CN109585011A - The Illnesses Diagnoses method and machine readable storage medium of chest pain patients - Google Patents
The Illnesses Diagnoses method and machine readable storage medium of chest pain patients Download PDFInfo
- Publication number
- CN109585011A CN109585011A CN201811260651.9A CN201811260651A CN109585011A CN 109585011 A CN109585011 A CN 109585011A CN 201811260651 A CN201811260651 A CN 201811260651A CN 109585011 A CN109585011 A CN 109585011A
- Authority
- CN
- China
- Prior art keywords
- pectoralgia
- variable
- chest pain
- pain patients
- diagnostic 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.)
- 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
- G16H50/20—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
- Health & Medical Sciences (AREA)
- Engineering & Computer Science (AREA)
- Biomedical Technology (AREA)
- Medical Informatics (AREA)
- Public Health (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Pathology (AREA)
- Epidemiology (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Investigating Or Analysing Biological Materials (AREA)
Abstract
The embodiment of the present invention provides a kind of Illnesses Diagnoses method of chest pain patients, belongs to field of medical device.This method comprises: the chest pain patients are directed to, current signature variable needed for obtaining pectoralgia diagnostic model；And according to the current signature variable and the pectoralgia diagnostic model, judge whether the chest pain patients suffer from acute coronary syndrome or dissection of aorta.The Illnesses Diagnoses method and machine readable storage medium of the chest pain patients can accurately detect whether chest pain patients suffer from acute coronary syndrome or dissection of aorta, and spend less.
Description
Technical field
The present invention relates to Medical Devices, more particularly to the Illnesses Diagnoses method and machine readable storage of a kind of chest pain patients
Medium.
Background technique
A kind of complicated multiplicity of the reason of pectoralgia is that emergency treatment is common and the illness of threat to life, causes pectoralgia, including acute hat
Arteries and veins syndrome (Acute coronary syndrome, ACS), dissection of aorta (acute aortic dissection), lung
Embolism (PE), pneumothorax, pericarditis, pericardial tamponade and esophageal rupture etc., wherein ACS institute in the disease of these serious threat to life
Accounting example highest is 27.4%, and the misdiagnosis rate of myocardial infarction (AMI) is 3%~5%, the disease incidence of dissecting aneurysm of aorta
Sick for every about 0.5~1 human hair of 100,000 people, case fatality rate is up to 50% in 48h.American Heart Association reports this disease year hair for 2006
Sick rate is every about 25~30 human hair of 1,000,000 people disease, and the country there is no detailed statistics.Dissection of aorta seriously jeopardizes at these
Proportion 0.1% in the disease of life, but if its death rate of mistaken diagnosis is more than 90%.ACS and dissection of aorta symptom and the heart
Electrograph inspection performance is extremely similar, even if can not further differentiate that this is to lead for dissection of aorta echocardiography
Cause the big factor of the death rate increased one.Research finds no matter clinical to the research of acute chest pain both at home and abroad or basis aspect at present
Focus primarily upon the related fields with the detection of the markers such as d-dimer.There is not yet both at home and abroad to young people's patients with acute chest pain
Big data research, also have no the risk factor of China youth's crowd's acute chest pain and the risk stratification of young patients with acute chest pain
Method still relies on CT Aortography and MRI aortic blood for making a definite diagnosis for ACS and Aortic Dissection at present
Pipe is imaged or through esophagus Color Sonography.But the such property made a definite diagnosis detection means can not be conventional in emergency ward, community's basic hospital
It is equipped with, in addition inspection fee is expensive, can not carry out simply and easily bedside detection, this is also that Aortic Dissection mistaken diagnosis is dead
Die the main reason for rate is high.
Summary of the invention
The purpose of the embodiment of the present invention is that the Illnesses Diagnoses method and machine readable storage medium of a kind of chest pain patients are provided,
The Illnesses Diagnoses method and machine readable storage medium of the chest pain patients can accurately detect whether chest pain patients suffer from acute hat
Arteries and veins syndrome or dissection of aorta, and spend less.
To achieve the goals above, the embodiment of the present invention provides a kind of Illnesses Diagnoses method of chest pain patients, this method packet
It includes: for the chest pain patients, current signature variable needed for obtaining pectoralgia diagnostic model；And become according to the current signature
Amount and the pectoralgia diagnostic model, judge whether the chest pain patients suffer from acute coronary syndrome or dissection of aorta.
Preferably, the pectoralgia diagnostic model obtains in the following manner: spy needed for determining the pectoralgia diagnostic model
Levy variable；The historical data of characteristic variable needed for obtaining the pectoralgia diagnostic model and the history testing result of pectoralgia；And
Based on machine learning method, according to the history of the historical data of characteristic variable needed for the pectoralgia diagnostic model and the pectoralgia
Testing result establishes the pectoralgia diagnostic model.
Preferably, characteristic variable needed for the determination pectoralgia diagnostic model includes: acquisition change relevant to pectoralgia
Amount；The variable relevant to pectoralgia is normalized；The variable after normalization is sieved by recursive feature null method
Choosing obtains the first selection variables；And become using first selection variables as feature needed for the pectoralgia diagnostic model
Amount.
Preferably, characteristic variable needed for the determination pectoralgia diagnostic model includes: by random forests algorithm
Index MeanDecreaseAccuracy and MeanDecreaseGini are respectively to the importance of the variable relevant to pectoralgia
Screening and sequencing is carried out, the first sequence of importance sequence and the second sequence of importance sequence are obtained；Since importance highest, from institute
It states and takes out the variable work for being greater than default value and identical quantity in the first sequence of importance and second sequence of importance respectively
For the second selection variables and third selection variables, keep second selection variables identical with the third selection variables；It will be described
Second selection variables or the third selection variables are as characteristic variable needed for the pectoralgia diagnostic model.
Preferably, characteristic variable needed for the pectoralgia diagnostic model includes following human body indicators: plasma D-dimer content is surveyed
Fixed, the quantitative determination of troponin T, creatine kinase, creatine kinase isozyme, aspartate aminotransferase, urea, blood platelet meter
Number, glucose, creatinine, seralbumin, total protein, bilirubin direct and sodium.
Preferably, the variable relevant to pectoralgia be selected from the group comprising following human body indicators: gender, the age, urea,
Neutrophil leucocyte, gamma-glutamyl based transferase, lymphocyte, creatinine, monocyte, glucose, eosinophil, serum
Uric acid, basophilic granulocyte, creatine kinase, white blood cell count(WBC), lactic dehydrogenase, red blood cell count(RBC), calcium, hemoglobinometry,
Sodium, hematid specific volume measurement, potassium, mean corpuscular volume (MCV), chloride, platelet count, Phos, the blood red egg of mean corpuscular
Before Bai Liang, magnesium, mean corpuscular hemoglobin concentration (MCHC), troponin T, erythrocyte volume distribution width measurement CV, brain natriuretic peptide
When body, mean platelet volume measurement, creatine kinase isozyme quantitative determination, alanine aminotransferase, plasma prothrombin
Between measurement, aspartate aminotransferase, plasma fibrinogen measurement, total protein, plasma prothrombin mobility measurement, blood
Pure albumen, international standardization ratio, total bilirubin, thrombin time test, bilirubin direct, plasma D-dimer determination with
And alkaline phosphatase.
Preferably, after acquisition variable relevant to pectoralgia, this method further include: use multiple interpolating method pair
The variable relevant to pectoralgia carries out missing values and fills up.
Preferably, the machine learning algorithm includes: in logistic regression algorithm, random forests algorithm and SVM algorithm
At least one.
Preferably, this method further include: obtain the variable for having significant correlation with the diagnostic result of pectoralgia；Examine the chest
Whether characteristic variable needed for pain diagnostic model is the variable for having significant correlation with the diagnostic result of pectoralgia.
The embodiment of the present invention also provides a kind of machine readable storage medium, which is characterized in that the machine readable storage medium
On be stored with instruction, the instruction be used for so that machine executes the Illnesses Diagnoses method of chest pain patients described above.
Through the above technical solutions, using the Illnesses Diagnoses method and machine readable storage of chest pain patients provided by the invention
Medium, current signature variable needed for obtaining pectoralgia diagnostic model for chest pain patients, and chest is judged by pectoralgia diagnostic model
Whether pain patient's suffers from acute coronary syndrome or dissection of aorta, can accurately detect whether chest pain patients suffer from acute hat
Arteries and veins syndrome or dissection of aorta, and spend less.
The other feature and advantage of the embodiment of the present invention will the following detailed description will be given in the detailed implementation section.
Detailed description of the invention
Attached drawing is to further understand for providing to the embodiment of the present invention, and constitute part of specification, under
The specific embodiment in face is used to explain the present invention embodiment together, but does not constitute the limitation to the embodiment of the present invention.Attached
In figure:
Fig. 1 is the flow chart for the chest pain patients Illnesses Diagnoses method that one embodiment of the invention provides；
Fig. 2 is the preparation method of characteristic variable needed for the pectoralgia diagnostic model that one embodiment of the invention provides；
Fig. 3 be another embodiment of the present invention provides pectoralgia diagnostic model needed for characteristic variable preparation method；
The variable with MeanDecreaseAccuracy progress screening and sequencing that Fig. 4 A one embodiment of the invention provides shows
It is intended to；
The schematic diagram for the variable that screening and sequencing is carried out with MeanDecreaseGini that Fig. 4 B one embodiment of the invention provides；
Fig. 5 is the relation schematic diagram of the pectoralgia diagnostic model accuracy that one embodiment of the invention provides and variable quantity；
Fig. 6 is the flow chart of the preparation method for the pectoralgia diagnostic model that one embodiment of the invention provides；
Fig. 7 is the ROC curve diagram that one embodiment of the invention provides.
Specific embodiment
It is described in detail below in conjunction with specific embodiment of the attached drawing to the embodiment of the present invention.It should be understood that this
Locate described specific embodiment and be merely to illustrate and explain the present invention embodiment, is not intended to restrict the invention embodiment.
Fig. 1 is the flow chart for the chest pain patients Illnesses Diagnoses method that one embodiment of the invention provides.As shown in Figure 1, the party
Method includes:
Step S11, for the chest pain patients, current signature variable needed for obtaining pectoralgia diagnostic model；And
Whether step S12 judges the chest pain patients according to the current signature variable and the pectoralgia diagnostic model
With acute coronary syndrome or dissection of aorta.
In embodiments of the present invention, it is examined mainly for the acute coronary syndrome and dissection of aorta that cause pectoralgia
It is disconnected, first have to the acquisition that data information is carried out for chest pain patients.The data information of acquisition is mainly needed for pectoralgia diagnostic model
Current signature variable.In embodiments of the present invention, human body indicators that certain and pectoralgia of chest pain patients is relevant are as pectoralgia
Current signature variable needed for diagnostic model.Variable (that is, human body indicators) relevant to pectoralgia may include: white blood cell count(WBC),
The routine blood indexes such as neutrophil leucocyte, lymphocyte, the routine urinalysis index such as serum uric acid, urea, creatine kinase, creatinine, total egg
The blood biochemistry index such as white, seralbumin, thrombin time test, serum PT in patients, international standardization ratio,
The Blood Coagulations such as plasma prothrombin mobility measurement, the indexs such as d-dimer are 58 total.Then missing can be deleted
It is worth higher variable, for example, in embodiments of the present invention, characteristic variable 14 that missing values are higher than 25% are deleted, finally, with chest
Bitterly relevant variable is selected from the group comprising following human body indicators:
Gender (Gender), age (Age), urea (UREA), neutrophil leucocyte (NEUT), gamma-glutamyl based transferase
(γ-GGT), lymphocyte (LYM), creatinine (CRE), monocyte (MONO), glucose (GLU), eosinophil
(EOS), serum uric acid (SUA), basophilic granulocyte (BG), creatine kinase (CK), white blood cell count(WBC) (WBC), lactic dehydrogenase
(LDH), red blood cell count(RBC) (RBC), calcium (CA), hemoglobinometry (HGB), sodium (NA), hematid specific volume measure (HCT), potassium
(K), mean corpuscular volume (MCV) (MCV), chloride (CL), platelet count (BPC), Phos (PHOS), mean corpuscular blood
Lactoferrin amount (MCH), magnesium (MG), mean corpuscular hemoglobin concentration (MCHC) (MCHC), troponin T (CTnT), erythrocyte volume
The dispersion of distribution measures (CV), Natriuretic Peptide (NT-PROBNP), mean platelet volume measurement (MPV), the same work of creatine kinase
Enzyme quantitative determines (CKIQD), alanine aminotransferase (ALT), serum PT in patients (PT), aspartic acid amino
Transferase (AST), plasma fibrinogen measurement (FG), total protein (TP), plasma prothrombin mobility measure (PPAD), blood
Pure albumen (SA), international standardization ratio (INR), total bilirubin (TBIL), thrombin time test (TT), bilirubin direct
(DBIL), plasma D-dimer determination (DD) and alkaline phosphatase (ALP).
Wherein, plasma D-dimer determination (DD) is a specific fibrinolytic process markup object, as long as body is intravascular
The thrombosis and Fibrinolysis activity, d-dimer for having activation will increase.Myocardial infarction, pulmonary embolism, dissection of aorta, the heart
Colic pain, inflammation etc. can lead to d-dimer raising.Positive predictive value of the d-dimer in Pulmonary Embolism Patients is 20%, but yin
Property predicted value be 100%.Its sensibility illustrates a possibility that failing to pinpoint a disease in diagnosis pulmonary embolism very little up to 100%.In other words, d-dimer yin
Property patient can exclude pulmonary embolism substantially, and d-dimer positive patient needs row CTPA to further clarify.This plasma D-dimer content
The hypersensitivity of detection and high negative predictive value are also confirmed in other researchs.Myocardial infarction, angina pectoris, aorta clamp
When layer, d-dimer can also be increased, and d-dimer positive predictive value is respectively 37.8%, 13.3%, 6.7%, negative predictive value
Respectively 41.4%, 65.5%, 82.8%, illustrate that myocardial infarction, angina pectoris or active cannot be diagnosed as when the d-dimer positive
Arteries and veins interlayer cannot also exclude above-mentioned disease completely when negative, at this time need to further combined with electrocardiogram dynamic change, troponin,
The inspections such as myocardium enzyme, aorta CT further clarify pectoralgia reason.
Creatine kinase isozyme (CK-MB) to determine myocardial infarction time, area, send out myocardial infarction and Reperfu- sion again
Situation all has important value.The raising of CK-MB and 30 days and 6 months case fatality rate are significant related, and the increase of risk factors
Start from CK-MB raising.It may be the independent risk factor for predicting prognostic value that this conclusion prompt CK-MB, which is increased,.STEMI is examined
Disconnected is that benchmark is dynamically changed to typical electrocardiogram, and myocardial injury markers are as existing for the effect for confirming diagnosis；And
NSTEMI lacks the specificity of electrocardiogram, and myocardial injury markers play an important role.
White blood cell count(WBC) (WBC): being the independent risk factor of coronary heart disease, and the raising prompt myocardial ischemia of numerical value fills again
Note damage and the generation of endpoints, while there is significant correlativity with recent and late mortality rate.
Troponin T (CTnT): being the unique structural proteins of heart with cardiac troponin T and Troponin I, is cardiac muscle
The sensitivity of meronecrosis and special biochemical marker.Probably there is the patient of 23.5% dissection of aorta to have troponin liter
Height, the raised dissection of aorta patient of troponin are 4 times higher than patient's case fatality rate of troponin normal range (NR).Troponin I
It increases very common in the patient of dissection of aorta.Troponin I may indicate cardiac hemodynamic in dissection of aorta
Pressure is larger, but may not be able to reflect that prognosis is poor.Aorta and coronary artery interlayer can lead to troponin raising, on the one hand may be used
It can be the small infarct for causing cardiac muscle, on the other hand may be that coronary artery interlayer makes deficiency myocardial blood supply and cardiac muscle is caused to lack
Oxygen.
Neutrophil leucocyte (NEUT)/lymphocyte (LYM): occurrence and development process of the inflammatory reaction in atherosclerosis
In play an important role, while be also cause ST-Elevation Acute Myocardial Infarction (ST-segment elevation
Myocardial infarction, STEMI) myocardial ischemia/reperfusion injury (myocardial ischemia-
Reperfusion injury, MIRI) an important factor for, and leucocyte is one of key player of inflammatory reaction.Dialogue at present
The research of classified counting of leucocyte is more and more, there is that researches show that neutrophil leucocyte/lymphocyte ratio (neutrophil to
Lymphocyte ratio, NLR) there is important value to acute coronary syndrome diagnosis and prognosis.Myocardial ischemia-reperfusion damage
Wound is a complicated pathophysiological process, and neutrophil leucocyte passes through cell factor, complement, oxygen radical, protease, sticks
Each process of the participation reperfusion injury such as molecule.
Seralbumin (SA): the serum albumin levels of Acute Coronary Syndrome Patients are significantly reduced compared with healthy person, and
With Acute Coronary Syndrome Patients lesion degree aggravate serum albumin levels decrease, prompt serum albumin levels with
The acute coronary syndrome state of an illness is closely related, and acute coronary syndrome coincident with severity degree of condition is in patients serum's albumin level
It is negatively correlated.Existing studies have shown that seralbumin and dissection of aorta non-correlation.
Creatinine (CRE): since usually complicated hypertension, kidney are chronically at high perfusion state to dissection of aorta patient, and
Interlayer can often involve the arteria renalis, and controlled hypotension will lead to patient's arteria renalis blood supply reduction again while treatment, so as to cause AKI
Generation, serum creatinine level increase.
Platelet count (BPC): blood platelet is the important component for participating in blood coagulation.When dissection of aorta occurs, inner membrance tearing
Make subendothelial tissue exposure with the formation of false chamber, the release of Endothelin, thromboplastin and other factors leads to platelet activation, aggregation
And the formation of blood clot, meanwhile, the reduction of platelet count and extensive blood coagulation in vivo cause the consumption of blood platelet to have.A type is acute
Platelet count and Death risk are in significant negatively correlated when dissection of aorta patient is admitted to hospital.Platelet count when being admitted to hospital≤
The A type patients with acute aortic dissection hospital mortality of 119 × 109/L is significantly higher than blood platelet > 119 × 109/L patient.
In embodiments of the present invention, current signature needed for variable relevant to pectoralgia can be used as pectoralgia diagnostic model becomes
Amount.But in order to improve diagnosis efficiency, shorten the time that sufferer medical treatment for the first time touches treatment, the present invention also provides to above-mentioned change
Amount is screened in the method for characteristic variable needed for the more accurate pectoralgia diagnostic model of acquisition, specifically will be explained below.
After the characteristic variable needed for obtaining pectoralgia diagnostic model, pectoralgia diagnostic model is substituted them in, to make chest
Pain diagnostic model according to the chest pain patients that the variable of substitution obtains acquiring these variables suffer from whether with acute coronary syndrome or
The conclusion of dissection of aorta.
Fig. 2 is the preparation method of characteristic variable needed for the pectoralgia diagnostic model that one embodiment of the invention provides.Such as Fig. 2 institute
Show, this method comprises:
Step S21 obtains variable relevant to pectoralgia；
The variable relevant to pectoralgia is normalized in step S22；
Step S23 screens the variable after normalization by recursive feature null method to obtain the first selection variables；With
And
Step S24, using first selection variables as characteristic variable needed for the pectoralgia diagnostic model.
As described above, in addition to directly using human body indicators relevant to pectoralgia as feature needed for pectoralgia diagnostic model
Other than variable, it can also be screened using the method for the present embodiment.
In inventive embodiments, firstly, being carried out first to each index to eliminate the influence of dimension between different human body index
Normalized, such as normalized mode are the maximum value using the numerical value of certain index divided by the index, or use other sides
Formula.Then, feature selecting is carried out using recursive feature null method, recursive feature null method main thought is building model repeatedly
(such as Random Forest model) then selects the feature of best (or worst), the feature selecting elected is come out, then
This process is repeated in remaining feature, until all features all traverse.
For 46 human body indicators relevant to pectoralgia listed above, the first selection variables after completing feature selecting are
33, the precision of prediction highest of model at this time.First selection variables include: plasma D-dimer determination (DD), troponin T
(CTnT), creatine kinase (CK), creatine kinase isozyme quantitative determination (CKIQD), aspartate aminotransferase (AST), urine
Plain (UREA), platelet count (BPC), glucose (GLU), creatinine (CRE), seralbumin (SA), total protein (TP), directly
Bilirubin (DBIL), sodium (NA), total bilirubin (TBIL), erythrocyte volume distribution width measure (CV), international standardization ratio
(INR) etc..Characteristic variable needed for first selection variables can be used as pectoralgia diagnostic model.
Fig. 3 be another embodiment of the present invention provides pectoralgia diagnostic model needed for characteristic variable preparation method.Such as Fig. 3
It is shown, this method comprises:
Step S31 passes through the index MeanDecreaseAccuracy and MeanDecreaseGini of random forests algorithm
Screening and sequencing is carried out to the importance of the variable relevant to pectoralgia respectively, obtains the first sequence of importance sequence and the second weight
The property wanted generic sequence；
Step S32 divides from first sequence of importance and second sequence of importance since importance highest
The variable of default value and identical quantity Qu Chu be greater than as the second selection variables and third selection variables, make second sieve
Select variable identical with the third selection variables, and quantity is less than the quantity of first selection variables；
Step S33, using second selection variables or the third selection variables as needed for the pectoralgia diagnostic model
Characteristic variable.
In the present embodiment, first using random forests algorithm index MeanDecreaseAccuracy and
MeanDecreaseGini carries out screening and sequencing to the importance of variable relevant to pectoralgia described above respectively.
The sequence thinking of MeanDecreaseAccuracy are as follows: the value of a variable is become random number, random forest
The reduction degree of forecasting accuracy.The value is bigger, and the importance for indicating the variable is bigger.
The sequence thinking of MeanDecreaseGini are as follows: calculate each variable on each node of classification tree observation it is different
The influence of matter, thus the importance of comparison variable.The value is bigger, and the importance for indicating the variable is bigger.
It is sorted by the above method, obtains the first sequence of importance sequence and the second sequence of importance sequence.Such as Fig. 4 A-4B
It is shown, the signal for the variable that screening and sequencing is carried out with MeanDecreaseAccuracy that Fig. 4 A one embodiment of the invention provides
Figure；The schematic diagram for the variable that screening and sequencing is carried out with MeanDecreaseGini that Fig. 4 B one embodiment of the invention provides.
Fig. 5 is the relation schematic diagram of the pectoralgia diagnostic model accuracy that one embodiment of the invention provides and variable quantity.By
It is found that after variable quantity reaches 10, the accuracy of model tends to be steady Fig. 5, with increasing for variable number, has small
Ascendant trend.In order to be balanced between shortening consultation time and accuracy, need variable as few as possible still cannot
Accuracy is abandoned, it is therefore preferable that variable will at least take 10, to accomplish to guarantee that accuracy variation is little, naturally it is also possible to take
Other values.
Therefore, for the first sequence of importance sequence and the second sequence of importance sequence, since importance highest, respectively
It takes out and is greater than the variable of default value (10 or other values described above) and identical quantity as the second selection variables and the
Three selection variables keep the second selection variables identical with the third selection variables, and quantity is less than the quantity of the first selection variables
(quantity greater than the first selection variables is inaccurate instead).For the first sequence of importance sequence and the second sequence of importance sequence
Column, meet above-mentioned condition variable be first 13 since importance highest, i.e. the second selection variables and third selection variables,
Including following human body indicators: plasma D-dimer determination (DD), troponin T (CTnT), creatine kinase (CK), creatine kinase are same
Work enzyme quantitative determines (CKIQD), aspartate aminotransferase (AST), urea (UREA), platelet count (BPC), glucose
(GLU), creatinine (CRE), seralbumin (SA), total protein (TP), bilirubin direct (DBIL) and sodium (NA).
Second selection variables or third selection variables are identical, therefore feature needed for all can serve as pectoralgia diagnostic model becomes
Amount.And it is possible to see, the second selection variables or third selection variables are all contained in the first selection variables, the second screening
13 human body indicators of variable or third selection variables are more accurate.
In addition, the present invention can also verify characteristic variable needed for pectoralgia diagnostic model obtained above.Such as
Judge whether they with pectoralgia diagnostic result have significant correlation, specific as follows:
Using variable relevant to pectoralgia, variable test mould is constructed by multi-class logistic regression, SVM and random forest
Type, by five folding cross validations (data set is divided into 5 parts, in turn will wherein 4 parts be used as training data, 1 part as test number
According to, tested, be used to testing algorithm accuracy, be common test method) obtain the phase relation of spearman related coefficient
Numerical value (Corr value) and significance (P value), as shown in 1 data set basic condition of table and correlation analysis, according to can in table 1
Know, it is the age (Age), neutrophil leucocyte (NEUT), lymphocyte (LYM), monocyte (MONO), white blood cell count(WBC) (WBC), red
Cell count (RBC), hemoglobinometry (HGB), hematid specific volume measure (HCT), platelet count (BPC), phenylalanine ammonia
Based transferase (ALT), aspartate aminotransferase (AST), total protein (TP), seralbumin (SA), total bilirubin
(TBIL), bilirubin direct (DBIL), urea (UREA), creatinine (CRE), glucose (GLU), creatine kinase (CK), lactic acid are de-
Hydrogen enzyme (LDH), calcium (CA), sodium (NA), potassium (K), troponin T (CTnT), creatine kinase isozyme quantitative determination (CKIQD),
Serum PT in patients (PT), international standardization ratio (INR), coagulates at plasma prothrombin mobility measurement (PPAD)
Hemase time measurement (TT), plasma D-dimer determination (DD) have significant correlation with pectoralgia diagnostic result (P value is less than 0.05).
As it can be seen that plasma D-dimer determination (DD) described above, troponin T (CTnT), creatine kinase (CK), the same work of creatine kinase
Enzyme quantitative determines (CKIQD), aspartate aminotransferase (AST), urea (UREA), platelet count (BPC), glucose
(GLU), creatinine (CRE), seralbumin (SA), total protein (TP), bilirubin direct (DBIL) and sodium (NA) are all contained in
Interior, characteristic variable needed for illustrating the pectoralgia diagnostic model that method above obtains has significant correlation with pectoralgia diagnostic result.
1 data set basic condition of table and correlation analysis
Fig. 6 is the flow chart of the preparation method for the pectoralgia diagnostic model that one embodiment of the invention provides.As shown in fig. 6, should
Method includes:
Step S61, characteristic variable needed for determining the pectoralgia diagnostic model；
Step S62, the history detection of the historical data and pectoralgia of characteristic variable needed for obtaining the pectoralgia diagnostic model
As a result；And
Step S63 is based on machine learning method, according to the historical data of characteristic variable needed for the pectoralgia diagnostic model
With the history testing result of the pectoralgia, the pectoralgia diagnostic model is established.
In the present embodiment, characteristic variable needed for determining pectoralgia diagnostic model by method as described above first.Then,
The historical data of characteristic variable needed for obtaining the pectoralgia diagnostic model and the history testing result of pectoralgia, specific as follows:
The historical data of characteristic variable needed for obtaining the pectoralgia diagnostic model and the history testing result institute needle of pectoralgia
Pair patient meet following standard:
1,18 years old≤age≤45 year old；
2, meet European Society of Cardiology, American Society of Cardiology's foundation, American Heart Association and world Heart Federation
The improved ACS diagnostic criteria of standard in 2007 of Joint Task group, is diagnosed as ACS patient.
3, using imageological examination, that is, echocardiogram appropriate, CT, MRI or the clearly following information of Aortography:
1) it must determine or exclude dissection of aorta；
2) clear interlayer involves position, involves aorta ascendens or is only limited to descending aorta or the arch of aorta；
3) if it is possible, finding some anatomical features of interlayer as far as possible.
4, it avoids having history of operation, history of intracranialing hemorrhage at no distant date, avoids having for nearly 2 months tumour or myocardial infarction, pernicious swollen
Tumor, fever, acute or chronic diseases associated with inflammation, blood disease and autoimmune disease.
Selection certain hospital emergency department on January 1,1 day to 2017 January in 2011 of the embodiment of the present invention and outpatient service it is quasi- examine ACS and
The patient that Aortic Dissection is accepted for medical treatment totally 529, wherein ACS patient's number be 275, women 43 (15.64%), male
232 (84.36%), the age be 36.85 (18-45) years old, dissection of aorta be 243, women 22 (9.05%), male
221 (90.95%), the age was at 36.59 (18-45) years old.All ACS patients underwent ECG examination, myocardium enzyme damage mark
Object five inspections are diagnosed as ACS, all dissection of aorta patients underwent enhancings referring to acute coronary syndrome diagnostic criteria
CT is diagnosed as dissection of aorta.
In the history testing result annotation process of pectoralgia, the index that uses is the patient when the of time medical checking information
One-shot measurement value and the first time diagnosis for checking information.Utilize the serum cardiac troponin T marker in patient test's information
Index, patient or electrocardiogram prompt " ST Duan Taigao " by the value of troponin T greater than 0.1 are labeled as Acute Coronary Syndrome
Sign, in diagnosis include " acute coronary syndrome ", " ACS ", " acute myocardial infarction AMI " patient also with flesh calcium
Albumen T and ECG results are verified, if meeting above-mentioned condition, emergency treatment coronary syndrome are labeled it as, if not
Meet above-mentioned condition, then it is not labeled.To the patient for carrying out CT examination, " aorta increasing will be included in diagnosis description
Include " dissection of aorta " in width ", the patient of " true room vacation chamber " or " aorta ascendens " and diagnosis and clearly indicates its type
Patient be labeled as dissection of aorta.
For above-mentioned patient, take variable relevant to pectoralgia as the history number of characteristic variable needed for pectoralgia diagnostic model
According to taking history testing result of corresponding the marked testing result data as pectoralgia.
Such as meet that data are imperfect, and data have the case where missing, can using delete include missing values sample or missing
Fill up two methods.It deletes the sample comprising missing values and has lost more samples although ensure that the authenticity of data set
This.On the contrary, missing values complementing method ensure that sample size, but the authenticity of data set is by interference from human factor, it cannot be guaranteed that.
Multiple interpolating method is common missing values complementing method, derives from Bayesian Estimation, it is believed that value to be filled up
It is value random, that value has been observed from other.When carrying out data filling, firstly, generating k using the value observed
Partial data set.Secondly, for statistical analysis to each complete interpolation data collection.Finally, being selected using score function
It selects, selects a most suitable interpolation data collection.
Subsequently, based on machine learning method, by the historical data of characteristic variable needed for pectoralgia diagnostic model and pectoralgia
History testing result is brought into the formula of machine learning method, can establish pectoralgia diagnostic model.
Machine learning algorithm described above may include: in logistic regression algorithm, random forests algorithm and SVM algorithm
At least one.
The principle of SVM algorithm are as follows:
SVM is also known as support vector machines, is a kind of common method of discrimination.Commonly used to carry out pattern-recognition, classification and
Regression analysis belongs to the learning model for having supervision.Its main thought be linear can a point situation analyzed, for linearly not
The case where can dividing, converts high dimensional feature for the sample of low-dimensional input space linearly inseparable by using non-linear map
Space makes its linear separability, so that high-dimensional feature space linearly divides the nonlinear characteristic of sample using linear algorithm
Analysis is possibly realized.Here non-linear map is known as kernel function:
Wherein,For mapping function, the optimization problem of non-linear SVM can be obtained at this time:
So thatThe optimal solution acquiredIt calculatesThe categorised decision function finally obtained are as follows:
When constructing classifier using SVM method, need to select a kernel function first, herein using sigmoid core
Function
K (x, z)=tanh (β xTY+ θ), β > 0, θ < 0
It is that it is determined the result is that global optimum, while ensure that the good generalization ability to unknown sample.
The principle of random forests algorithm are as follows:
Each classification tree in random forest is binary tree, and generation follows top-down recurrence division principle, i.e.,
Successively training set is divided since root node；In binary tree, root node includes whole training datas, not according to node
Purity minimum principle is split into left sibling and right node, they separately include a subset of training data, according to same rule
Then node continues to divide, and stops growing until meeting branch's stopping rule.If the classification data on node n is all from same
One classification, impurity level I (n)=0 of point.Impurity level measure is Gini criterion, i.e., hypothesis P (ω j) is to belong to ω on node n
J class number of samples accounts for the frequency of training sample sum, then Gini criterion is expressed as
Specific algorithm process is as follows:
(1) N indicates original training set number of samples, mallFor indicating the number of variable.
(2) have using bootstrap method randomly select k new self-service sample sets with putting back to, and thus construct k decision
Tree, the sample not being pumped to every time constitute the outer data (out-of-bag, OOB) of k bag.
(3) each self-service sample set randomly selects m at each node of every one tree for establishing a decision treetry
A variable (mtry<mall), the variable of a most classification capacity is then selected, the threshold value of variable classification is by checking each
Classification point determines.
(4) each tree is grown to the maximum extent, does not do any trimming.
(5) more classification trees of generation are formed into random forest, new data is differentiated with random forest grader
With classification, classification results by Tree Classifier ballot it is how many depending on.
In random forest building process, self-service sample set is used for the formation of each Tree Classifier, and sampling generates every time
The outer data (OOB) of bag be used to the accuracy of prediction classification, the OOB for being summarized to obtain error rate to each prediction result estimates
Then meter assesses the accuracy that assembled classifier differentiates.In addition, in random forest, applied self-service sample set is from original
Training sample concentration randomly select, variable applied by every one tree is also from all variable mallIn randomly select, twice with
Machine process makes random forest have more stable error rate, while the performance of classifier is measured using the outer data of bag.
Most important parameter is m in random foresttry, LiawA etc., which passes through to test, to be thoughtIt is one relatively good
Selection.The important parameter of other two in random forest is the number ntree and leaf node nodesize for constructing classification tree
Size, this research are studied using ntree=500.
Pectoralgia diagnostic model is constructed respectively using logistic regression, SVM method and random forest method, is tested by five foldings intersection
Card obtains model evaluation index and ROC curve as shown in table 2 and Fig. 7.
2 model evaluation index of table
According to table 2, the prediction accuracy for the model established using three kinds of machine learning methods is above 85%, wherein
The risk forecast model better effect of random forests algorithm building, susceptibility, specificity and accuracy be above 90%, show with
Machine forest risk forecast model can not only be to acute coronary syndrome discriminating power with higher, while to dissection of aorta
Also discriminating power with higher.
See from Fig. 7, for the model that three kinds of machine learning methods are established, when negative and positive class rate (FPR) is lower, real class
Rate (TPR) can achieve higher level, wherein ideal with the model that random forest method is established.
For the practical generalization ability of further research model, acute coronary syndrome and active are extracted again from emergency department
Arteries and veins interlayer patient amounts to 13 people, wherein 9 people of Acute Coronary Syndrome Patients, 4 people of dissection of aorta patient.It is collected simultaneously patient's sheet
Plasma D-dimer determination (DD), troponin T (CTnT), creatine kinase (CK), creatine kinase in inspection project when secondary medical
Isodynamic enzyme quantitative determines (CKIQD), aspartate aminotransferase (AST), urea (UREA), platelet count (BPC), grape
Sugared (GLU), creatinine (CRE), seralbumin (SA), total protein (TP), bilirubin direct (DBIL) and this 13, sodium (NA)
The first time measured value of index, the model that the three kinds of machine learning methods then obtained using training are established is respectively to patient disease
Risk is predicted that obtained result and evaluation index difference is as shown in table 3.
3 case verification of table
By case verification result it is found that the prediction accuracy value for the model that three kinds of machine learning methods are established is identical, together
When all there is certain generalization ability, but since verify data acquires less, and there is non-equilibrium, cannot compare well
The predictive ability of model.
Present medical treatment has had been introduced into the epoch of big data, and so-called medical treatment big data refers to that medical treatment & health big data can be with
It is related to whole hospitals, health organ and all groups of a country, passes through the daily medical tube of hospital, health organ
Reason, clinical diagnosis, progress note, medication record check that inspection result record can collect numerous medical treatment with high value
Data.Under information-based big data era, for the cause of disease, Related Risk Factors, the danger of young patients with acute chest pain Severe acute disease
Danger layering, treatment method, prognosis etc. carry out the summary of system, and it is flat to establish standardization early warning, layering, prognostic model and health control
Platform gets more prime time to prevent and rescuing young people's pectoralgia Severe acute disease.Pass through model evaluation index Average Accuracy
It is with average Kappa value it is found that best using the prediction model effect of random forest method building.Its accuracy rate can achieve
90.17%, this shows the patient for meaning Aortic Dissection and acute coronary syndrome, is established by big data pre-
Surveying model can achieve 90.17% to the diagnosis rate of patient under the intervention without invasive detection means.This cures reduction
The waste of resource is treated, diagnosis and treatment preferably are carried out to China youth's patients with acute chest pain, reduces acute coronary syndrome and aorta
It is significant to improve people from China youth general level of the health for the case fatality rate of interlayer.
According to Random Forest model obtain variable important feature selecting as a result, it has been found that, first 13 in 46 characteristic variables
Feature is screened out simultaneously, is plasma D-dimer determination (DD), troponin T (CTnT), creatine kinase (CK), flesh respectively
Acid kinase isodynamic enzyme quantitative determines (CKIQD), aspartate aminotransferase (AST), urea (UREA), platelet count
(BPC), glucose (GLU), creatinine (CRE), seralbumin (SA), total protein (TP), bilirubin direct (DBIL) and sodium
(NA), there is apparent diagnostic value for identifying dissection of aorta and acute coronary syndrome.
By this result of study we it is found that plasma D-dimer content and acute coronary syndrome and Aortic Dissection
Antidiastole have maximum relevance.Plasma D-dimer content is after the activated fibrin stabilizing factor of fibrin monomer is crosslinked, through fibre
The selective degradation product that lyase hydrolysis generates.Plasma D-dimer content fibrinolytic process has specificity, can be living to intracorporal fibrinolytic
Property and blood coagulation activity reflected, therefore can be used as in clinic a kind of marker for judging Aortic Dissection.It can incite somebody to action
Plasma D-Dimer levels, which increase, to be used to carry out the judgement of hypercoagulative state, while also can be used as secondary fibrinolytic and thrombosis
Marker.Currently, a large amount of clinical researches it has been proved that ACS patient there is also the disorders of blood coagulation and Balance of Fibrinolysis System, therefore theory pushes away
Plasma D-Dimer levels raising should be had by surveying ACS patient, but plasma D-Dimer levels and the relationship of ACS are unlike blood plasma D- bis-
The relationship of aggressiveness and dissection of aorta.
The embodiment of the present invention also provides a kind of machine readable storage medium, and finger is stored on the machine readable storage medium
It enables, which is used for so that machine executes the judgment method of chest pain patients dissection of aorta described above.
Through the above technical solutions, using the Illnesses Diagnoses method and machine readable storage of chest pain patients provided by the invention
Medium, current signature variable needed for obtaining pectoralgia diagnostic model for chest pain patients, and chest is judged by pectoralgia diagnostic model
Whether pain patient suffers from acute coronary syndrome or dissection of aorta, can accurately detect whether chest pain patients suffer from acute coronary
Syndrome or dissection of aorta, and spend less.
The optional embodiment of the embodiment of the present invention is described in detail in conjunction with attached drawing above, still, the embodiment of the present invention is simultaneously
The detail being not limited in above embodiment can be to of the invention real in the range of the technology design of the embodiment of the present invention
The technical solution for applying example carries out a variety of simple variants, these simple variants belong to the protection scope of the embodiment of the present invention.
It is further to note that specific technical features described in the above specific embodiments, in not lance
In the case where shield, it can be combined in any appropriate way.In order to avoid unnecessary repetition, the embodiment of the present invention pair
No further explanation will be given for various combinations of possible ways.
It will be appreciated by those skilled in the art that implementing the method for the above embodiments is that can pass through
Program is completed to instruct relevant hardware, which is stored in a storage medium, including some instructions are used so that single
Piece machine, chip or processor (processor) execute all or part of the steps of each embodiment the method for the application.And it is preceding
The storage medium stated includes: USB flash disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory
The various media that can store program code such as (RAM, Random Access Memory), magnetic or disk.
In addition, any combination can also be carried out between a variety of different embodiments of the embodiment of the present invention, as long as it is not
The thought of the embodiment of the present invention is violated, equally should be considered as disclosure of that of the embodiment of the present invention.
Claims (10)
1. a kind of Illnesses Diagnoses method of chest pain patients, which is characterized in that this method comprises:
For the chest pain patients, current signature variable needed for obtaining pectoralgia diagnostic model；And
According to the current signature variable and the pectoralgia diagnostic model, judge whether the chest pain patients suffer from acute coronary
Syndrome or dissection of aorta.
2. the Illnesses Diagnoses method of chest pain patients according to claim 1, which is characterized in that the pectoralgia diagnostic model is logical
Cross following manner acquisition:
Characteristic variable needed for determining the pectoralgia diagnostic model；
The historical data of characteristic variable needed for obtaining the pectoralgia diagnostic model and the history testing result of pectoralgia；And
Based on machine learning method, according to the historical data of characteristic variable needed for the pectoralgia diagnostic model and the pectoralgia
History testing result establishes the pectoralgia diagnostic model.
3. the Illnesses Diagnoses method of chest pain patients according to claim 2, which is characterized in that the determination pectoralgia is examined
Characteristic variable needed for disconnected model includes:
Obtain variable relevant to pectoralgia；
The variable relevant to pectoralgia is normalized；
The variable after normalization is screened by recursive feature null method to obtain the first selection variables；And
Using first selection variables as characteristic variable needed for the pectoralgia diagnostic model.
4. the Illnesses Diagnoses method of chest pain patients according to claim 2, which is characterized in that the determination pectoralgia is examined
Characteristic variable needed for disconnected model includes:
By the index MeanDecreaseAccuracy and MeanDecreaseGini of random forests algorithm respectively to it is described with
The importance of the relevant variable of pectoralgia carries out screening and sequencing, obtains the first sequence of importance sequence and the second sequence of importance sequence
Column；
Since importance highest, takes out and be greater than in advance respectively from first sequence of importance and second sequence of importance
If numerical value and the variable of identical quantity as the second selection variables and third selection variables, make second selection variables and described
Third selection variables are identical；
Using second selection variables or the third selection variables as characteristic variable needed for the pectoralgia diagnostic model.
5. according to the Illnesses Diagnoses method of chest pain patients described in any one of claim 2-4 claim, feature exists
In characteristic variable needed for the pectoralgia diagnostic model includes following human body indicators:
Plasma D-dimer determination, troponin T, creatine kinase, creatine kinase isozyme quantitative determination, aspartic acid amino turn
Move enzyme, urea, platelet count, glucose, creatinine, seralbumin, total protein, bilirubin direct and sodium.
6. the Illnesses Diagnoses method of chest pain patients according to claim 3 or 4, which is characterized in that described related to pectoralgia
Variable be selected from the group comprising following human body indicators:
Gender, age, urea, neutrophil leucocyte, gamma-glutamyl based transferase, lymphocyte, creatinine, monocyte, grape
Sugar, eosinophil, serum uric acid, basophilic granulocyte, creatine kinase, white blood cell count(WBC), lactic dehydrogenase, red blood cell meter
It is number, calcium, hemoglobinometry, sodium, hematid specific volume measurement, potassium, mean corpuscular volume (MCV), chloride, platelet count, inorganic
Phosphorus, mean corpuscular hemoglobin, magnesium, mean corpuscular hemoglobin concentration (MCHC), troponin T, erythrocyte volume distribution are wide
Degree measurement CV, Natriuretic Peptide, mean platelet volume measurement, creatine kinase isozyme quantitative determination, alanine amino turn
Shifting enzyme, serum PT in patients, aspartate aminotransferase, plasma fibrinogen measurement, total protein, blood plasma are solidifying
The measurement of hemase original mobility, seralbumin, international standardization ratio, total bilirubin, thrombin time test, direct gallbladder are red
Element, plasma D-dimer determination and alkaline phosphatase.
7. the Illnesses Diagnoses method of chest pain patients according to claim 3, which is characterized in that in the acquisition and pectoralgia phase
After the variable of pass, this method further include:
Missing values are carried out to the variable relevant to pectoralgia using multiple interpolating method to fill up.
8. the Illnesses Diagnoses method of chest pain patients according to claim 2, which is characterized in that the machine learning algorithm packet
It includes:
At least one of logistic regression algorithm, random forests algorithm and SVM algorithm.
9. the Illnesses Diagnoses method of chest pain patients according to claim 4, which is characterized in that this method further include:
Obtain the variable for having significant correlation with the diagnostic result of pectoralgia；
Whether characteristic variable needed for examining the pectoralgia diagnostic model is that the diagnostic result with pectoralgia has significant correlation
Variable.
10. a kind of machine readable storage medium, which is characterized in that be stored with instruction on the machine readable storage medium, the instruction
For making machine perform claim require the Illnesses Diagnoses method of chest pain patients described in any one of 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811260651.9A CN109585011A (en) | 2018-10-26 | 2018-10-26 | The Illnesses Diagnoses method and machine readable storage medium of chest pain patients |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811260651.9A CN109585011A (en) | 2018-10-26 | 2018-10-26 | The Illnesses Diagnoses method and machine readable storage medium of chest pain patients |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109585011A true CN109585011A (en) | 2019-04-05 |
Family
ID=65920882
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811260651.9A Pending CN109585011A (en) | 2018-10-26 | 2018-10-26 | The Illnesses Diagnoses method and machine readable storage medium of chest pain patients |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109585011A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111180064A (en) * | 2019-12-25 | 2020-05-19 | 北京亚信数据有限公司 | Evaluation method and device for auxiliary diagnosis model and computing equipment |
| CN111834003A (en) * | 2020-04-18 | 2020-10-27 | 李智敏 | Novel method, system and equipment for preliminary screening diagnosis of coronary pneumonia and influenza pneumonia |
| CN114882988A (en) * | 2022-03-22 | 2022-08-09 | 复旦大学附属中山医院 | Acute aortic dissection diagnosis and prediction system |
| CN115662612A (en) * | 2022-09-23 | 2023-01-31 | 中国人民解放军南部战区总医院 | Doctor auxiliary system for rapid differential diagnosis of high-risk acute chest pain |
| CN117594131A (en) * | 2024-01-17 | 2024-02-23 | 北京市心肺血管疾病研究所 | Device for identifying or assisting in identifying acute chest pain type and application thereof |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103733065A (en) * | 2011-06-02 | 2014-04-16 | 阿尔玛克诊断有限公司 | Molecular diagnostic test for cancer |
| CN105102631A (en) * | 2012-12-03 | 2015-11-25 | 阿尔玛克诊断有限公司 | Molecular diagnostic test for cancer |
| CN105543369A (en) * | 2016-01-13 | 2016-05-04 | 金锋 | Mental disorder biomarker and application thereof |
| CN105814571A (en) * | 2013-11-07 | 2016-07-27 | 曼迪奥研究有限公司 | Methods and systems of evaluating a risk of lung cancer |
| US20170032099A1 (en) * | 2014-04-08 | 2017-02-02 | Centre Hospitalier Universitaire D'angers | Prognostic tests for hepatic disorders |
| CN107480441A (en) * | 2017-08-04 | 2017-12-15 | 方芳 | A kind of modeling method and system of children's septic shock prognosis prediction based on SVMs |
-
2018
- 2018-10-26 CN CN201811260651.9A patent/CN109585011A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103733065A (en) * | 2011-06-02 | 2014-04-16 | 阿尔玛克诊断有限公司 | Molecular diagnostic test for cancer |
| CN105102631A (en) * | 2012-12-03 | 2015-11-25 | 阿尔玛克诊断有限公司 | Molecular diagnostic test for cancer |
| CN105814571A (en) * | 2013-11-07 | 2016-07-27 | 曼迪奥研究有限公司 | Methods and systems of evaluating a risk of lung cancer |
| US20170032099A1 (en) * | 2014-04-08 | 2017-02-02 | Centre Hospitalier Universitaire D'angers | Prognostic tests for hepatic disorders |
| CN105543369A (en) * | 2016-01-13 | 2016-05-04 | 金锋 | Mental disorder biomarker and application thereof |
| CN107480441A (en) * | 2017-08-04 | 2017-12-15 | 方芳 | A kind of modeling method and system of children's septic shock prognosis prediction based on SVMs |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111180064A (en) * | 2019-12-25 | 2020-05-19 | 北京亚信数据有限公司 | Evaluation method and device for auxiliary diagnosis model and computing equipment |
| CN111834003A (en) * | 2020-04-18 | 2020-10-27 | 李智敏 | Novel method, system and equipment for preliminary screening diagnosis of coronary pneumonia and influenza pneumonia |
| CN114882988A (en) * | 2022-03-22 | 2022-08-09 | 复旦大学附属中山医院 | Acute aortic dissection diagnosis and prediction system |
| CN115662612A (en) * | 2022-09-23 | 2023-01-31 | 中国人民解放军南部战区总医院 | Doctor auxiliary system for rapid differential diagnosis of high-risk acute chest pain |
| CN117594131A (en) * | 2024-01-17 | 2024-02-23 | 北京市心肺血管疾病研究所 | Device for identifying or assisting in identifying acute chest pain type and application thereof |
| CN117594131B (en) * | 2024-01-17 | 2024-04-16 | 北京市心肺血管疾病研究所 | Device for identifying or assisting in identifying acute chest pain type and application thereof |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109585011A (en) | The Illnesses Diagnoses method and machine readable storage medium of chest pain patients | |
| Tucker et al. | Early diagnostic efficiency of cardiac troponin I and troponin T for acute myocardial infarction | |
| Bellavia et al. | Comparison of right ventricular longitudinal strain imaging, tricuspid annular plane systolic excursion, and cardiac biomarkers for early diagnosis of cardiac involvement and risk stratification in primary systematic (AL) amyloidosis: a 5-year cohort study | |
| JP4486260B2 (en) | Method and apparatus for predicting the presence of hemostatic dysfunction in patient samples | |
| CN113327679A (en) | Pulmonary embolism clinical risk and prognosis scoring method and system | |
| US6277070B1 (en) | Medical analysis and treatment method and system | |
| Stahl et al. | Can capillary whole blood glucose and venous plasma glucose measurements be used interchangeably in diagnosis of diabetes mellitus? | |
| Devarajan | NGAL for the detection of acute kidney injury in the emergency room | |
| Sun et al. | Value of SOFA, APACHE IV and SAPS II scoring systems in predicting short-term mortality in patients with acute myocarditis | |
| US11348687B2 (en) | Personalizing a biomarker signal and medical procedures by determining plasma volume variation of one or more markers | |
| CN112669960A (en) | Liver fibrosis prediction model construction method based on machine learning method, prediction system, device and storage medium | |
| CN105765382A (en) | Methods and kits for predicting the risk of having a cardiovascular disease or event | |
| Kramer et al. | A multimodal score accurately classifies Fontan failure and late mortality in adult Fontan patients | |
| CN108133754B (en) | The forecasting system of bleeding risk after a kind of thrombolysis | |
| CN111341452A (en) | Multi-system atrophy and disability prediction method, model building method, device and equipment | |
| WO2017177192A1 (en) | White blood cell population dynamics | |
| Deng et al. | Characterizing risk of in-hospital mortality following subarachnoid hemorrhage using machine learning: a retrospective study | |
| Hong et al. | Red cell distribution width as an independent predictor of exercise intolerance and ventilatory inefficiency in patients with chronic heart failure | |
| Kataoka | Changes in red blood cell volume during transition of heart failure status: a reflection of cellular hydration status? | |
| Fang et al. | Prediction of acute kidney injury after total aortic arch replacement with serum cystatin C and urine N-acetyl-β-d-glucosaminidase: A prospective observational study | |
| RU2650212C1 (en) | Method of estimation of overdiagnosis of myocardial infarction | |
| Plebani et al. | Clinical evaluation of a new quantitative method for specific IgE antibodies | |
| CN110070942A (en) | A kind of chronic liver disease risk evaluating system promoting tree-model based on gradient | |
| Yilmaz et al. | Evaluation of the effect of hyperchloremia on the prognosis and mortality of medical intensive care patients: a single-center study | |
| CN108872583A (en) | Hypertensive disorder in pregnancy correlating markings compositions and its application |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190405 |
|
| RJ01 | Rejection of invention patent application after publication |