A data interaction method for virtual simulation of anesthesiology crisis skills
By slicing and clustering the physiological parameter data of virtual patients, the problems of complexity and low efficiency of data management in traditional data storage methods are solved, efficient data management and storage are realized, and the integrity of virtual patient data is ensured. and analytical convenience.
Patent Information
- Application Number
- CN202410377975.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-03-29
- Publication Date
- 2024-12-13
- Estimated Expiration
- 2044-03-29
AI Technical Summary
The traditional data storage method of virtual simulation of anesthesia crisis skills adopts fixed-sized slice processing, which increases the complexity of data management, reduces the data management and storage efficiency, and affects the efficiency of data interaction.
By slicing the physiological parameter data of virtual patients and splitting and merging according to the similarity of the data, the slice data is classified and stored using clustering algorithm to ensure the integrity and efficiency of the data.
It improves the management and storage efficiency of virtual patient data, ensures the integrity of data, facilitates learners to analyze physiological data of virtual patients, and improves the efficiency of data interaction.
Smart Images
Figure CN118277802B_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of data processing, and in particular to a data interaction method for virtual simulation of anesthesiology crisis skills. Background Art
[0002] Data interaction in virtual simulation of anesthesiology crisis skills is very important, as it allows learners to interact with the virtual environment, simulate crisis scenarios, and obtain feedback. Using virtual reality headsets, learners can immerse themselves in virtual anesthesia operating rooms or emergency situations, and interact with objects and patients in the virtual environment through handles or gesture control. Since a large amount of data is generated during the virtual simulation process, including the physiological parameters of virtual patients, learners' operation records, etc., in order to facilitate the storage of these data and facilitate interaction.
[0003] Traditionally, when storing data for virtual simulation of anesthesiology crisis skills, the collected data is usually sliced into fixed-size slices. If the number of slices is too small, the amount of data in each slice will be too much, making it difficult to analyze the data. If the number of slices is too large, the management complexity will increase, reducing the efficiency of data management and storage, and thus affecting the efficiency of data interaction. Summary of the invention
[0004] In order to solve the above problems, the present invention provides a data interaction method for virtual simulation of anesthesiology crisis skills.
[0005] A data interaction method for virtual simulation of anesthesiology crisis skills of the present invention adopts the following technical solution:
[0006] An embodiment of the present invention provides a data interaction method for virtual simulation of anesthesiology crisis skills, the method comprising the following steps:
[0007] Acquire physiological parameter data of a virtual patient, and divide the physiological parameter data according to time to obtain a plurality of slice data of the physiological parameter data, wherein the slice data includes a plurality of physiological indicator data, and the physiological indicator data includes a plurality of data points;
[0008] According to the physiological indicator value of each data point in each physiological indicator data of the slice data, a number of slice data to be split are obtained; according to the physiological indicator values of adjacent data points in each physiological indicator data of the slice data to be split, a number of splitting nodes of each physiological indicator data of each slice data to be split are obtained; according to the order of the splitting nodes in the physiological indicator data, a number of sub-slice data of each slice data to be split are obtained;
[0009] According to the lengths and sub-slice data of two adjacent slice data, first slice data and second slice data of two adjacent slice data are obtained; according to the physiological index value of each data point in each item of physiological index data in the first slice data and the second slice data, similarity of two adjacent slice data is obtained; threshold judgment is performed on the similarity of all two adjacent slice data, and several merged slice data of physiological parameter data are obtained by merging;
[0010] All merged slice data of physiological parameter data are clustered and stored in categories.
[0011] Furthermore, the step of obtaining a plurality of slice data to be split according to the physiological indicator value of each data point in each item of physiological indicator data of the slice data comprises the following specific steps:
[0012]
[0013] In the formula, x a,i,j is the physiological index value of the jth data point in the ith physiological index data of the ath slice data, is the mean value of the physiological index values of all data points in the ith physiological index data of the ath slice data, n a,i is the number of data points in the i-th physiological index data of the a-th slice data, N a is the total number of physiological index data in the ath slice data, norm[] is the linear normalization function, Q a is the discrete degree of the ath slice data;
[0014] The discrete degree of each slice data is obtained, and a number of slice data to be split are obtained through threshold judgment.
[0015] Furthermore, the step of obtaining a plurality of slice data to be split by threshold judgment includes the following specific steps:
[0016] A first threshold is preset, denoted as th1, if Q a ≥th1, the ath slice data is taken as a slice data to be split.
[0017] Furthermore, the step of obtaining a plurality of splitting nodes for each item of physiological indicator data of each slice data to be split according to the physiological indicator values of adjacent data points in each item of physiological indicator data of the slice data to be split includes the following specific steps:
[0018]
[0019] In the formula, S b,p,q is the variance of the physiological index values of the first q data points in the p-th physiological index data of the b-th slice data to be split, S b,p,q-1is the variance of the physiological index values of the first q - 1 data points in the p-th physiological index data of the b-th slice data to be split, where q is a reference value, C b,p,q is the degree of outlier of the q-th data point in the p-th physiological index data of the b-th slice data to be split, norm() is a linear normalization function, and || is to take the absolute value;
[0020] Obtain the degree of outlier of each data point in each physiological index data of each slice data to be split, and obtain several splitting nodes of each physiological index data of each slice data to be split through threshold judgment.
[0021] Furthermore, the specific steps included in obtaining several splitting nodes of each physiological index data of each slice data to be split through threshold judgment are as follows:
[0022] Preset a second threshold, denoted as th2. If C b,p,q ≥ th2, take the q-th data point as a splitting node of the p-th physiological index data of the b-th slice data to be split, and start from the (q + 1)-th data point in the p-th physiological index data to re-select the first q data points, obtain the degree of outlier of the 2×q-th data point and compare it with th2. If the degree of outlier of the 2×q-th data point is greater than or equal to th2, also take the 2×q-th data point as a splitting node, and finally obtain several splitting nodes of the p-th physiological index data of the b-th slice data to be split; If C b,p,q < th2, increase the value of the reference value q by one, obtain the degree of outlier of the q-th data point after the change of the q value and continue to perform threshold judgment until the degree of outlier of the data point is greater than or equal to th2 to obtain a splitting node, and continue to calculate the degree of outlier of the data points after the change of the q value and perform threshold judgment backward, and finally obtain several splitting nodes of the p-th physiological index data of the b-th slice data to be split.
[0023] Furthermore, the specific steps included in obtaining several sub-slice data of each slice data to be split according to the order of the splitting nodes in the physiological index data are as follows:
[0024] Denote any slice data to be split as the target slice data to be split, obtain the index of each splitting node in each physiological index data of the target slice data to be split, where the index is the order of the splitting node in the physiological index data, denote the sequence composed of all the indexes of the splitting nodes as the index sequence, use the DBSCAN clustering algorithm to cluster the index sequence to obtain several clusters, denote the cluster with the largest number of indexes as the target cluster, and denote the mean value of all the indexes in the target cluster as Denote Round up to obtain the splitting node of the target slice data to be split, split to obtain two sub-slice data of the target slice data to be split, obtain the degree of dispersion of each sub-slice data. If the degree of dispersion of the sub-slice data is greater than or equal to the first threshold, take the sub-slice data as a slice data to be split and split it again to obtain two sub-slice data, and finally obtain several sub-slice data of the target slice data to be split; if the degree of dispersion of the sub-slice data is less than the first threshold, the sub-slice data will not be split again.
[0025] Further, the obtaining of the first slice data and the second slice data of two adjacent slice data according to the lengths of two adjacent slice data and the sub-slice data includes the following specific steps:
[0026] For any two adjacent slice data on the left and right, denote the slice data on the left as slice data A and the slice data on the right as slice data B. If slice data A has multiple sub-slice data, take the sub-slice data closest to slice data B as the first initial slice data; if slice data A has no sub-slice data, take slice data A as the first initial slice data; if slice data B has multiple sub-slice data, take the sub-slice data closest to slice data A as the second initial slice data; if slice data B has no sub-slice data, take slice data B as the second initial slice data;
[0027] Denote the length of the first initial slice data as L1 and the length of the second initial slice data as L2. If L1 > L2, obtain the slice data adjacent to the second initial slice data and with a length of L2 in the first initial slice data as the first slice data, and directly take the second initial slice data as the second slice data; if L1 < L2, obtain the slice data adjacent to the first initial slice data and with a length of L1 in the second initial slice data as the second slice data, and directly take the first initial slice data as the first slice data; if L1 = L2, take the first initial slice data as the first slice data and the second initial slice data as the second slice data; thus obtaining the first slice data and the second slice data of slice data A and slice data B.
[0028] Further, the obtaining of the similarity of two adjacent slice data according to the physiological index values of each data point in each physiological index data of the first slice data and the second slice data includes the following specific steps:
[0029]
[0030] In the formula, m 1,i,k is the physiological index value of the k-th data point in the i-th physiological index data of the first slice data, is the mean value of the physiological index values of all data points in the i-th physiological index data of the first slice data, m2,i,k is the physiological index value of the kth data point in the i-th item of physiological index data of the second slice data, is the mean value of the physiological index values of all data points in the ith physiological index data of the second slice data, num i is the number of data points in the i-th physiological index data, N is the total number of physiological index data of the slice data, norm{} is the linear normalization function, R 1,2 is the similarity between slice data A and slice data B.
[0031] Furthermore, the similarity between all two adjacent slice data is judged by threshold value, and the merged slice data of the physiological parameter data is obtained by merging, and the specific steps include the following:
[0032] A third threshold is preset, denoted as th3. If R 1,2 ≥th3, merge the slice data A and the slice data B in chronological order to obtain a merged slice data, compare the similarities of all two adjacent slice data with the third threshold, and merge to obtain multiple merged slice data; for several slice data that have not been merged, take each slice data that has not been merged as a merged slice data, and finally obtain several merged slice data of physiological parameter data.
[0033] Furthermore, the clustering and classified storage of all merged slice data of the physiological parameter data includes the following specific steps:
[0034] Any merged slice data is recorded as the target merged slice data, the standard deviation of the physiological indicator values of all data points in each physiological indicator data of the target merged slice data is obtained, and several standard deviations of the target merged slice data are obtained, and all standard deviations are arranged according to the order of obtaining the physiological indicator data to obtain the standard deviation sequence of the target merged slice data; the standard deviation sequence of each merged slice data is obtained, and all merged slice data are clustered according to the standard deviation sequence by the ISODATA algorithm, and the distance measurement adopts the DTW distance between the standard deviation sequences of the merged slice data to obtain several slice clusters, and the merged slice data in each slice cluster is stored separately.
[0035] The beneficial effect of the technical solution of the present invention is that the traditional data slice storage method uses slice data of a fixed size. In virtual patient simulation data, the slice data of a fixed size will destroy the integrity of physiological data, which is not conducive to learners' analysis of physiological patients.
[0036] The present invention obtains slice data by slicing the data, and continuously splits and merges the slice data, so that data with a large degree of similarity are in the same slice, and clusters the slice data, and classifies and stores the slice data. The present invention ensures the integrity of the data of the virtual patient in a continuous time, so as to facilitate the learner to analyze the different physiological data of the virtual patient caused by various operations, and improve the management and storage of physiological parameter data information. BRIEF DESCRIPTION OF THE DRAWINGS
[0037] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0038] Figure 1 A flowchart of the steps of a data interaction method for virtual simulation of anesthesiology crisis skills provided by an embodiment of the present invention. DETAILED DESCRIPTION
[0039] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the data interaction method for virtual simulation of anesthesiology crisis skills proposed by the present invention, its specific implementation method, structure, characteristics and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.
[0040] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
[0041] The following is a detailed description of a data interaction method for virtual simulation of anesthesiology crisis skills provided by the present invention in conjunction with the accompanying drawings.
[0042] See also Figure 1 , which shows a step flow chart of a data interaction method for virtual simulation of anesthesiology crisis skills provided by an embodiment of the present invention, the method comprising the following steps:
[0043] Step S001: Acquire physiological parameter data of a virtual patient, and divide the physiological parameter data according to time to obtain a plurality of slice data of the physiological parameter data.
[0044] It should be noted that the purpose of this embodiment is to cluster and fragment the virtual patient physiological parameter data for storage to improve data management and storage efficiency. Before starting processing and analysis, the data needs to be collected and preprocessed.
[0045] Specifically, the physiological parameter data of the virtual patient is obtained through the physiological model, wherein the physiological parameter data includes multiple physiological index data: heart rate, blood pressure, respiratory rate, blood oxygen saturation, body temperature, a total of 5 physiological index data, wherein each physiological index data contains multiple data points, and each data point represents the physiological index value of the corresponding physiological index data at each moment; the obtained physiological parameter data is divided into a slice data every minute in the time sequence, and multiple slice data of the physiological parameter data are obtained. It should be noted that if the physiological parameter data is divided to the last less than one minute, the remaining physiological parameter data is directly used as a slice data.
[0046] It should be noted that the acquisition frequency of the physiological parameter data of the virtual patient through the physiological model is 1 time / second. The physiological parameter data of the virtual patient refers to the physiological parameter data of the patient simulated by the physiological model in the virtual simulation environment. The acquired physiological parameter data of the virtual patient is the data sent by the sending end of the interactive data. The data analysis module is the receiving end of the interactive data, which is used to receive and analyze the physiological parameter data of the virtual patient sent by the sending end.
[0047] At this point, the physiological parameter data of the virtual patient and multiple slice data of the physiological parameter data are obtained.
[0048] Step S002: obtain a number of slice data to be split according to the physiological indicator value of each data point in each physiological indicator data of the slice data; obtain a number of splitting nodes for each physiological indicator data of each slice data to be split according to the physiological indicator values of adjacent data points in each physiological indicator data of the slice data to be split; and obtain a number of sub-slice data for each slice data to be split according to the order of the splitting nodes in the physiological indicator data.
[0049] It should be noted that the purpose of slicing the physiological parameter data is to facilitate the management of the physiological parameter data. However, due to the pre-slicing of the physiological parameter data, the slicing is only performed based on time. During the virtual operation of the learner, the virtual patient will present different physiological characteristics. Ultimately, the physiological characteristics of the virtual patient need to be analyzed to evaluate the learner's operation. Therefore, in the analysis process, it is more convenient to assign data with higher similarity to the same slice, so it is necessary to analyze each slice data, continuously split and merge the slice data, and assign data with higher similarity to the same slice data.
[0050] It should be noted that since the slice data of the physiological parameter data is sliced according to time, the change characteristics of the data are not taken into account. Therefore, there will be data with large differences in the physiological index values in the same slice data. When analyzing the data, it is more desirable to analyze similar data and judge the state of the virtual patient based on whether the time of the similar data obeys the periodic distribution. Therefore, it is necessary to perform adaptive splitting on the slice data with large differences.
[0051] Specifically, according to the physiological index value of each data point in each physiological index data of the slice data, a number of slice data to be split are obtained, as follows:
[0052]
[0053] In the formula, x a,i,j is the physiological index value of the jth data point in the ith physiological index data of the ath slice data, is the mean value of the physiological index values of all data points in the ith physiological index data of the ath slice data, n a,i is the number of data points in the i-th physiological index data of the a-th slice data, N a is the total number of physiological index data of the ath slice data, norm[] is a linear normalization function, and the normalized object is all slice data Q a is the discrete degree of the ath slice data.
[0054] It should be noted that It represents the discrete degree of the i-th physiological index data of the a-th slice data. The larger its value is, the higher the discrete degree of the i-th physiological index data is, which leads to a higher discrete degree of the a-th slice data.
[0055] Furthermore, the discrete degree of each slice data is obtained, and a number of slice data to be split are obtained through threshold judgment, as follows:
[0056] A first threshold is preset, denoted as th1. This embodiment takes th1=0.6 as an example. If Q a ≥th1, the ath slice data is taken as a slice data to be split.
[0057] It should be noted that Q a ≥th1 indicates that the ath slice data has a high degree of discreteness and needs to be subsequently split. Therefore, the ath slice data is taken as a slice data to be split.
[0058] It should be noted that the splitting nodes of each physiological index data item are calculated by analyzing the discrete degree change of each physiological index data item in each slice data to be split, and the range between every two splitting nodes is recorded as a splitting interval.
[0059] Specifically, according to the physiological indicator values of adjacent data points in each physiological indicator data of each slice data to be split, a number of splitting nodes of each physiological indicator data of each slice data to be split are obtained, as follows:
[0060]
[0061] In the formula, S b,p,q is the variance of the physiological index values of the first q data points in the p-th physiological index data of the b-th slice data to be split, S b,p,q-1 is the variance of the physiological index value of the first q-1 data points in the p-th physiological index data of the b-th slice data to be split, q is a reference value, and the initial value of the reference value q is preset to 6 in this embodiment for description, C b,p,q is the outlier degree of the qth data point in the pth physiological index data of the bth slice data to be split, norm() is a linear normalization function, and the normalized object is all the data points in the pth physiological index data of the bth slice data to be split. || is the absolute value.
[0062] It should be noted that all data points are all data points determined after the reference value, and the purpose of introducing the reference value is to prevent too little data from having a significant impact on the calculation results.
[0063] Furthermore, the degree of outlier of each data point in each physiological indicator data of each slice data to be split is obtained, and several splitting nodes of each physiological indicator data of each slice data to be split are obtained by threshold judgment, as follows:
[0064] A second threshold is preset, denoted as th2. This embodiment takes th2=0.7 as an example. If C b,p,q ≥th2, C b,p,q is the outlier degree of the qth data point in the pth physiological indicator data of the bth slice data to be split, the qth data point is used as a splitting node of the pth physiological indicator data of the bth slice data to be split, and the first q data points are reselected from the q+1th data point in the pth physiological indicator data, and the outlier degree of the 2×qth data point is obtained and compared with th2. If the outlier degree of the 2×qth data point is greater than or equal to th2, the 2×qth data point is also used as a splitting node, and finally several splitting nodes of the pth physiological indicator data of the bth slice data to be split are obtained; if C b,p,q<th2, increment the value of the reference value q by one, and obtain the outlier degree of the q-th data point after the change of the q value and continue the threshold judgment until the outlier degree of the data point is greater than or equal to th2, obtaining a splitting node, and continue to calculate the outlier degree of the data points after the change of the q value and the threshold judgment backward, finally obtaining several splitting nodes of the p-th physiological index data of the b-th slice data to be split.
[0065] Further, according to the order of the splitting nodes in the physiological index data, several sub-slice data of each slice data to be split are obtained, specifically as follows:
[0066] Denote any slice data to be split as the target slice data to be split, obtain the index of each splitting node in each physiological index data of the target slice data to be split, where the index is the order of the splitting node in the physiological index data, denote the sequence composed of all the indexes of the splitting nodes as the index sequence, use the DBSCAN clustering algorithm to cluster the index sequence, obtain several clusters, denote the cluster with the largest number of indexes as the target cluster, and denote the mean value of all the indexes in the target cluster as Take Round up as the splitting node of the target slice data to be split, split to obtain two sub-slice data of the target slice data to be split, obtain the dispersion degree of each sub-slice data, if the dispersion degree of the sub-slice data is greater than or equal to the first threshold, take the sub-slice data as a slice data to be split and split again to obtain two sub-slice data, finally obtaining several sub-slice data of the target slice data to be split; if the dispersion degree of the sub-slice data is less than the first threshold, the sub-slice data will not be split anymore.
[0067] Thus, several sub-slice data of any slice data to be split are obtained.
[0068] Step S003: According to the lengths of two adjacent slice data and the sub-slice data, obtain the first slice data and the second slice data of the two adjacent slice data, according to the physiological index values of each data point in each physiological index data of the first slice data and the second slice data, obtain the similarity of the two adjacent slice data, perform threshold judgment on the similarities of all two adjacent slice data, and merge to obtain several merged slice data of the physiological parameter data.
[0069] It should be noted that the above steps perform adaptive splitting processing on all the slice data to be split. When slicing the physiological parameter data in Step 1, only slicing is performed according to time, which cannot ensure the similarity of the data in the same slice data. Therefore, it is necessary to perform adaptive splitting processing on the slice data to be split to obtain multiple sub-slice data. Since in the simulation process, the physiological characteristics of the virtual patient will have a certain degree of stability, there are similar characteristics within a certain range. In the above steps, the data is adaptively split according to the degree of dispersion of the data in the slice data to be split. There may be a certain similarity between adjacent slice data, and the data with a greater similarity can be merged to reduce the number of slices and facilitate storage.
[0070] Specifically, according to the lengths of two adjacent slice data and the sub-slice data, the first slice data and the second slice data of the two adjacent slice data are obtained as follows:
[0071] For any two adjacent slice data on the left and right, the slice data on the left is denoted as slice data A, and the slice data on the right is denoted as slice data B. If slice data A has multiple sub-slice data, the sub-slice data closest to slice data B is used as the first initial slice data. If slice data A has no sub-slice data, slice data A is used as the first initial slice data; if slice data B has multiple sub-slice data, the sub-slice data closest to slice data A is used as the second initial slice data. If slice data B has no sub-slice data, slice data B is used as the second initial slice data.
[0072] It should be noted that for the obtained first initial slice data and second initial slice data, in order to better perform subsequent similarity analysis, it is necessary to ensure that the lengths of the first initial slice data and the second initial slice data are the same. Therefore, it is necessary to unify the lengths of the first initial slice data and the second initial slice data.
[0073] Specifically, the length of the first initial slice data is denoted as L1, and the length of the second initial slice data is denoted as L2. If L1 > L2, the slice data adjacent to the second initial slice data and with a length of L2 is obtained from the first initial slice data as the first slice data, and the second initial slice data is directly used as the second slice data; if L1 < L2, the slice data adjacent to the first initial slice data and with a length of L1 is obtained from the second initial slice data as the second slice data, and the first initial slice data is directly used as the first slice data; if L1 = L2, the first initial slice data is used as the first slice data, and the second initial slice data is used as the second slice data; thus, the first slice data and the second slice data of slice data A and slice data B are obtained.
[0074] It should be noted that since the two sub-slice data are the physiological characteristic data of the virtual patient in a continuous time, and the physiological characteristic data in a continuous time have local similarities, that is, the closer the distance between the data, the higher the similarity, the part of the slice data with a larger data volume that is closest to the slice data with a smaller data volume is taken for calculation.
[0075] Specifically, according to the physiological index value of each data point in each item of physiological index data in the first slice data and the second slice data, the similarity of two adjacent slice data is obtained, as follows:
[0076]
[0077] In the formula, m 1,i,k is the physiological index value of the kth data point in the ith item of physiological index data of the first slice data, is the mean value of the physiological index values of all data points in the ith physiological index data of the first slice data, m 2,i,k is the physiological index value of the kth data point in the i-th item of physiological index data of the second slice data, is the mean value of the physiological index values of all data points in the ith physiological index data of the second slice data, num i is the number of data points in the i-th physiological index data, N is the total number of physiological index data of the slice data, norm{} is a linear normalization function, and the normalized object is the data points of all two adjacent slice data. R 1,2 is the similarity between slice data A and slice data B.
[0078] Furthermore, a threshold value is judged on the similarity of all two adjacent slice data, and a plurality of merged slice data of the physiological parameter data are obtained by merging, as follows:
[0079] A third threshold is preset, denoted as th3. This embodiment takes th3=0.7 as an example. If R 1,2 ≥th3, merge the slice data A and the slice data B in chronological order to obtain a merged slice data, compare the similarities of all two adjacent slice data with the third threshold, and merge to obtain multiple merged slice data; for several slice data that have not been merged, take each slice data that has not been merged as a merged slice data, and finally obtain several merged slice data of physiological parameter data.
[0080] At this point, several merged slice data of physiological parameter data are obtained.
[0081] Step S004: clustering and classifying all merged slice data of physiological parameter data.
[0082] Specifically, any merged slice data is recorded as the target merged slice data, the standard deviation of the physiological indicator values of all data points in each physiological indicator data of the target merged slice data is obtained, and several standard deviations of the target merged slice data are obtained, and all standard deviations are arranged in the order of obtaining the physiological indicator data to obtain the standard deviation sequence of the target merged slice data; the standard deviation sequence of each merged slice data is obtained, and all merged slice data are clustered according to the standard deviation sequence by the ISODATA algorithm, and the distance metric uses the DTW distance between the standard deviation sequences of the merged slice data to obtain several slice clusters, and the merged slice data in each slice cluster is stored separately to improve the management and storage efficiency of the physiological parameter data. It should be noted that the merged slice data in the same slice clustering result have certain similarities, and the management and storage efficiency of the physiological parameter data is improved by separate storage, so as to better interact.
[0083] Through the above steps, a data interaction method for virtual simulation of anesthesiology crisis skills is completed.
[0084] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.
Claims
1. A data interaction method for virtual simulation of anesthesiology crisis skills, characterized in that: The method comprises the following steps: Acquire physiological parameter data of a virtual patient, and divide the physiological parameter data according to time to obtain a plurality of slice data of the physiological parameter data, wherein the slice data includes a plurality of physiological indicator data, and the physiological indicator data includes a plurality of data points; According to the physiological indicator value of each data point in each physiological indicator data of the slice data, a number of slice data to be split are obtained; according to the physiological indicator values of adjacent data points in each physiological indicator data of the slice data to be split, a number of splitting nodes of each physiological indicator data of each slice data to be split are obtained; according to the order of the splitting nodes in the physiological indicator data, a number of sub-slice data of each slice data to be split are obtained; According to the lengths and sub-slice data of two adjacent slice data, first slice data and second slice data of two adjacent slice data are obtained; according to the physiological index value of each data point in each item of physiological index data in the first slice data and the second slice data, similarity of two adjacent slice data is obtained; threshold judgment is performed on the similarity of all two adjacent slice data, and several merged slice data of physiological parameter data are obtained by merging; Clustering and classifying all merged slice data of physiological parameter data; The specific steps of obtaining a plurality of slice data to be split according to the physiological indicator value of each data point in each item of physiological indicator data of the slice data are as follows: In the formula, x a,i,j is the physiological index value of the jth data point in the ith physiological index data of the ath slice data, is the mean value of the physiological index values of all data points in the ith physiological index data of the ath slice data, n a,i is the number of data points in the i-th physiological index data of the a-th slice data, N a is the total number of physiological index data in the ath slice data, norm[] is the linear normalization function, Q a is the discrete degree of the ath slice data; Obtain the discrete degree of each slice data, and obtain a number of slice data to be split through threshold judgment; The specific steps of obtaining a number of slice data to be split by threshold judgment are as follows: A first threshold is preset, denoted as th1. If Q a ≥th1, take the ath slice data as a slice data to be split; The specific steps of obtaining a plurality of splitting nodes of each physiological indicator data of each slice data to be split according to the physiological indicator values of adjacent data points in each physiological indicator data of the slice data to be split are as follows: In the formula, S b,p,q is the variance of the physiological index values of the first q data points in the p-th physiological index data of the b-th slice data to be split, S b,p,q-1 is the variance of the physiological index value of the first q-1 data points in the p-th physiological index data of the b-th slice data to be split, q is the reference value, C b,p,q is the outlier degree of the qth data point in the pth physiological index data of the bth slice data to be split, and || is the absolute value; Obtaining the degree of outlier of each data point in each physiological indicator data of each slice data to be split, and obtaining a number of splitting nodes of each physiological indicator data of each slice data to be split by threshold judgment; The specific steps of obtaining a number of splitting nodes for each physiological indicator data of each slice data to be split by threshold judgment are as follows: Preset a second threshold, denoted as th2. If C b,p,q ≥ th2, take the q-th data point as a splitting node of the p-th physiological index data of the b-th slice data to be split, and start selecting the first q data points again from the (q + 1)-th data point in the p-th physiological index data. Obtain the outlier degree of the 2×q-th data point and compare it with th2. If the outlier degree of the 2×q-th data point is greater than or equal to th2, take the 2×q-th data point as a splitting node as well. Finally, obtain several splitting nodes of the p-th physiological index data of the b-th slice data to be split; if C b,p,q < th2, increment the value of the reference value q by one, obtain the outlier degree of the q-th data point after the change of the q value and continue the threshold judgment until the outlier degree of the data point is greater than or equal to th2 to obtain a splitting node. Then continue to calculate the outlier degree of the data points after the change of the q value and perform the threshold judgment backward. Finally, obtain several splitting nodes of the p-th physiological index data of the b-th slice data to be split; According to the order of the splitting nodes in the physiological index data, a plurality of sub-slice data of each slice data to be split are obtained, The specific steps involved are as follows: Any slice data to be split is recorded as the target slice data to be split, and the index of each split node in each physiological indicator data of the target slice data to be split is obtained, where the index is the order of the split node in the physiological indicator data, and the sequence composed of the indexes of all split nodes is recorded as the index sequence. The index sequence is clustered using the DBSCAN clustering algorithm to obtain several clusters, and the cluster containing the largest number of indexes is recorded as the target cluster, and the mean of all indexes in the target cluster is recorded as Will Round up as the splitting node of the target slice data to be split, split to obtain two sub-slice data of the target slice data to be split, obtain the discrete degree of each sub-slice data, if the discrete degree of the sub-slice data is greater than or equal to the first threshold, take the sub-slice data as a slice data to be split and split to obtain two sub-slice data again, and finally obtain a plurality of sub-slice data of the target slice data to be split; if the discrete degree of the sub-slice data is less than the first threshold, the sub-slice data is no longer split; The step of obtaining the first slice data and the second slice data of two adjacent slice data according to the lengths and sub-slice data of two adjacent slice data includes the following specific steps: For any two adjacent slice data, the slice data on the left is recorded as slice data A, and the slice data on the right is recorded as slice data B. If slice data A has multiple sub-slice data, the sub-slice data closest to slice data B is used as the first initial slice data. If slice data A has no sub-slice data, slice data A is used as the first initial slice data. If slice data B has multiple sub-slice data, the sub-slice data closest to slice data A is used as the second initial slice data. If slice data B has no sub-slice data, slice data B is used as the second initial slice data. Let the length of the first initial slice data be denoted as L1, and the length of the second initial slice data be denoted as L2. If L1 > L2, then obtain slice data adjacent to the second initial slice data and with a length of L2 in the first initial slice data as the first slice data, and directly use the second initial slice data as the second slice data; if L1 < L2, then obtain slice data adjacent to the first initial slice data and with a length of L1 in the second initial slice data as the second slice data, and directly use the first initial slice data as the first slice data; if L1 = L2, then use the first initial slice data as the first slice data and the second initial slice data as the second slice data; thus obtaining the first slice data and the second slice data of slice data A and slice data B.
2. A data interaction method for virtual simulation of anesthesiology crisis skills according to claim 1, characterized in that: The specific steps for obtaining the similarity between two adjacent slice data according to the physiological index values of each data point in each physiological index data of the first slice data and the second slice data are as follows: In the formula, m 1,i,k is the physiological index value of the kth data point in the ith item of physiological index data of the first slice data, is the mean value of the physiological index values of all data points in the ith physiological index data of the first slice data, m 2,i,k is the physiological index value of the kth data point in the i-th item of physiological index data of the second slice data, is the mean value of the physiological index values of all data points in the ith physiological index data of the second slice data, num i is the number of data points in the i-th physiological index data, N is the total number of physiological index data of the slice data, R 1,2 is the similarity between slice data A and slice data B.
3. A data interaction method for virtual simulation of anesthesiology crisis skills according to claim 2, characterized in that: The specific steps for performing a threshold judgment on the similarity between all pairs of adjacent slice data and merging to obtain several merged slice data of the physiological parameter data are as follows: A third threshold is preset, denoted as th3. If R 1,2 ≥th3, merge the slice data A and the slice data B in chronological order to obtain a merged slice data, compare the similarities of all two adjacent slice data with the third threshold, and merge to obtain multiple merged slice data; for several slice data that have not been merged, take each slice data that has not been merged as a merged slice data, and finally obtain several merged slice data of physiological parameter data.
4. A data interaction method for virtual simulation of anesthesiology crisis skills according to claim 1, characterized in that: The specific steps for clustering and classifying and storing all the merged slice data of the physiological parameter data are as follows: Denote any one of the merged slice data as the target merged slice data, obtain the standard deviation of the physiological index values of all data points in each physiological index data of the target merged slice data to obtain several standard deviations of the target merged slice data, arrange all the standard deviations in the order of acquisition of the physiological index data to obtain the standard deviation sequence of the target merged slice data; obtain the standard deviation sequence of each merged slice data, perform clustering on all the merged slice data according to the standard deviation sequence by the ISODATA algorithm, and use the DTW distance between the standard deviation sequences of the merged slice data as the distance metric to obtain several slice clusters, and store the merged slice data in each slice cluster separately.