CN114155955A - Airway obstruction severity assessment method and system - Google Patents
Airway obstruction severity assessment method and system Download PDFInfo
- Publication number
- CN114155955A CN114155955A CN202111455643.1A CN202111455643A CN114155955A CN 114155955 A CN114155955 A CN 114155955A CN 202111455643 A CN202111455643 A CN 202111455643A CN 114155955 A CN114155955 A CN 114155955A
- Authority
- CN
- China
- Prior art keywords
- index
- patient
- airway
- severity
- airway obstruction
- 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
Images
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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- 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
- G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
- G16H10/60—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for patient-specific data, e.g. for electronic patient records
-
- 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
Abstract
The invention discloses a method and a system for evaluating the severity of airway obstruction, wherein the method comprises the following steps: acquiring instant diagnosis data of a current patient; extracting key index characteristic data from the instant diagnosis data; and inputting the key index characteristic data into an airway obstruction severity evaluation model which is constructed in advance based on a patient medical data source, and obtaining the airway obstruction severity score of the patient according to the output of the airway obstruction severity evaluation model. The invention can assist related rescue personnel to quickly and correctly evaluate and timely and effectively treat the obstruction of the airway of the patient, and improve the rescue ability under emergency and sudden conditions.
Description
Technical Field
The invention relates to the field of data processing, in particular to a method and a system for evaluating the severity of airway obstruction.
Background
Emergency medicine is a comprehensive clinical subject of a scientific management system for timely, rapid and effective treatment and research of acute diseases in the aspect of the whole body of a machine body by taking the development of modern medical science and treatment measures of clinical medicine as means. With the development of medical science, emergency medicine becomes an independent novel comprehensive medical subject and is one of the most rapidly developed clinical subjects at present. The emergency medical science is different from the conventional outpatient medical science, has higher requirements on diagnosis and treatment of disease conditions, and requires a doctor to make a correct diagnosis and quickly treat the disease conditions of a patient in a short time through limited disease condition information of the patient, so that the method is very meaningful for providing a quick, efficient and accurate disease condition analysis method for the emergency doctor.
Airway obstruction is one of the important causes of traumatic death, and timely and effective airway management is crucial to reducing mortality. The core problem of the trauma airway management is the accurate identification of the early stage of airway obstruction and the reason. The core technology of airway management is to protect the airway, open the airway, and establish artificial ventilation, the primary purpose of which is to ensure good ventilation and oxygenation. In emergency situations, a decision must be made as early as possible as to whether the patient needs airway management. However, for patients with spontaneous breathing, making this decision is often significantly delayed, and the condition of severe airway obstruction once it occurs can pose a significant threat to the lives of the injured persons. Patients often become quickly critical if they can identify the airway correctly and intervene effectively at an early stage. Therefore, how to make emergency medical care personnel quickly and accurately judge whether the patient has the airway obstruction and the severity of the airway obstruction is an important subject in the industry.
Disclosure of Invention
The invention provides a method and a system for evaluating the severity of airway obstruction, which are used for assisting related rescue workers to quickly and correctly evaluate and timely and effectively treat the airway obstruction of a patient and improve the rescue ability under emergency conditions.
Therefore, the invention provides the following technical scheme:
a method of assessing the severity of airway obstruction, the method comprising:
acquiring instant diagnosis data of a current patient;
extracting key index characteristic data from the instant diagnosis data;
inputting the key index characteristic data into an airway obstruction severity evaluation model which is constructed in advance based on a patient medical data source, and obtaining an airway obstruction severity score of the patient according to the output of the airway obstruction severity evaluation model;
the airway obstruction severity evaluation model comprises an input layer, three hidden layers and an output layer, wherein the three hidden layers are fully connected with the output layer, and a random inactivation mechanism is used between the adjacent hidden layers.
Optionally, the constructing an airway obstruction severity assessment model based on the patient medical data source comprises:
acquiring a patient medical data source and extracting patient treatment related data from the medical data source;
performing correlation analysis on the patient diagnosis related data to obtain a key index set;
carrying out normalization processing and missing value filling on the indexes in the key index set to obtain a key index sample set;
determining an airway intervention means label and an airway severity corresponding to the airway intervention means label of each sample in the key index sample set;
and training according to the key index sample set and the severity of the air passage to obtain an air passage obstruction severity evaluation model.
Optionally, the patient medical data source comprises any one or more of: an emergency database, a nursing system and a hospital database.
Optionally, the patient visit related data comprises any one or more of: patient basic information, patient airway intervention records, patient detection indexes and consciousness state information of the patient;
the patient airway intervention record comprises: nasal catheter oxygen inhalation, mask oxygen inhalation, oropharyngeal airway or tracheal intubation;
the patient detection indicators include: vital signs, blood routine, blood and qi analysis, blood coagulation function, and emergency biochemistry;
the consciousness state information of the patient includes: mental state, whether restlessness.
Optionally, the performing correlation analysis on the patient visit related data to obtain a key index set includes:
carrying out correlation analysis on the patient detection indexes to obtain a candidate key index set;
sorting each candidate index in the candidate key index set and the correlation coefficient of the severity of the airway obstruction according to the absolute value, selecting the candidate index of which the absolute value of the correlation coefficient is more than or equal to a set threshold value to add into the recommended index set, and otherwise adding into the residual index set;
classifying the importance degree of each index in the recommendation index set by combining medical knowledge, and adding the indexes in the recommendation index set into a key index set;
selecting part of indexes from the residual index set and adding the selected part of indexes into the key index set;
adding a mental state label and a restlessness label to each training sample in the set of key indicators.
Optionally, the normalizing the indexes in the key index set includes:
if the index is in a normal interval, taking the index as 0;
if the index is higher than the normal interval and lower than the extreme danger upper limit value, mapping the index value into a value in an interval of 0 to 1;
if the index is higher than or equal to the extreme danger upper limit value, taking the index as 1;
if the index is lower than the normal interval and higher than the extremely dangerous lower limit value, mapping the index value into a value in an interval from 0 to-1;
and if the index is lower than or equal to the extreme danger lower limit value, taking the index as-1.
Optionally, the obtaining of the airway obstruction severity assessment model trained according to the key index sample set and the airway intervention means label includes:
and according to the key index sample set and the airway intervention means label, adopting MAE as a loss function, and using an Adam optimizer for iterative training to obtain an airway obstruction severity evaluation model.
An airway obstruction severity assessment system, the system comprising:
the data acquisition module is used for acquiring the instant diagnosis data of the current patient;
the characteristic extraction module is used for extracting key index characteristic data from the instant diagnosis data;
the evaluation module is used for inputting the key index characteristic data into an airway obstruction severity evaluation model which is constructed in advance based on a patient medical data source, and obtaining the airway obstruction severity score of the patient according to the output of the airway obstruction severity evaluation model;
the airway obstruction severity evaluation model comprises an input layer, three hidden layers and an output layer, wherein the three hidden layers are fully connected with the output layer, and a random inactivation mechanism is used between the adjacent hidden layers.
Optionally, the system further comprises:
the model building module is used for building an airway obstruction severity evaluation model; the model building module comprises:
the data acquisition unit is used for acquiring a medical data source of a patient and extracting the data related to the patient treatment from the medical data source;
the data analysis unit is used for carrying out correlation analysis on the patient diagnosis related data to obtain a key index set;
the data processing unit is used for carrying out normalization processing and missing value filling on the indexes in the key index set to obtain a key index sample set;
the label determining unit is used for determining the airway intervention means label of each sample in the key index sample set and the airway severity corresponding to the airway intervention means label;
and the training unit is used for training according to the key index sample set and the severity of the air passage to obtain an air passage obstruction severity evaluation model.
Optionally, the data analysis unit includes:
the data analysis subunit is used for carrying out correlation analysis on the patient detection indexes to obtain a candidate key index set;
the sorting processing subunit is used for sorting each candidate index in the candidate key index set according to the absolute value of the correlation coefficient between the candidate index and the severity of the airway obstruction, selecting the candidate index with the absolute value of the correlation coefficient greater than or equal to a set threshold value to add into the recommendation index set, and otherwise adding into the residual index set;
the screening subunit is used for grading the importance degree of each index in the recommendation index set by combining medical knowledge and then adding the indexes in the graded recommendation index set into the key index set;
selecting part of indexes from the residual index set and adding the selected part of indexes into the key index set;
and the label adding subunit is used for adding a mental state label and a restless label to each training sample in the key index set.
The method and the system for evaluating the severity of the airway obstruction provided by the embodiment of the invention construct an airway obstruction severity evaluation model in advance based on a patient medical data source, extract key index features from the current instant diagnosis data of a patient, input the key index features into the airway obstruction severity evaluation model, and obtain the score of the severity of the airway obstruction of the patient according to the output of the airway obstruction severity evaluation model. Thereby supplementary first aid medical personnel of intelligence whether have the air flue to block the disease to the patient and carry out quick accurate judgement, guide the treatment personnel to block quick accurate aassessment and timely effective processing to the air flue, improve the air flue and block the efficiency of treatment.
Drawings
FIG. 1 is a schematic structural diagram of an airway obstruction severity assessment model in an embodiment of the invention;
FIG. 2 is a flow chart of the construction of an airway obstruction severity assessment model in an embodiment of the present invention;
FIG. 3 is a flowchart illustrating missing value filling of indicators in the key indicator set according to an embodiment of the present invention;
FIG. 4 is a flow chart of a method of assessing the severity of airway obstruction in an embodiment of the present invention;
FIG. 5 is a schematic structural diagram of an airway obstruction severity assessment system according to an embodiment of the present invention;
FIG. 6 is a schematic structural diagram of a model building module in an embodiment of the present invention.
Detailed Description
The embodiment of the invention provides an airway obstruction severity evaluation method and system, wherein an airway obstruction severity evaluation model is constructed in advance based on a patient medical data source, key index features are extracted from instant diagnosis data of a current patient, then the key index features are input into the airway obstruction severity evaluation model, and an airway obstruction severity score of the patient is obtained according to the output of the airway obstruction severity evaluation model. Thereby supplementary first aid medical personnel of intelligence whether have the air flue to block the disease to the patient and carry out quick accurate judgement, guide the treatment personnel to block quick accurate aassessment and timely effective processing to the air flue, improve the air flue and block the efficiency of treatment.
In the embodiment of the invention, the airway obstruction severity evaluation model adopts a machine learning model which adopts a multi-layer perceptron structure, and as shown in fig. 1, the model comprises an input layer, three hidden layers and an output layer. Wherein, the three hidden layers are fully connected with the output layer, and a random inactivation mechanism is used between the adjacent hidden layers to calculate through a Linear rectification function (RecU) activation function. The ReLU activation function generally refers to a non-linear function represented by a ramp function and its variants.
First, a process of constructing an airway obstruction severity evaluation model in the embodiment of the present invention will be described in detail.
Fig. 2 is a flowchart of constructing an airway obstruction severity assessment model according to an embodiment of the present invention, including the following steps:
The patient medical data sources may include, but are not limited to, any one or more of: an emergency database, a nursing system and a hospital database.
The patient visit related data may include, but is not limited to, any one or more of the following: patient basic information (such as age, sex and the like), patient airway intervention records (such as nasal catheter oxygen inhalation, mask oxygen inhalation, oropharyngeal airway or tracheal intubation), patient detection indexes (such as vital signs, blood routine, blood and gas analysis, blood coagulation function, emergency treatment biochemistry and the like), and consciousness state information (such as mental state, whether restlessness and the like) of a patient.
And 202, performing correlation analysis on the patient visit related data to obtain a key index set.
In specific implementation, a Pearson (Pearson) correlation analysis technology can be adopted to analyze detection index sets of vital signs, blood routine, blood-gas analysis, blood coagulation function and the like, and a candidate key index set is formed by combining medical priori knowledge.
The Pearson correlation coefficient is a numerical value between-1 and describes the linear correlation degree between the two variables, and if the absolute value of the correlation coefficient of the two variables is closer to 1, the correlation degree is stronger; if the correlation coefficient is 0, it indicates that there is no linear correlation between the two variables.
The Pearson correlation coefficient calculation formula is as follows:
x and Y in the formula represent two variables respectively, cov (X, Y) calculates covariance of variables, sigma X, sigmaYCalculating the variance of the variables X and Y respectively; the formula can be decomposed into the rightmost form,
respectively represent the average values of variables X and Y, and Xi and Yi respectively represent the ith sample values of X and Y, and the total number is n samples.
And sorting each candidate index in the candidate key index set and the correlation coefficient of the severity of the airway obstruction according to the absolute value, selecting the index of which the absolute value of the correlation coefficient is more than or equal to a threshold value t, adding the index into the recommendation index set, and otherwise, adding the residual index set.
The index weight obtained by the Pearson correlation coefficient calculation formula can describe the importance degree of the index to the airway obstruction phenomenon to a certain extent, but certain deviation exists. Through analyzing the medical record of the patients, the phenomena of multiple injuries, head injuries, chest and abdomen injuries and the like of most patients with airway obstruction are discovered. Due to the existence of multiple injuries, physiological indexes of other diseases of the patient are changed, and whether the patient has airway obstruction or not is judged. For example: the correlation weight of the red blood cell count is high, but the index has little influence on the airway, and if the weight of the index is set to be large, the generalization capability of the model is deteriorated. Therefore, when the key index set is selected, the index weight given by the correlation analysis needs to be revised in combination with the medical priori knowledge, and the weight of part of the indexes with small correlation with the disease condition in medicine is reduced. Therefore, the key index set is determined, firstly, the importance degree of each index in the recommendation index set is graded by combining medical knowledge, and the graded recommendation index set is added into the key index set; screening all indexes in the rest index sets, selecting part of indexes with great medical significance to the disease condition, and adding the selected indexes into the key index set.
For example, for each index in the key index set, all indexes are divided into 3 levels according to the ranking result of the Pearson correlation coefficient and medical knowledge, and the indexes are respectively given different weights, and the indexes under the same level have the same weight. The key index set is shown in table 1.
Table 1:
in addition, a large number of experiments show that physiological phenomena such as the coma degree and restlessness of the patient have certain influence on the judgment of the airway obstruction, so that a mental state label and a restlessness label are added in a centralized way in key indexes and used for describing the current mental state of the patient.
And 203, performing normalization processing and missing value filling on the data in the key index set to obtain a key index sample set.
For different patients, due to the difference of individual differences of age, sex, physical quality and the like, various physiological indexes have deviation within a certain range. Taking the respiration rate as an example: 60-100 were normal levels, with less than 50 or more than 120 being more severe. Since the evaluation interval of the normal range is large, individual differences have a large influence on the model. Therefore, data can be processed by adopting a normalization method according to the prior medical prior knowledge so as to reduce the influence of individual difference on the model.
The specific normalization processing method comprises the following steps:
if a certain index is in a normal interval, taking the index as 0;
if the index is higher than or equal to the upper limit value of extreme danger (lethal risk exists), the index is set to be 1;
if the index is higher than the normal interval and lower than the extreme danger upper limit value, mapping the index value into an interval from 0 to 1;
if the index is lower than or equal to the lower extreme danger limit value, the index is valued as-1;
and if the index is lower than the normal interval and higher than the extremely dangerous lower limit value, mapping the index value into an interval from 0 to-1.
For example, the Respiratory Rate (RR) in basic vital signs, recorded as-1 when RR <8, indicates that the RR is too low to be extremely dangerous; the recorded value is scaled to be between-1 and 0 when RR is more than or equal to 8 and RR is less than 12; the value is normal when RR is more than or equal to 12 and RR is less than 24, and the value is recorded as 0; the recorded value is scaled to be between 0 and 1 when RR is more than or equal to 24 and RR is less than 34; when RR ≧ 34 is recorded as 1, it indicates that the RR was too high to be extremely dangerous. For other detection indexes, the values of the normal interval and the extreme values of overhigh and overlow can be calculated by combining a statistical method with medical priori knowledge.
Since the patient may not detect a certain index for various reasons, partial index records may be missing. For a filling scheme for missing data, after patient data is normalized, a K-Nearest Neighbor (KNN) predictive missing value filling method is first used. For patients with a certain index k missing, the KNN filling method firstly needs to select patient data with no index k missing and less index k missing from all patient data as a training sample set T. Removing the index k to be filled of the patient, calculating the similarity between the patient and each sample data in the T, selecting N sample data most similar to the patient, and calculating the average value of the index as a filling value. If the patient missing value is too large to compare with the similarity of other samples, or if there are no N sample data similar to the patient in the training sample, the 0 value filling method (normal value filling method) is adopted. This filling method can be understood as if the patient does not test the indicator, then the indicator is normal by default. In the application process, the number of detection indexes of a patient is increased along with the detection progress, and the prediction accuracy is increased. If only basic vital sign data exists at the beginning of emergency treatment, other indexes are all absent, and the accuracy of the obtained prediction result is poor. With the more and more acquired indexes, the prediction accuracy is higher and higher, so that for patients with excessive deficiency values (patients with less detection indexes), the 0 value filling method can be used for performing relatively accurate initial diagnosis on the patients.
Fig. 3 shows a flowchart of missing value filling for the indicators in the key indicator set in the embodiment of the present invention, which includes the following steps:
And step 304, calculating the similarity between the patient and each sample in the training sample set T by using the indexes except the index k.
In step 305, it is determined whether there are at least N samples highly similar to the patient, i.e. the difference between each index is small. If so, go to step 306; otherwise, step 307 is executed. Too high value of N can lead to a large number of samples to be difficult to match, too low value is easily influenced by individual cases, and N is generally 5-10 in experiments.
And step 306, filling the patient index k by taking the average value of the index k of the most similar N samples as a filling value.
By the above-described filling of patient loss indices, a complete index corresponding to each patient can be obtained. The complete index of each patient and the corresponding mental state label and agitation label are used as a key index sample to obtain a key index sample set.
With continued reference to fig. 2, at step 204, an airway intervention instrument signature and an airway severity corresponding to the airway intervention instrument signature for each sample in the sample set of key metrics is determined.
Airway severity is set according to the intervention used, such as: no airway intervention means the least severe, airway severity is set to 1; the intervention means is nasal catheter oxygen inhalation, and the severity of the airway is set to be 2; the intervention means was either intubation or tracheotomy, indicating the highest severity, set at 10.
And step 205, training according to the key index sample set and the severity of the airway to obtain an airway obstruction severity evaluation model.
The airway obstruction severity assessment model can be iteratively trained using an Adam optimizer, using MAE (mean absolute error) as a loss function. The MAE formula is defined as follows:
wherein, N is the data volume,
indicating the true severity of the obstruction of the patient's airway,
the predicted severity of the patient's airway obstruction is represented, and by minimizing the MAE value, the model is made to fit better to the training data.
It should be noted that the patient or patient mentioned in the above modeling process refers to a patient in a general sense, i.e. a patient with a medical record, and not to a specific patient or a certain class of patients.
According to the method for evaluating the severity of the airway obstruction provided by the embodiment of the invention, the model for evaluating the severity of the airway obstruction is utilized, and aiming at the current patient, related rescue workers can be assisted to quickly and correctly evaluate and timely and effectively treat the airway obstruction of the patient, so that the rescue ability under emergency and sudden conditions is improved.
Fig. 4 is a flowchart of a method for evaluating the severity of airway obstruction according to an embodiment of the present invention, which includes the following steps:
The point-of-care diagnostic data includes: basic vital signs, detection data, consciousness status labels, etc.
Similar to the extraction of a key index set when an airway obstruction severity evaluation model is established, for the current patient, key index characteristic data is extracted from the instant diagnosis data of the current patient. The key index features mainly include indexes shown in table 1:
and 403, inputting the key index characteristic data into an airway obstruction severity evaluation model, and obtaining an airway obstruction severity score of the patient according to the output of the airway obstruction severity evaluation model.
The output of the model is the airway obstruction severity score, which can be a floating point type number.
The method for evaluating the severity of the airway obstruction provided by the embodiment of the invention is characterized in that an airway obstruction severity evaluation model is constructed in advance based on a medical data source of a patient, key index features are extracted from instant diagnosis data of the current patient, then the key index features are input into the airway obstruction severity evaluation model, and the airway obstruction severity score of the patient is obtained according to the output of the airway obstruction severity evaluation model. Thereby supplementary first aid medical personnel of intelligence whether have the air flue to block the disease to the patient and carry out quick accurate judgement, guide the treatment personnel to block quick accurate aassessment and timely effective processing to the air flue, improve the air flue and block the efficiency of treatment.
Correspondingly, the embodiment of the invention also provides an airway obstruction severity assessment system, which is a structural schematic diagram as shown in fig. 5.
In this embodiment, the system includes the following modules:
a data obtaining module 501, configured to obtain instant diagnosis data of a current patient;
a feature extraction module 502, configured to extract key index feature data from the instant diagnosis data;
and the evaluation module 503 is configured to input the key index feature data into an airway obstruction severity evaluation model constructed in advance based on a patient medical data source, and obtain an airway obstruction severity score of the patient according to an output of the airway obstruction severity evaluation model.
In this embodiment of the present invention, the airway obstruction severity assessment model comprises an input layer, three hidden layers, an output layer, wherein the three hidden layers are fully connected with the output layer, and a random inactivation mechanism is used between adjacent hidden layers.
The airway obstruction severity evaluation model can be established by a corresponding model construction module, and fig. 6 is a structural schematic diagram of the model construction module in the embodiment of the invention.
The model building module comprises:
the data acquisition unit 601 is used for acquiring a medical data source of a patient and extracting data related to the patient visit from the medical data source;
a data analysis unit 602, configured to perform correlation analysis on the patient visit related data to obtain a key index set;
a data processing unit 603, configured to perform normalization processing and missing value filling on the indexes in the key index set, so as to obtain a key index sample set;
a label determining unit 604, configured to determine an airway intervention means label of each sample in the key index sample set and an airway severity corresponding to the airway intervention means label;
and the training unit 605 is configured to train according to the key index sample set and the severity of the airway to obtain an airway obstruction severity evaluation model.
The data analysis unit 602 may specifically include the following sub-units:
the data analysis subunit is used for carrying out correlation analysis on the patient detection indexes to obtain a candidate key index set;
the sorting processing subunit is used for sorting each candidate index in the candidate key index set according to the absolute value of the correlation coefficient between the candidate index and the severity of the airway obstruction, selecting the candidate index with the absolute value of the correlation coefficient greater than or equal to a set threshold value to add into the recommendation index set, and otherwise adding into the residual index set;
the screening subunit is used for grading the importance degree of each index in the recommendation index set by combining medical knowledge and adding the indexes in the graded recommendation index set into a key index set; selecting part of indexes which have great medical significance to the disease condition from the residual index set, and adding the selected indexes into the key index set;
and the label adding subunit is used for adding a mental state label and a restless label to each training sample in the key index set.
It should be noted that, in the specific implementation, the model building module may be a part of the system of the present invention, or may be independent of the system of the present invention, and the embodiment of the present invention is not limited thereto. In addition, more contents on the principle and working mode of each unit and subunit in the model building module may refer to the related description in the foregoing embodiment of the method of the present invention, and are not described herein again.
The airway obstruction severity evaluation system provided by the embodiment of the invention is characterized in that an airway obstruction severity evaluation model is constructed in advance based on a patient medical data source, key index features are extracted from instant diagnosis data of a current patient, then the key index features are input into the airway obstruction severity evaluation model, and an airway obstruction severity score of the patient is obtained according to the output of the airway obstruction severity evaluation model. Thereby supplementary first aid medical personnel of intelligence whether have the air flue to block the disease to the patient and carry out quick accurate judgement, guide the treatment personnel to block quick accurate aassessment and timely effective processing to the air flue, improve the air flue and block the efficiency of treatment.
The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. Furthermore, the above-described system embodiments are merely illustrative, wherein modules and units illustrated as separate components may or may not be physically separate, i.e., may be located on one network element, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Those skilled in the art will appreciate that all or part of the steps in the above method embodiments may be implemented by a program to instruct relevant hardware to perform the steps, and the program may be stored in a computer-readable storage medium, referred to herein as a storage medium, such as: ROM/RAM, magnetic disk, optical disk, etc.
Accordingly, the embodiment of the present invention further provides a system for an airway obstruction severity assessment method, where the system is an electronic device, such as a mobile terminal, a computer, a tablet device, a personal digital assistant, and the like. The electronic device may include one or more processors, memory; wherein the memory is used for storing computer executable instructions and the processor is used for executing the computer executable instructions to realize the method of the previous embodiments.
The foregoing detailed description of the embodiments of the present invention has been presented for purposes of illustration and description, and is intended to be exemplary only and is not intended to be exhaustive or to be exhaustive of the embodiments of the invention. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present invention without any creative effort shall fall within the protection scope of the present invention, and the content of the present description shall not be construed as limiting the present invention. Therefore, any modification, equivalent replacement, improvement and the like made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. A method for assessing the severity of airway obstruction, the method comprising:
acquiring instant diagnosis data of a current patient;
extracting key index characteristic data from the instant diagnosis data;
inputting the key index characteristic data into an airway obstruction severity evaluation model which is constructed in advance based on a patient medical data source, and obtaining an airway obstruction severity score of the patient according to the output of the airway obstruction severity evaluation model;
the airway obstruction severity evaluation model comprises an input layer, three hidden layers and an output layer, wherein the three hidden layers are fully connected with the output layer, and a random inactivation mechanism is used between the adjacent hidden layers.
2. The method of claim 1, wherein constructing the airway obstruction severity assessment model based on the patient medical data source comprises:
acquiring a patient medical data source and extracting patient treatment related data from the medical data source;
performing correlation analysis on the patient diagnosis related data to obtain a key index set;
carrying out normalization processing and missing value filling on the indexes in the key index set to obtain a key index sample set;
determining an airway intervention means label and an airway severity corresponding to the airway intervention means label of each sample in the key index sample set;
and training according to the key index sample set and the severity of the air passage to obtain an air passage obstruction severity evaluation model.
3. The method of claim 2, wherein the patient medical data source comprises any one or more of: an emergency database, a nursing system and a hospital database.
4. The method of claim 2, wherein the patient encounter-related data comprises any one or more of: patient basic information, patient airway intervention records, patient detection indexes and consciousness state information of the patient;
the patient airway intervention record comprises: nasal catheter oxygen inhalation, mask oxygen inhalation, oropharyngeal airway or tracheal intubation;
the patient detection indicators include: vital signs, blood routine, blood and qi analysis, blood coagulation function, and emergency biochemistry;
the consciousness state information of the patient includes: mental state, whether restlessness.
5. The method of claim 4, wherein performing a correlation analysis on the patient visit related data to obtain a set of key indicators comprises:
carrying out correlation analysis on the patient detection indexes to obtain a candidate key index set;
sorting each candidate index in the candidate key index set and the correlation coefficient of the severity of the airway obstruction according to the absolute value, selecting the candidate index of which the absolute value of the correlation coefficient is more than or equal to a set threshold value to add into the recommended index set, and otherwise adding into the residual index set;
classifying the importance degree of each index in the recommendation index set by combining medical knowledge, and adding the indexes in the recommendation index set into a key index set;
selecting part of indexes from the residual index set and adding the selected part of indexes into the key index set;
adding a mental state label and a restlessness label to each training sample in the set of key indicators.
6. The method of claim 2, wherein the normalizing the metrics in the set of key metrics comprises:
if the index is in a normal interval, taking the index as 0;
if the index is higher than the normal interval and lower than the extreme danger upper limit value, mapping the index value into a value in an interval of 0 to 1;
if the index is higher than or equal to the extreme danger upper limit value, taking the index as 1;
if the index is lower than the normal interval and higher than the extremely dangerous lower limit value, mapping the index value into a value in an interval from 0 to-1;
and if the index is lower than or equal to the extreme danger lower limit value, taking the index as-1.
7. The method of claim 2, wherein training the airway obstruction severity assessment model according to the sample set of key indicators and the airway intervention tool label comprises:
and according to the key index sample set and the airway intervention means label, adopting MAE as a loss function, and using an Adam optimizer for iterative training to obtain an airway obstruction severity evaluation model.
8. An airway obstruction severity assessment system, the system comprising:
the data acquisition module is used for acquiring the instant diagnosis data of the current patient;
the characteristic extraction module is used for extracting key index characteristic data from the instant diagnosis data;
the evaluation module is used for inputting the key index characteristic data into an airway obstruction severity evaluation model which is constructed in advance based on a patient medical data source, and obtaining the airway obstruction severity score of the patient according to the output of the airway obstruction severity evaluation model;
the airway obstruction severity evaluation model comprises an input layer, three hidden layers and an output layer, wherein the three hidden layers are fully connected with the output layer, and a random inactivation mechanism is used between the adjacent hidden layers.
9. The system of claim 8, further comprising:
the model building module is used for building an airway obstruction severity evaluation model; the model building module comprises:
the data acquisition unit is used for acquiring a medical data source of a patient and extracting the data related to the patient treatment from the medical data source;
the data analysis unit is used for carrying out correlation analysis on the patient diagnosis related data to obtain a key index set;
the data processing unit is used for carrying out normalization processing and missing value filling on the indexes in the key index set to obtain a key index sample set;
the label determining unit is used for determining the airway intervention means label of each sample in the key index sample set and the airway severity corresponding to the airway intervention means label;
and the training unit is used for training according to the key index sample set and the severity of the air passage to obtain an air passage obstruction severity evaluation model.
10. The system of claim 9, wherein the data analysis unit comprises:
the data analysis subunit is used for carrying out correlation analysis on the patient detection indexes to obtain a candidate key index set;
the sorting processing subunit is used for sorting each candidate index in the candidate key index set according to the absolute value of the correlation coefficient between the candidate index and the severity of the airway obstruction, selecting the candidate index with the absolute value of the correlation coefficient greater than or equal to a set threshold value to add into the recommendation index set, and otherwise adding into the residual index set;
the screening subunit is used for grading the importance degree of each index in the recommendation index set by combining medical knowledge and then adding the indexes in the graded recommendation index set into the key index set;
selecting part of indexes from the residual index set and adding the selected part of indexes into the key index set;
and the label adding subunit is used for adding a mental state label and a restless label to each training sample in the key index set.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111455643.1A CN114155955A (en) | 2021-12-01 | 2021-12-01 | Airway obstruction severity assessment method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111455643.1A CN114155955A (en) | 2021-12-01 | 2021-12-01 | Airway obstruction severity assessment method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114155955A true CN114155955A (en) | 2022-03-08 |
Family
ID=80455681
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111455643.1A Pending CN114155955A (en) | 2021-12-01 | 2021-12-01 | Airway obstruction severity assessment method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114155955A (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109273093A (en) * | 2018-09-14 | 2019-01-25 | 苏州贝斯派生物科技有限公司 | A kind of construction method and building system of Kawasaki disease risk evaluation model |
| CN111276242A (en) * | 2020-01-20 | 2020-06-12 | 吉林大学 | Disease diagnosis and disease state evaluation modeling method for patients in intensive care unit of hospital |
| CN112365978A (en) * | 2020-11-10 | 2021-02-12 | 北京航空航天大学 | Method and device for establishing early risk assessment model of tachycardia event |
| CN113077900A (en) * | 2021-03-11 | 2021-07-06 | 杭州市妇产科医院 | Diabetes early risk assessment method, device, computer equipment and medium |
| CN113517066A (en) * | 2020-08-03 | 2021-10-19 | 东南大学 | Depression assessment method and system based on candidate gene methylation sequencing and deep learning |
-
2021
- 2021-12-01 CN CN202111455643.1A patent/CN114155955A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109273093A (en) * | 2018-09-14 | 2019-01-25 | 苏州贝斯派生物科技有限公司 | A kind of construction method and building system of Kawasaki disease risk evaluation model |
| CN111276242A (en) * | 2020-01-20 | 2020-06-12 | 吉林大学 | Disease diagnosis and disease state evaluation modeling method for patients in intensive care unit of hospital |
| CN113517066A (en) * | 2020-08-03 | 2021-10-19 | 东南大学 | Depression assessment method and system based on candidate gene methylation sequencing and deep learning |
| CN112365978A (en) * | 2020-11-10 | 2021-02-12 | 北京航空航天大学 | Method and device for establishing early risk assessment model of tachycardia event |
| CN113077900A (en) * | 2021-03-11 | 2021-07-06 | 杭州市妇产科医院 | Diabetes early risk assessment method, device, computer equipment and medium |
Non-Patent Citations (7)
| Title |
|---|
| JORGE L.M. AMARAL等: "Machine learning algorithms and forced oscillationmeasurements to categorise the airway obstruction severity in chronic obstructive pulmonary disease", 《COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE》 * |
| YOUFANG HAN: "Intelligent injury prediction for traumatic airway obstruction", 《MEDICAL & BIOLOGICAL ENGINEERING & COMPUTING》 * |
| 刘星等: "重症超声评估急性呼吸窘迫综合征患者病情严重程度及预后的价值分析", 《中国医药导报》 * |
| 樊寻梅等: "TBFV环用于诊断小儿上气道梗阻(附8例报告)", 《实用儿科临床杂志》 * |
| 王芳等: "呼吸道梗阻风险评估表在预防颈椎前路手术患者术后呼吸道梗阻中的应用", 《中国实用医药》 * |
| 王远征: "流速-容量环辅助诊断上气道梗阻一例", 《实用临床医学》 * |
| 胡红梅等: "气管镜在ICU诊断上气道阻塞一例并文献复习", 《健康之路》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107438399B (en) | Cardiovascular deterioration early warning score | |
| Huddar et al. | Predicting complications in critical care using heterogeneous clinical data | |
| CN110459328B (en) | Clinical monitoring equipment | |
| US20120239435A1 (en) | Automated annotation of clinical data | |
| CN110289061B (en) | Time sequence prediction method for traumatic hemorrhagic shock injury | |
| US20110046972A1 (en) | Method and system for health-centered medicine | |
| CN111261281A (en) | Model training method, mechanical ventilation phenotype identification method and breathing machine | |
| RU2698007C1 (en) | Automatic system for supporting medical solutions in a combined pathology | |
| Ho et al. | Imputation-enhanced prediction of septic shock in ICU patients | |
| Huddar et al. | Predicting postoperative acute respiratory failure in critical care using nursing notes and physiological signals | |
| CN116631562B (en) | Method and device for generating discharge records of electronic medical records and electronic equipment | |
| CN114023441A (en) | Severe AKI early risk assessment model and device based on interpretable machine learning model and development method thereof | |
| CN107391901A (en) | Establish the method and server of public ward conditions of patients assessment models | |
| Al-Mualemi et al. | A deep learning-based sepsis estimation scheme | |
| CN111180026A (en) | Special diagnosis and treatment view system and method | |
| Islam et al. | Predictive analysis for risk of stroke using machine learning techniques | |
| Ramgopal et al. | Empirically derived age-based vital signs for children in the out-of-hospital setting | |
| KR102169637B1 (en) | Method for predicting of mortality risk and device for predicting of mortality risk using the same | |
| CN112967803A (en) | Early mortality prediction method and system for emergency patients based on integrated model | |
| CN114155955A (en) | Airway obstruction severity assessment method and system | |
| CN114068016A (en) | Airway obstruction intervention recommendation method and system | |
| CN115862897A (en) | Syndrome monitoring method and system based on clinical data | |
| Wu et al. | Study of the relationship between ICU patient recovery and tcm treatment in acute phase: a retrospective study based on Python data mining technology | |
| Berge et al. | Combining unsupervised, supervised, and rule-based algorithms for text mining of electronic health records-a clinical decision support system for identifying and classifying allergies of concern for anesthesia during surgery | |
| Saleh | Artificial intelligence and statistical techniques to predict probability of injury survival |
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 |