and each sample is that there are two the two-dimensional array (f of element for a toolm
(n), n).
In the present embodiment, according to treated, original signal establishes input matrix,
N indicates characteristic, m representative sample number, it can be seen that each sample is that there are two the two-dimensional arrays of element for a tool
(fm(n), n).
In some embodiments, training simultaneously establishes prediction model.
In the present embodiment, referring to Figure 1, since input feature vector has temporal characteristics, i.e. fm(n) strictly right with time value n
It answers, so needing to construct the network of an integrated form to embody the temporal characteristics of input data.As can be seen from Figure 1 this is integrated
Network is made of the four identic learning machine that transfinites (Extreme Learning Machine, ELM) sub-networks, transfinite
Habit machine principle is as follows:
Fig. 2 is referred to, is single ELM network, it is assumed that have N number of arbitrary sample (xj, tj), wherein
xj=[xj1, xj2... xin]T∈Rn
tj=[tj1, tj2... tjm]T∈Rm
There is the neural networks with single hidden layer of L hidden node that can be expressed as one
Wherein, g (x) is activation primitive, ωi=[wil, wi2... win]TIt is the input weight of i-th of Hidden unit, biIt is
The biasing of i-th of Hidden unit, βi=[βi1, βi2... βim]TIt is the output weight of i-th of Hidden unit.ωi·xjIndicate ωi
And xjInner product, the target of neural networks with single hidden layer study can be expressed as so that the error of output is minimum
There is ωi、xjAnd biSo that:
It can be indicated with matrix:
H β=T
Wherein, H is the output of hidden node, and β is output weight, and T is desired output.
In order to training neural networks with single hidden layer, it is intended that obtainSo that
Wherein, i=1,2 ..., L this are equivalent to minimize loss function
Traditional some algorithms based on gradient descent method can be used to solve such problems, but it is basic based on
The learning algorithm needs of gradient adjust all parameters during iteration.And in ELM algorithm, once input weight ωiWith
The biasing b of hidden layeriIt is determined at random, the output matrix H of hidden layer is just now uniquely determined.Training neural networks with single hidden layer can convert
To solve a linear system: H β=T.And exporting weight can be determined
It is the Moore-Penrose generalized inverse matrix of matrix H.
In some embodiments, training simultaneously establishes prediction model.
Firstly, extracting matrix FmnIn each column construct the input feature vector matrix of each preposition sub-network of the prediction model,
Input value is fm(i) and i, wherein 1≤i≤n, m=1,2 ..., m acquire each preposition subnet using the learning machine principle that transfinites
The output matrix of network, related coefficient are generated by computer random.
In the present embodiment, for the integrated network model of Fig. 1, it is assumed that shared n preposition sub-networks, each sub-network hidden layer
Neuron number be that (note: the neuron number of each sub-network can be the same or different L1, herein for convenience
It calculates, it is assumed that identical.)
It (1) is that each preposition sub-network generates one group of (w at random first with computeri, bi, βi)p, p=1,2,3...n,
Activation primitive selectionAnd extract matrix fmnIn each column construct the input of each sub-network
Eigenmatrix fn.By taking first sub-network as an example, extraction f firstmnIn first row:
Computer random generates coefficient matrix:
Bias matrix:
Coefficient matrix:
(2) each hidden node output matrix H is acquiredn.For solving first preposition sub-network
Seek the output matrix y of each preposition sub-networkn, for solving first preposition sub-network:
Secondly, the input feature vector by the output of preposition sub-network as postposition output network, and combine correct lung function
Index is acquired the output weight of postposition output network with the learning machine principle that transfinites again, exports weight with this and establish prediction model.
In the present embodiment, the output matrix of n preposition sub-networks is established:
(3) assume that postposition network hidden node number is L2, to postposition network, using y as input feature vector, repeat step (1)
(2) postposition network hidden node output matrix is established:
Seek HbackBroad sense Moore-Penrose generalized inverse matrix
Export weightWherein
Thus model is established.
In some embodiments, prediction model is established further include:
The model of different performance is obtained by adjusting the number for the learning machine network hidden node that individually transfinites in prediction model.
In the present embodiment, model solution process is completed model solution according to the learning machine principle that transfinites and is established, and the model is logical
Cross the model for adjusting the number acquisition different performance for the learning machine sub-network hidden node that transfinites.
Particular embodiments described above has carried out further in detail the purpose of the present invention, technical scheme and beneficial effects
Describe in detail bright, it should be understood that the above is only a specific embodiment of the present invention, is not intended to restrict the invention, it is all
Within the spirit and principles in the present invention, any modification, equivalent substitution, improvement and etc. done should be included in protection of the invention
Within the scope of.
Claims (10)
1. a kind of lung function index prediction meanss characterized by comprising
One lung function instrument acquires incomplete breath signal；
One processing system determines lung function index Prediction Parameters according to the breath signal, and according to prediction model.
2. lung function index prediction meanss according to claim 1, which is characterized in that the processing system prestores a prediction
Model, the prediction model are the learning machine network structure that transfinites of integrated form, including n preposition sub-networks and a postposition output
Network realizes the prediction of incomplete breath signal.
3. lung function index prediction meanss according to claim 2, which is characterized in that the n preposition sub-networks are n
The independent learning machine network that transfinites works side by side, and the postposition output network is the learning machine network that transfinites.
4. lung function index prediction meanss according to claim 2 or 3, which is characterized in that the number of the preposition sub-network
Measure n according to every time test acquisition data volume and in conjunction with function relevant to the data volume depending on.
5. lung function index prediction meanss according to claim 2, which is characterized in that the output of the preposition sub-network is made
The input feature vector of network is exported for the postposition, the postposition output network exports lung function index Prediction Parameters.
6. a kind of lung function index determines method characterized by comprising
Lung function instrument acquires incomplete breath signal, as original signal imput process system；
Original signal characteristic is converted to form new data characteristics as input feature vector value；
Establish input feature vector matrix；
It trains and establishes prediction model；
External data input prediction model simultaneously exports lung function index through operation.
7. lung function index determines method according to claim 6, which is characterized in that described to convert shape to original signal characteristic
The data characteristics of Cheng Xin includes:
It is proposed a kind of airway resistance function based on gaussian probability distribution:
Wherein a0, μ, σ are Constant Model parameter, and V (i) indicates i moment turbine rotary frequency in the incomplete breath signal collection process
Rate；
Driving pressure: P (i)=E (FVC-kV (i)), wherein E indicates driving constant, and k indicates conversion constant, by demarcating
It arrives, FVC indicates the incomplete breath signal parameter measured；
F (i) is acquired as input feature vector value using f (i)=E (FVC-kV (i))/R (i), and input feature vector matrix is established with this.
8. lung function index according to claim 7 determines method, which is characterized in that described to establish input feature vector matrix
Are as follows:
Wherein, n indicates characteristic, m representative sample number, and each sample is that there are two the two-dimensional array (f of element for a toolm(n),
n)。
9. lung function index according to claim 6 determines method, which is characterized in that the training simultaneously establishes prediction model
Include:
Extract matrix FmnIn each column construct the input feature vector matrix of each preposition sub-network of the prediction model, input value is
fm(i) and i, wherein 1≤i≤n, m=1,2 ..., m acquire the output of each preposition sub-network using the learning machine principle that transfinites
Matrix, related coefficient are generated by computer random；
Input feature vector by the output of preposition sub-network as postposition output network, and lung function index is combined, again to transfinite
Learning machine principle acquires the output weight of postposition output network, exports weight with this and establishes prediction model.
10. lung function index according to claim 6 or 9 determines method, which is characterized in that the training simultaneously establishes prediction
Model includes:
The model of different performance is obtained by adjusting the number for the learning machine network hidden node that individually transfinites in the prediction model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811413177.9A CN109659031B (en) | 2018-11-23 | 2018-11-23 | Lung function index prediction device and determination method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811413177.9A CN109659031B (en) | 2018-11-23 | 2018-11-23 | Lung function index prediction device and determination method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109659031A true CN109659031A (en) | 2019-04-19 |
| CN109659031B CN109659031B (en) | 2023-05-09 |
Family
ID=66112424
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811413177.9A Active CN109659031B (en) | 2018-11-23 | 2018-11-23 | Lung function index prediction device and determination method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109659031B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112071424A (en) * | 2020-08-31 | 2020-12-11 | 吾征智能技术(北京)有限公司 | Construction method and system of disease prediction model based on vital capacity data |
| CN113469227A (en) * | 2021-06-18 | 2021-10-01 | 南京润楠医疗电子研究院有限公司 | Forced expiration total amount prediction method |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050284476A1 (en) * | 2004-06-24 | 2005-12-29 | Blanch Paul B | Method and apparatus for non-invasive prediction of intrinsic positive end-expiratory pressure (PEEPi) in patients receiving ventilator support |
| WO2013186634A2 (en) * | 2012-06-14 | 2013-12-19 | Singapore Health Services Pte Ltd. | Predicting acute cardiopulmonary events and survivability of a patient |
| CN106897570A (en) * | 2017-03-02 | 2017-06-27 | 山东师范大学 | A kind of COPD test system based on machine learning |
| CN108109697A (en) * | 2017-12-20 | 2018-06-01 | 中国科学院合肥物质科学研究院 | A kind of pulmonary function test (pft) system and method based on expiration mathematical model |
-
2018
- 2018-11-23 CN CN201811413177.9A patent/CN109659031B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050284476A1 (en) * | 2004-06-24 | 2005-12-29 | Blanch Paul B | Method and apparatus for non-invasive prediction of intrinsic positive end-expiratory pressure (PEEPi) in patients receiving ventilator support |
| WO2013186634A2 (en) * | 2012-06-14 | 2013-12-19 | Singapore Health Services Pte Ltd. | Predicting acute cardiopulmonary events and survivability of a patient |
| CN106897570A (en) * | 2017-03-02 | 2017-06-27 | 山东师范大学 | A kind of COPD test system based on machine learning |
| CN108109697A (en) * | 2017-12-20 | 2018-06-01 | 中国科学院合肥物质科学研究院 | A kind of pulmonary function test (pft) system and method based on expiration mathematical model |
Non-Patent Citations (1)
| Title |
|---|
| 郑小禄等: "基于超限学习机的通信网络弹性预测方法", 《通信技术》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112071424A (en) * | 2020-08-31 | 2020-12-11 | 吾征智能技术(北京)有限公司 | Construction method and system of disease prediction model based on vital capacity data |
| CN113469227A (en) * | 2021-06-18 | 2021-10-01 | 南京润楠医疗电子研究院有限公司 | Forced expiration total amount prediction method |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109659031B (en) | 2023-05-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP2023164839A (en) | Method for analysis of cough sound using disease signature to diagnose respiratory disease | |
| Bellet et al. | The 6-minute walk test in outpatient cardiac rehabilitation: validity, reliability and responsiveness—a systematic review | |
| Mori et al. | Colored 3-dimensional analyses of respiratory resistance and reactance in COPD and asthma | |
| Enright et al. | Quality of spirometry test performance in children and adolescents*: experience in a large field study | |
| CN106897570A (en) | A kind of COPD test system based on machine learning | |
| CN108670200A (en) | A kind of sleep sound of snoring classification and Detection method and system based on deep learning | |
| CN101478912A (en) | Systems and methods for analyzing and assessing dementia and dementia -type disorders | |
| CN109659031A (en) | Lung function index prediction meanss and determining method | |
| KR102477592B1 (en) | Sarcopenia artificial intelligence diagnosis system using body fluid samples and method thereof | |
| Veezhinathan et al. | Detection of obstructive respiratory abnormality using flow–volume spirometry and radial basis function neural networks | |
| CN106572813A (en) | Pattern recognition system for quantifying the likelihood of the contribution of multiple possible forms of chronic disease to patient reported dyspnea | |
| CN111973188A (en) | Method for estimating respiratory mechanics parameter based on neural network | |
| Dinh et al. | Facial soft tissue thicknesses prediction using anthropometric distances | |
| Agarwal et al. | Design and development of a low-cost spirometer with an embedded web server | |
| CN108665066A (en) | Portable respirator flow curve scaling method based on BP neural network | |
| Ngo et al. | Blockchained federated learning for privacy and security preservation: Practical example of diagnosing cerebellar ataxia | |
| Mahesh et al. | Assessment and classification of normal and restrictive respiratory conditions through pulmonary function test and neural network | |
| CN111261264A (en) | Intelligent physique testing device and method and storage medium | |
| O'Donnell et al. | The flow-ratio index: an approach for measuring the influence of age and cigarette smoking on maximum expiratory flow-volume curve configuration | |
| Shen et al. | A computational approach to estimate postmortem interval using postmortem computed tomography of multiple tissues based on animal experiments | |
| Shehab et al. | Accurate Prediction of Pulmonary Fibrosis Progression Using EfficientNet and Quantile Regression: A High Performing Approach | |
| Tzelepis et al. | Expiratory effort enhancement and peak expiratory flow in humans | |
| Priyadharshini et al. | Automatic identification and classification of brain tumor using MRI images through transfer learning | |
| KUMAR et al. | Artificial neural network based approach for diagnosis of respiratory system using model based parameters of maximum expiratory flow-volume curve | |
| Li et al. | Study on Multi-frequency Characteristics of Lung Cancer Conductivity |
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 |