CN118072929A - Real-time data intelligent management method for portable sterile surgical instrument package storage equipment - Google Patents
Real-time data intelligent management method for portable sterile surgical instrument package storage equipment Download PDFInfo
- Publication number
- CN118072929A CN118072929A CN202410480647.2A CN202410480647A CN118072929A CN 118072929 A CN118072929 A CN 118072929A CN 202410480647 A CN202410480647 A CN 202410480647A CN 118072929 A CN118072929 A CN 118072929A
- Authority
- CN
- China
- Prior art keywords
- environmental state
- state sequence
- historical
- sequence
- surgical instrument
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000007726 management method Methods 0.000 title claims abstract description 11
- 230000007613 environmental effect Effects 0.000 claims abstract description 428
- 230000002159 abnormal effect Effects 0.000 claims abstract description 42
- 238000000034 method Methods 0.000 claims abstract description 34
- 238000013528 artificial neural network Methods 0.000 claims abstract description 18
- 238000005516 engineering process Methods 0.000 claims abstract description 8
- 238000012549 training Methods 0.000 claims description 36
- 238000004364 calculation method Methods 0.000 claims description 15
- 230000001373 regressive effect Effects 0.000 claims description 9
- 238000010606 normalization Methods 0.000 claims description 4
- 230000000694 effects Effects 0.000 description 7
- 238000012544 monitoring process Methods 0.000 description 7
- 230000006870 function Effects 0.000 description 6
- 238000005070 sampling Methods 0.000 description 6
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 description 5
- 229910052760 oxygen Inorganic materials 0.000 description 5
- 239000001301 oxygen Substances 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 125000004122 cyclic group Chemical group 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 208000015181 infectious disease Diseases 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 238000009423 ventilation Methods 0.000 description 2
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 description 1
- 230000002411 adverse Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013523 data management Methods 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000004321 preservation Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000012876 topography Methods 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the technical field of medical information, in particular to a real-time data intelligent management method for portable sterile surgical instrument package storage equipment. The method comprises the steps of determining the matching degree based on the similarity between an acquired environmental state sequence of a surgical instrument set at the current moment and a historical environmental state sequence and the stability of the historical environmental state sequence, selecting part of the historical environmental state sequence from the historical environmental state sequence, and predicting the environmental state sequence to obtain a predicted environmental state sequence; splicing the environment state sequence and the predicted environment state sequence to obtain a complete predicted sequence; inputting the complete prediction sequence into a regressor weighted by the matching degree to obtain a prediction anomaly coefficient; and inputting the environmental state sequence and the prediction anomaly coefficient into a neural network to obtain the prediction validity period of the surgical instrument package. According to the invention, the regression device is used for determining the prediction abnormal coefficient, and the validity period of the surgical instrument package is predicted in real time through the neural network technology, so that the prediction validity period is obtained.
Description
Technical Field
The invention relates to the technical field of medical information, in particular to a real-time data intelligent management method for portable sterile surgical instrument package storage equipment.
Background
Under the scene that climate and topography are complicated changeable, including extremely cold, damp and hot, high altitude etc. adverse circumstances under, surgical operation apparatus package has very big storage problem, if the improper extremely easy emergence of storage mildews, can't ensure aseptic validity, leads to the risk surge of sick and wounded surgical operation treatment apparatus related infection. Therefore, how to manage surgical instruments and realize the considerable, measurable, adjustable and controllable storage environment of the surgical instrument package is a problem to be solved at present.
At present, the mobile numerical control storage device can be used for controlling environmental conditions such as temperature, humidity, ventilation and the like so as to keep the sterile state of the instrument package and ensure the quality requirement of surgical instruments. The movable numerical control storage device is a novel technology, can realize the functions of classifying and placing the surgical instrument bags, identifying clearly, quickly and accurately searching and the like, can reach the validity period of 180 days under specific conditions, and can save space, manpower and material resource cost.
However, in an actual environment, the storage environment of all the surgical instruments cannot be guaranteed to be constant, for example, when conditions such as external temperature, humidity and the like are changed drastically, the internal environment of the storage device may fluctuate, and the storage effect is reduced; or when the storage device is frequently switched on and off, the storage device is communicated with the external environment, so that the internal environment is unstable, and the quality and the preservation time of the instrument can be influenced.
The existing method for predicting the effective period of the surgical instrument is mostly based on a neural network trained by existing data, the accuracy and the reliability of a model depend on the number and the quality of samples of training data, but the available environmental data of the surgical instrument storage package in practice are limited, the effect of model training is limited, and the effective period prediction of the surgical instrument has real-time requirements, so that the dependence of the model on data samples is further increased, and finally the generalization capability of the model is often lacking due to insufficient data sample size; moreover, because of the complexity of the environmental data, it may be more difficult for the neural network to analyze the valid data, resulting in a large error in the final predicted result.
Disclosure of Invention
In order to solve the technical problem that effective data are difficult to analyze due to higher complexity of environmental data, and a predicted result finally obtained has larger error, the invention aims to provide a real-time data intelligent management method for portable sterile surgical instrument package storage equipment, and the adopted technical scheme is as follows:
acquiring an environmental state sequence of the surgical instrument package at the current moment;
Determining the matching degree of the environmental state sequence and the historical environmental state sequence at the current moment based on the similarity of the environmental state sequence and the historical environmental state sequence at the current moment and the stability of the historical environmental state sequence of the surgical instrument package;
Selecting part of historical environmental state sequences from the historical environmental state sequences according to the matching degree, and predicting the environmental state sequences to obtain predicted environmental state sequences;
Splicing an environmental state sequence of the surgical mechanical package at the current moment and the predicted environmental state sequence to obtain a complete predicted sequence; inputting the complete prediction sequence into a regressor weighted by the matching degree to obtain a prediction anomaly coefficient; and inputting the environment state sequence and the prediction abnormal coefficient into a trained neural network to obtain the prediction validity period of the surgical instrument package.
Preferably, the determining the matching degree of the environmental state sequence at the current time and the historical environmental state sequence based on the similarity of the environmental state sequence at the current time and the historical environmental state sequence and the stability of the historical environmental state sequence of the surgical instrument package includes:
Comparing the difference between the environmental state sequence of the surgical instrument package at the current moment and the historical environmental state sequence, and determining similar parameters of the environmental state sequence and the historical environmental state sequence of the surgical instrument package at the current moment;
determining the preferred parameters of the historical environmental state sequence by combining the stability and the frequent change condition of the historical environmental state sequence;
And adjusting the similar parameters according to the preferred parameters, and determining the matching degree of the environmental state sequence at the current moment and the historical environmental state sequence.
Preferably, the comparing the difference between the environmental state sequence of the surgical instrument package at the current time and the historical environmental state sequence, and determining the similarity parameters of the environmental state sequence of the surgical instrument package at the current time and the historical environmental state sequence includes:
The calculation formula of the similar parameters is as follows:
；
wherein, Is a first similarity; /(I)Is a second degree of similarity; /(I)Is a similar parameter; /(I)For the environmental state sequence/>, of the surgical instrument package at the current momentTotal number of elements in (a); /(I)For the environmental state sequence/>An ith environmental state parameter; The method comprises the steps of presetting standard environmental state parameters; /(I) For the historical environmental state sequence/>The i-th historical environmental state parameter; m is a historical environmental state sequence/>Total number of elements in (a); /(I)For the historical environmental state sequence/>A first historical environmental status parameter; /(I)For the environmental state sequence/>The difference between the i-th environmental state parameter and the i-1-th environmental state parameter; /(I)For the historical environmental state sequence/>The difference between the first historical environmental state parameter and the first-1 environmental state parameter; /(I)For the environmental state sequence/>The difference between the i-th environmental state parameter and the i+1-th environmental state parameter; /(I)For the historical environmental state sequence/>The difference between the first historical environmental state parameter and the first +1st environmental state parameter; i is the environmental state sequence/>A sequence number of the environmental status parameter; l is the historical environmental state sequence/>A serial number of a medium history environmental status parameter; u is a union symbol; max is the maximum value symbol.
Preferably, the determining the preferred parameters of the historical environmental state sequence according to the stability and the frequent change of the historical environmental state sequence includes:
The calculation formula of the preferred parameters is as follows:
；
wherein, Is a preferred parameter; /(I)For the current historical environmental state sequence/>First order difference of the first historical environmental state parameter; /(I)For the current historical environmental state sequence/>Second order difference of the first historical environmental state parameter; m is a historical environmental state sequence/>Total number of elements in (a); /(I)Is the number of historical environmental state sequences; /(I)To divide the historical environmental state sequence/>The number of other historical environmental state sequences; /(I)For the historical environmental state sequence/>And divide the historical environmental state sequence/>Other than the h-th historical environmental state sequence.
Preferably, the adjusting the similar parameters according to the preferred parameters, determining the matching degree between the environmental state sequence at the current time and the historical environmental state sequence includes:
and taking the normalized value of the product of the preferred parameter and the similar parameter as the matching degree of the environmental state sequence at the current moment and the historical environmental state sequence.
Preferably, the inputting the complete prediction sequence into a regressor weighted by the matching degree to obtain a prediction anomaly coefficient includes:
weights of weak regressors The acquisition method of (1) comprises the following steps: and carrying out weighted summation on the initial weight of the historical environmental state sequence in the weak regressor training set and the matching degree corresponding to the historical environmental state sequence, wherein the result value of the weighted summation is the weight of the weak regressor.
Preferably, the inputting the complete prediction sequence into a regressor weighted by the matching degree to obtain a prediction anomaly coefficient, further includes:
Determining a stored abnormal value of the historical environment state sequence according to the historical environment state sequence and the validity period duration;
Training a weak regressor by the historical environmental state sequence and the corresponding stored outlier; inputting the complete prediction sequence into each weak regressor to obtain initial abnormal values of each weak regressor;
and carrying out weighted summation on the initial abnormal value corresponding to each weak regressive and the weight of each weak regressive to obtain the prediction abnormal coefficient of the complete prediction sequence of the surgical instrument package.
Preferably, the determining the stored outlier of the historical environmental state sequence according to the historical environmental state sequence and the validity period duration includes:
the calculation formula of the stored abnormal value of the historical environment state sequence is as follows:
；
wherein, For the historical environmental state sequence/>Is stored with an outlier; /(I)Is a normalization function; /(I)For the historical environmental state sequence/>Total number of elements in (a); /(I)For the historical environmental state sequence/>A first historical environmental status parameter; The method comprises the steps of presetting standard environmental state parameters; /(I) For the historical environmental state sequence/>The standard validity period of the corresponding surgical instrument package; /(I)Historical environmental state sequence/>, for human taggingThe actual period of validity of the corresponding surgical kit.
Preferably, the selecting a part of the historical environmental state sequence from the historical environmental state sequences according to the matching degree, and predicting the environmental state sequence to obtain a predicted environmental state sequence includes:
Selecting historical environment state sequences with the matching degree of Top-k according to the sequence from large to small as a reference historical environment state sequence;
And carrying out weighted summation on the historical environmental state parameters in the reference historical environmental state sequence according to the normalized matching degree to obtain a predicted environmental state sequence.
Preferably, the acquiring the environmental state sequence of the surgical instrument package at the current moment includes:
And tracking the position of each surgical instrument package through an RFID technology, and determining the environmental state sequence of the surgical instrument package at the current moment under the current position.
The embodiment of the invention has at least the following beneficial effects:
According to the embodiment of the invention, the storage abnormal value can be constructed and calculated according to the storage environment state parameters and the actual validity period of the existing surgical instrument package, the complexity and the training effect of the model are improved, the defect of insufficient training data samples is overcome to a certain extent, and the characteristic recognition capability of the neural network is enhanced. Matching the environmental state sequence at the current moment with the historical environmental state sequence, screening part of the historical environmental state sequence from the environmental state sequence to predict the environmental state sequence, analyzing the matching sequence from various characteristics, improving the accuracy of the obtained predicted environmental state sequence, and improving the matching degree of the predicted environmental state sequence and the environmental state sequence at the current moment. According to the embodiment of the invention, the future change mode of the current environmental state sequence is predicted through the historical data sample, the predicted abnormal coefficient is finally obtained, the generalization capability of the model is enhanced, the risk of over-fitting of the model is reduced, meanwhile, the method has a good matching effect on different environmental state sequences, and the real-time requirement of the effective period prediction of the surgical instrument package is met.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions and advantages of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a method for intelligent management of real-time data of a portable sterile surgical kit storage device according to one embodiment of the present invention;
fig. 2 is a flowchart of a method for obtaining matching degree between an environmental state sequence and a historical environmental state sequence at a current moment according to an embodiment of the present invention.
Detailed Description
In order to further describe the technical means and effects adopted by the invention to achieve the preset aim, the following is a detailed description of the specific implementation, structure, characteristics and effects of the real-time data intelligent management method for the portable sterile surgical instrument set storage device according to the invention in combination with the accompanying drawings and the preferred embodiment. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
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.
The embodiment of the invention provides a specific implementation method of a real-time data intelligent management method of portable sterile surgical instrument package storage equipment, which is suitable for a surgical instrument package data management scene. The environmental state parameters at different positions can be obtained through various monitoring modules such as a temperature sensor, a humidity sensor, an oxygen sensor and the like under the scene. In order to solve the problem that the neural network may have a relatively high complexity of the environmental data, so that it is relatively difficult to analyze the effective data, and a relatively high error exists in the finally obtained prediction result. According to the embodiment of the invention, a data sample set is constructed through the existing surgical instrument package storage parameters, namely, the data sample set is constructed through the historical surgical instrument package storage parameters, the AdaBoost algorithm is utilized to train a plurality of weak regressors, then the data change mode prediction is carried out on the surgical instrument package in the current environment, the plurality of weak regressors are combined in a weighted mode according to the prediction result to obtain the prediction abnormal coefficient of the corresponding surgical instrument package, and finally the validity period of the surgical instrument package is predicted in real time through the neural network technology, so that the prediction validity period is obtained.
The following specifically describes a specific scheme of the portable sterile surgical instrument package storage equipment real-time data intelligent management method provided by the invention with reference to the accompanying drawings.
Referring to fig. 1, a flowchart of steps of a method for intelligently managing real-time data of a portable sterile surgical kit storage device according to an embodiment of the present invention is shown, where the method includes the following steps:
step S100, acquiring an environmental state sequence of the surgical instrument package at the current moment.
According to the invention, various monitoring modules such as a temperature sensor, a humidity sensor, an oxygen sensor and the like in the storage device are utilized to acquire the environmental state parameters of each surgical instrument pack at the current moment, and the environmental state parameters acquired within the preset sampling time range before the current moment form an environmental state sequence corresponding to the surgical instrument pack at the current moment. It should be noted that the preset sampling time range may be adjusted by the practitioner according to the actual situation. And the position, the state and the use condition of each surgical instrument bag are tracked and managed through an RFID technology, and the environmental state parameters of each surgical instrument bag at the current moment under the current position are recorded. When the environmental state parameters are acquired, the sampling frequency is one hour, and the environmental state parameters of each surgical instrument pack in a preset sampling time range are recorded as the same data sequence. And then the acquired data are transmitted to a monitoring system for analysis processing, wherein the sampling frequency is empirically set, and the sampling frequency can be adjusted according to requirements in specific application.
As one embodiment of the invention, the data collected by each sensor in the monitoring module can be used as one type of environmental state parameter, the following operations of step S200-step S400 are carried out on each type of environmental state parameter, each type of environmental state parameter corresponds to the prediction validity period of one surgical instrument package, and in order to solve the situation that the prediction validity period is distorted due to the deviation of the data collected by any sensor, the average operation is carried out on the prediction validity periods of the surgical instrument packages corresponding to different types of environmental state parameters, and the result value is used as the stable prediction validity period of the surgical instrument package.
As another embodiment of the invention, the data collected by each sensor in the monitoring module can be comprehensively evaluated to obtain an environment state sequence generated by combining a plurality of environment factors.
As other embodiments of the present invention, the data collected by each sensor in the monitoring module may also be formed into a multi-dimensional environmental state sequence, where the environmental state parameter collected by each sensor is used as a dimension. For example, the environmental state parameters collected by the temperature sensor, the humidity sensor and the oxygen sensor at the current time k are a1, a2 and a3 respectively, the environmental state parameters collected by the temperature sensor, the humidity sensor and the oxygen sensor at the previous time k-1 are b1, b2 and b3 respectively, and the environmental state parameters collected by the temperature sensor, the humidity sensor and the oxygen sensor at the time k-2 are b1, b2 and b3 respectively, so that the corresponding environmental state sequences are { (a 1, a2, a 3), (b 1, b2, b 3), (b 1, b2 and b 3) }. And carrying out subsequent operation on each dimension of the environment state sequence to obtain the prediction validity period of the surgical instrument package corresponding to each dimension, and taking the average value of the prediction validity periods of the surgical instrument packages corresponding to each dimension as the stable prediction validity period of the surgical instrument package.
Step S200, determining the matching degree of the environmental state sequence and the historical environmental state sequence at the current moment based on the similarity of the environmental state sequence and the historical environmental state sequence at the current moment and the stability of the historical environmental state sequence of the surgical instrument package.
When the surgical instrument package is predicted, the prediction model can not be constructed blindly by utilizing the existing data so as to predict the valid period of the existing surgical instrument package, and the prediction model can also be constructed by utilizing the historical data so as to predict the valid period of the existing surgical instrument package.
The self-adaptive enhancement algorithm (adaptiveboosting, adaBoost) is an iterative algorithm, the core idea is to train different weak classifiers for the same training set, and then combine the weak classifiers to form a stronger final strong classifier, so that the weak learner with the prediction precision slightly higher than that of the random conjecture can be enhanced into a strong learner with high prediction precision, and the method can be applied to all currently popular machine learning algorithms to further enhance the prediction precision of the original algorithm, and has good performance when processing a large-scale data set. The invention constructs a prediction model of the effective period of the surgical instrument package through an AdaBoost algorithm.
When the AdaBoost algorithm is utilized to construct a prediction model of the effective period of the surgical instrument package, the model training effect is poor because the environment parameter types of each surgical instrument package are more, the storage period is long, the parameter change mode is difficult to determine, and the training data sample obtained in practice can only cover a part of data modes. In addition, the validity period prediction of the surgical instrument has real-time requirements, the dependence of the model on data samples is further increased, the generalization capability of the model is easy to be lacking due to insufficient data sample size, and the prediction result error is larger.
Therefore, corresponding stored abnormal values can be quantized according to parameters of all surgical instrument packages in a training sample, future data mode changes of the existing surgical instrument packages are predicted by carrying out similarity matching on real-time environment state sequences of the existing surgical instrument packages, then a plurality of weak regressors obtained by an AdaBoost algorithm are combined in a weighted mode to obtain a strong regressor suitable for the current complete prediction sequence, and finally the stored abnormal values of the data sequences are predicted and the prediction validity period of the corresponding surgical instrument packages is obtained.
The process for processing the collected environmental state parameters of the surgical instrument package and predicting the validity period of the surgical instrument package in real time comprises the following steps: firstly, constructing a data sample set of the storage environment condition of the surgical instrument package; predicting the change mode of the existing data through the existing data sample to obtain the stored abnormal value of the data; and obtaining the final prediction validity period of the surgical instrument package by using the neural network.
The method comprises the steps of constructing a data sample set of the storage environment condition of the surgical instrument package, and specifically: before making the prediction, enough data is first needed as training samples. The environmental state parameters of the surgical kit have a plurality of dimensions, such as temperature, humidity, ventilation, etc., which collectively affect the actual expiration date of the surgical kit. Therefore, the multidimensional environmental state parameters of each surgical instrument set at each time are combined into a characteristic vector, and the characteristic vector is environmental data. For each surgical instrument package, the environmental data at all times can form a time sequence, and each instrument package corresponds to a standard validity period and an actual validity period. The standard expiration date is the expiration date printed on the surgical instrument package, and the actual expiration date is the actual expiration date of the surgical instrument package under the influence of the environment.
In order to quantify the pattern of change in the sequence of environmental states of each surgical kit, all of the historical environmental state sequences are set to have a stored outlier, the larger the stored outlier is, the more abnormal the corresponding historical environmental state sequence is. The storage outlier is obtained by the overall deviation and the validity period deviation of the historical environment state sequence, and the larger the deviation is, the larger the storage outlier is. The data sequence of any current surgical instrument set is recorded as an environment state sequence, and the stored abnormal value of the history environment state sequence can be obtained through analysis and calculation.
It should be noted that, the data collected by each sensor in the monitoring module is used as one type of environmental status parameter, the following operations from step S200 to step S400 are performed on each type of environmental status parameter, and each type of environmental status parameter corresponds to a prediction validity period of a surgical instrument package, so the following operations are only directed at any type of environmental status parameter.
The calculation formula of the stored abnormal value of the historical environment state sequence is as follows:
；
wherein, For the historical environmental state sequence/>Is stored with an outlier; /(I)Is a normalization function; /(I)For the historical environmental state sequence/>Total number of elements in (a); /(I)For the historical environmental state sequence/>A first historical environmental status parameter; The method comprises the steps of presetting standard environmental state parameters; /(I) For the historical environmental state sequence/>The standard validity period of the corresponding surgical instrument package; /(I)Historical environmental state sequence/>, for human taggingThe actual period of validity of the corresponding surgical kit.
Wherein, in a calculation formula for storing abnormal values of the historical environmental state sequence,The overall deviation of the reference sequence relative to the preset standard environmental state parameter is represented, and the larger the value is, the larger the abnormal value is stored; Indicating a deviation in the expiration date of the surgical kit, the greater the value, the greater the stored outlier.
After the stored abnormal values of the historical environmental state sequences of all the historical surgical instrument packages are obtained, training the historical data samples by using an AdaBoost algorithm to obtain a plurality of weak regressors related to environmental state parameters and the stored abnormal values, and then analyzing the historical data samples to obtain the matching degree and the predicted environmental state sequences.
As another embodiment of the invention, the historical data sample can be analyzed to obtain the matching degree and the predicted environmental state sequence, and then the abnormal coefficient of the historical environmental state sequence of the historical surgical instrument package in the historical data sample can be obtained, so as to obtain a plurality of weak regressors for environmental state parameters and stored abnormal values.
The specific steps of obtaining a plurality of weak regressors for environmental state parameters and stored outliers are:
(1) And selecting a decision tree regression model as a basic regressor, wherein the independent variable is environment data, and the dependent variable is a stored abnormal value.
(2) The existing data samples are used for training the basic regressor, and the loss of the regressor is calculated through the mean square error.
(3) The weight of each data sequence is updated according to the error magnitude so that the regressor predicts samples with larger errors before paying more attention to the next round of training.
(4) A plurality of basic regressors, i.e., weak regressors, are obtained through iterative training. In the embodiment of the invention, the iteration times are set to an empirical value.
Thus, a complete training data sample set and a plurality of weak regressors obtained by training are obtained, and then the data prediction operation of the next step can be carried out.
For any surgical instrument package in the portable storage device, in order to predict a relatively accurate validity period, the most effective way is to find a historical environment state sequence similar to the change mode of the surgical instrument package, and predict the surgical instrument package by means of the stored abnormal values and the corresponding validity periods of the historical environment state sequence.
However, the validity period of the surgical instrument package needs to be predicted in real time, the change mode of the data is difficult to capture, and meanwhile, the number of available training data samples is limited, so that the actual accuracy of a prediction model is greatly reduced. Therefore, when the validity period of a certain surgical instrument package is predicted, the future change mode of the data sequence in the existing data sample can be predicted, namely the possible change direction of the data sequence of the surgical instrument package is analyzed, and then the predicted data sequence is processed through a regressor obtained through training to obtain corresponding abnormal coefficients, so that the model accuracy is improved.
According to the logic, the environmental state sequence of the surgical instrument package at the current moment is recorded asFor the environmental state sequence/>Based on the similarity between the environmental state sequence and the historical environmental state sequence at the current time and the stability of the historical environmental state sequence, the matching degree between the environmental state sequence at the current time and the historical environmental state sequence is determined, please refer to fig. 2, and fig. 2 is a step flow chart of a method for obtaining the matching degree between the environmental state sequence at the current time and the historical environmental state sequence.
Further, the step of obtaining the matching degree between the environmental state sequence at the current moment and the historical environmental state sequence is as follows:
step S201, comparing the difference between the environmental state sequence of the surgical instrument package at the current moment and the historical environmental state sequence, and determining similar parameters of the environmental state sequence of the surgical instrument package at the current moment and the historical environmental state sequence.
The calculation formula of the similar parameters is as follows:
；
wherein, Is a first similarity; /(I)Is a second degree of similarity; /(I)Is a similar parameter; /(I)For the environmental state sequence/>, of the surgical instrument package at the current momentTotal number of elements in (a); /(I)For the environmental state sequence/>An ith environmental state parameter; The method comprises the steps of presetting standard environmental state parameters; /(I) For the historical environmental state sequence/>The i-th historical environmental state parameter; m is a historical environmental state sequence/>Total number of elements in (a); /(I)For the historical environmental state sequence/>A first historical environmental status parameter; /(I)For the environmental state sequence/>The difference between the i-th environmental state parameter and the i-1-th environmental state parameter; /(I)For the historical environmental state sequence/>The difference between the first historical environmental state parameter and the first-1 environmental state parameter; /(I)For the environmental state sequence/>The difference between the i-th environmental state parameter and the i+1-th environmental state parameter; /(I)For the historical environmental state sequence/>The difference between the first historical environmental state parameter and the first +1st environmental state parameter; i is the environmental state sequence/>A sequence number of the environmental status parameter; l is the historical environmental state sequence/>A serial number of a medium history environmental status parameter; u is a union symbol; max is the maximum value symbol.
Wherein,Representing the environmental state sequence/>With historical environmental state sequence/>In/>Differences in the individual data; /(I)Representing the environmental state sequence/>(1 /)Deviations of the individual environmental state parameters from preset standard environmental state parameters; /(I)Then represent the environmental state sequence/>With historical environmental state sequence/>The smaller the difference in data at the corresponding time instant, and the environmental state sequence/>When the deviation from the preset standard environmental state parameter is larger, the environmental state sequence/>With historical environmental state sequence/>The greater the similarity of (c).
Second similarity degreeFor the environmental state sequence/>With historical environmental state sequence/>Based on the similarity of data change patterns,/>Then represent the environmental state sequence/>Middle/>Personal data and historical environmental State sequence/>Middle/>The change modes of the data are similar, wherein the smaller the data difference is, the smaller the data difference is before and after time sequence is, and the closer the corresponding moments of the data are, the more similar the change modes of the data are.For the 1 st to Mth historical environmental state parameter/>, in the range of the 1 st to Mth historical environmental state parameterIs a maximum value of (a).Then the representation will be a sequence of environmental states/>In a historical environmental state sequence/>The similarity data of the maximum change mode in the sequence is screened out, and the average value is obtained to obtain the environmental state sequence/>With historical environmental state sequence/>Based on the overall similarity of the data change patterns.
Step S202, combining the stability and the frequent change condition of the historical environment state sequence, determining the preferred parameters of the historical environment state sequence.
And obtaining the similarity between the environmental state sequence of the surgical instrument package at the current moment and the historical environmental state sequence, wherein the sequence with high similarity is the sequence with similar change modes. However, predicting the future change mode of the environmental state sequence of the surgical instrument package at the current moment by the similarity is still not accurate enough, because the high similarity does not represent that the environmental state sequence at the current moment can be necessarily developed according to the change of the existing historical environmental state sequence.
Since the environment in the surgical kit storage device is in a steady state in most cases, it is considered more likely for the current surgical kit environmental data sequence to progress toward a more steady direction. Furthermore, among all the historical environmental state sequences, there must be a high frequency sequence that varies closely and a low frequency sequence that varies uniquely, and the pattern of variation of these high frequency sequences is also more likely to be the future pattern of variation of the current surgical instrument package.
Based on the stability analysis of each historical environmental state sequence in the data sample set for training, the preferred parameters of the historical environmental state sequence are determined, namely, the preferred parameters of the historical environmental state sequence are determined by combining the stability and frequent change condition of the historical environmental state sequence.
The calculation formula of the preferred parameters is as follows:
；
wherein, Is a preferred parameter; /(I)For the current historical environmental state sequence/>First order difference of the first historical environmental state parameter; /(I)For the current historical environmental state sequence/>Second order difference of the first historical environmental state parameter; m is a historical environmental state sequence/>Total number of elements in (a); /(I)Is the number of historical environmental state sequences; /(I)To divide the historical environmental state sequence/>The number of other historical environmental state sequences; /(I)For the historical environmental state sequence/>And divide the historical environmental state sequence/>Other than the h-th historical environmental state sequence.
Wherein, in the calculation formula of the preferred parametersReflects the historical environmental state sequence/>Average variation size of (a); /(I)Reflects the historical environmental state sequence/>Average change speed of (a); /(I)The degree of stability of the historical environmental state sequence is reflected in terms of both the magnitude of the change and the speed of the change, the greater the value, the greater the preferred parameter of the historical environmental state sequence; /(I)Representing a historical environmental state sequence/>The greater the average similarity with other historical environmental state sequences, the more frequently the change pattern of the historical environmental state sequence appears, namely the higher the frequency of the change pattern of the historical environmental state sequence, and the more likely the change pattern of the high-frequency sequence with more similar changes is called as the future change pattern of the surgical instrument bag, the higher the preference degree of the corresponding sequence, wherein/>As the weight, the larger the similarity is, the larger the weight is.
So far, the similarity of the sequence of environmental states of the surgical kit at the current moment to all the sequences of historical environmental states and the preferred parameters of these sequences of historical environmental states have been obtained.
And step S203, adjusting the similar parameters according to the preferred parameters, and determining the matching degree of the environmental state sequence at the current moment and the historical environmental state sequence.
When the similarity between the environmental state sequence and a certain historical environmental state sequence is high enough, and the preferred parameter of the historical environmental state sequence is also high, the matching degree between the environmental state sequence and the historical environmental state sequence is considered to be high. That is, the environmental state sequence is more likely to develop according to the change mode of the historical environmental state sequence, so that further, the matching degree of the environmental state sequence and the historical environmental state sequence is determined by combining the preferred parameter and the similar parameter.
Further, taking the normalized value of the product of the preferred parameter and the similar parameter as the matching degree of the environmental state sequence at the current moment and the historical environmental state sequence.
In other embodiments, the matching degree is calculated by the following formula:
；
wherein, For the historical environmental state sequence/>Matching degree with the environment state sequence at the current moment; /(I)For the environmental state sequence and the historical environmental state sequence/>Similar parameters of (2); /(I)For the historical environmental state sequence/>Is used as a parameter.
When the environmental state sequence is more similar to the historical environmental state sequence, the matching degree of the environmental state sequence and the historical environmental state sequence is higher, because the similar sequences can be said to be more matched in general; the preferred parameters of the historical environment state sequence reflect the stability degree and the occurrence frequency of the change mode of the historical environment state sequence, so that when the preferred parameters of the historical environment state sequence are larger, the reliability of the historical environment state sequence is higher, and the preferred parameters of the historical environment state sequence are adjusted to similar parameters to obtain the matching degree of the historical environment state sequence and the environment state sequence at the current moment.
And step S300, selecting part of the historical environmental state sequences from the historical environmental state sequences according to the matching degree, and predicting the environmental state sequences to obtain predicted environmental state sequences.
And selecting historical environment state sequences with the matching degree of Top-k according to the sequence from large to small as a reference historical environment state sequence. In the embodiment of the present invention, the value of k is 5, and in other embodiments, the value of k may be adjusted by an implementer according to the actual situation. And carrying out weighted summation on the historical environmental state parameters in the reference historical environmental state sequence according to the normalized matching degree to obtain a predicted environmental state sequence.
Further, the length of the reference historical environmental state sequence is reserved as the length of the environmental state sequence at the current moment. It should be noted that, elements in the reference historical environmental state sequence close to the current moment are reserved. The values of all the predicted environmental state parameters in the predicted environmental state sequence are as follows:
；
wherein, A c-th predicted environmental state parameter in the predicted environmental state sequence; /(I)A c-th predicted environmental state parameter in the j-th reference historical environmental state sequence; /(I)The matching degree between the j-th reference historical environmental state sequence and the environmental state sequence at the current moment is the matching degree; k is the number of predicted environmental state sequences, and in the embodiment of the invention, the value is 5.
The change rule of the reference historical environment state sequence with higher matching degree is more similar to the change rule of the current environment state sequence, the predicted environment state sequence predicted by the reference historical environment state sequence with higher matching degree is more accurate, and therefore the reference historical environment state sequence with higher matching degree is given greater weight.
Constructing a predicted environmental state sequence from predicted environmental state parameters。
Step S400, splicing an environmental state sequence of the surgical mechanical package at the current moment and the predicted environmental state sequence to obtain a complete predicted sequence; inputting the complete prediction sequence into a regressor weighted by the matching degree to obtain a prediction anomaly coefficient; and inputting the environment state sequence and the prediction abnormal coefficient into a trained neural network to obtain the prediction validity period of the surgical instrument package.
After obtaining the predicted environmental state sequence, splicing the environmental state sequence of the surgical mechanical package at the current momentAnd the predicted environmental state sequence/>A complete predicted sequence is obtained. I.e./>I.e. the complete predicted sequence of the sequence of environmental states.
Inputting the complete prediction sequence into a plurality of weak regressors trained before, and weighting and combining the output results of the weak regressors to obtain the prediction anomaly coefficient of the sequence;
Wherein the weights of the weak regressors are summed by the weights of each historical environmental state sequence when the weak regressors are trained The degree of matching with them is obtained. It should be noted that, for the Adaboost algorithm, a plurality of weak classifiers may be combined into one strong classifier. The basic idea is to increase the accuracy of the classifier by weighting and repeating training the data set; wherein the weights of the training samples are initialized to equal values, and if there are N samples, each training sample is initially assigned the same weight value: 1/N. Therefore, for the embodiment of the present invention, the initial weight of each historical environmental state sequence is 1/H, where H is the number of the historical environmental state sequences.
Then the firstAdjusted weights of the weak regressors/>The acquisition method of (1) comprises the following steps: and carrying out weighted summation on the initial weight of the training sample in the weak regressive training set and the matching degree corresponding to the initial weight, namely carrying out weighted summation on the initial weight of the historical environment state sequence in the weak regressive training set and the matching degree corresponding to the historical environment state sequence, wherein the result value of the weighted summation is the weight of the weak regressive.
The weight calculation formula of the weak regressor is as follows:
；
wherein, For/>Adjusted weights of the weak regressors/>；/>For/>Initial weights of the h historical environmental state sequences in the training set of the weak regressors; /(I)For/>The h historical environmental state sequence in the training set of the weak regressor and the environmental state sequence at the current moment/>Matching degree of (3); norm is a normalization function.
The calculation formula of the weight of the weak regressor represents the environmental state sequence at the current momentThe higher the matching degree is, the higher the weight of the historical environment state sequence is occupied in training, and the weak regressor pair/>The higher the fitness of the corresponding weight/>The larger;
Inputting the complete prediction sequence into a regressor weighted by the matching degree to obtain a prediction abnormal coefficient, and further: and determining the stored abnormal value of the historical environment state sequence according to the historical environment state sequence and the validity period duration. It should be noted that, in step S200, a calculation method of the stored outlier of the historical environmental status sequence has been already provided, which is not described herein. The weak regressor is trained from the historical environmental state sequence and the corresponding stored outliers. It should be noted that, specifically, after the stored outlier of the surgical instrument set is obtained, the step of training the historical data sample by using the AdaBoost algorithm to further obtain a plurality of weak regressors related to the environmental state parameters and the stored outlier is given in detail in step S200, which is not described herein. Inputting the complete prediction sequence into each weak regressor to obtain initial abnormal values of each weak regressor; and carrying out weighted summation on the initial abnormal value corresponding to each weak regressive and the weight of each weak regressive to obtain the prediction abnormal coefficient of the complete prediction sequence of the surgical instrument package. The sum of the weights of the weak regressors is 1.
After the predicted abnormal coefficient is obtained, the environment state sequence and the predicted abnormal coefficient are input into a trained neural network, and the predicted effective period of the surgical instrument package is obtained.
And obtaining a prediction abnormal coefficient of a complete prediction sequence corresponding to the environmental state sequence of the existing surgical instrument package by using a model trained by an AdaBoost algorithm, and obtaining the prediction validity period of the surgical instrument package through a cyclic neural network (Recurrentneural Network, RNN). The method comprises the following specific steps:
(1) The method comprises the steps of using a cyclic neural network, taking an environmental state sequence and a predicted abnormal coefficient of an existing surgical instrument package as an input sequence of the cyclic neural network, taking a predicted validity period as output, and dividing a data set into a training set and a testing set for training and evaluating a model.
(2) The loss function of the cyclic neural network adopts a mean square error loss function, and an Adam optimizer is used as an optimizer to supervise training and train the model by using a training set.
(3) The test set is used for evaluating the performance of the model, so that the model is well trained and can be deployed for actual prediction.
Finally, the predicted validity period of all the surgical instrument packages in the portable storage device can be obtained through the steps S100-S400, each surgical instrument package is tracked and marked by utilizing the RFID technology, and when the surgical instrument package exceeding the predicted validity period is taken out, an alarm system of the device immediately gives an alarm to remind medical staff to take measures in time for processing. And the validity period information can interact with a back-end information system in real time so as to realize accurate tracking management of infection control.
Further, since the prediction validity period of the surgical instrument package under the environmental state parameter of one type is obtained in the steps of the steps S200 to S400, in order to consider multiple environmental state parameters in a combined manner, the operations of the steps S200 to S400 may be performed on each environmental state parameter, the prediction validity period of the surgical instrument package corresponding to each environmental state parameter may be calculated, and then the prediction validity period may be averaged, and the result value may be used as the stable prediction validity period of the surgical instrument package. The mean value is to avoid the situation that the prediction validity period is distorted due to the deviation of the environmental state data acquired by any sensor, so that the mean value is calculated on the prediction validity period of the surgical instrument package corresponding to different environmental state parameters, and the result value is used as the stable prediction validity period of the surgical instrument package.
In summary, the present invention relates to the technical field of medical information. The method comprises the steps of determining the matching degree based on the similarity between an acquired environmental state sequence of a surgical instrument set at the current moment and a historical environmental state sequence and the stability of the historical environmental state sequence, selecting part of the historical environmental state sequence from the historical environmental state sequence, and predicting the environmental state sequence to obtain a predicted environmental state sequence; splicing the environment state sequence and the predicted environment state sequence to obtain a complete predicted sequence; inputting the complete prediction sequence into a regressor weighted by the matching degree to obtain a prediction anomaly coefficient; and inputting the environmental state sequence and the prediction anomaly coefficient into a neural network to obtain the prediction validity period of the surgical instrument package. According to the embodiment of the invention, the regression device is used for determining the prediction abnormal coefficient, and the validity period of the surgical instrument package is predicted in real time through the neural network technology, so that the prediction validity period is obtained.
It should be noted that: the sequence of the embodiments of the present invention is only for description, and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments.
Claims (10)
1. The intelligent management method for the real-time data of the portable sterile surgical instrument package storage equipment is characterized by comprising the following steps of:
acquiring an environmental state sequence of the surgical instrument package at the current moment;
Determining the matching degree of the environmental state sequence and the historical environmental state sequence at the current moment based on the similarity of the environmental state sequence and the historical environmental state sequence at the current moment and the stability of the historical environmental state sequence of the surgical instrument package;
Selecting part of historical environmental state sequences from the historical environmental state sequences according to the matching degree, and predicting the environmental state sequences to obtain predicted environmental state sequences;
Splicing an environmental state sequence of the surgical mechanical package at the current moment and the predicted environmental state sequence to obtain a complete predicted sequence; inputting the complete prediction sequence into a regressor weighted by the matching degree to obtain a prediction anomaly coefficient; and inputting the environment state sequence and the prediction abnormal coefficient into a trained neural network to obtain the prediction validity period of the surgical instrument package.
2. The method for intelligently managing real-time data of a portable sterile surgical instrument set storage device according to claim 1, wherein the determining the matching degree of the environmental state sequence at the current time and the historical environmental state sequence based on the similarity of the environmental state sequence and the historical environmental state sequence at the current time and the stability of the historical environmental state sequence of the surgical instrument set comprises:
Comparing the difference between the environmental state sequence of the surgical instrument package at the current moment and the historical environmental state sequence, and determining similar parameters of the environmental state sequence and the historical environmental state sequence of the surgical instrument package at the current moment;
determining the preferred parameters of the historical environmental state sequence by combining the stability and the frequent change condition of the historical environmental state sequence;
And adjusting the similar parameters according to the preferred parameters, and determining the matching degree of the environmental state sequence at the current moment and the historical environmental state sequence.
3. The method for intelligently managing real-time data of a portable sterile surgical instrument set storage device according to claim 2, wherein the step of comparing the difference between the environmental state sequence of the surgical instrument set at the current time and the historical environmental state sequence, and determining similar parameters of the environmental state sequence of the surgical instrument set at the current time and the historical environmental state sequence, comprises the steps of:
The calculation formula of the similar parameters is as follows:
；
wherein, Is a first similarity; /(I)Is a second degree of similarity; /(I)Is a similar parameter; /(I)For the environmental state sequence/>, of the surgical instrument package at the current momentTotal number of elements in (a); /(I)For the environmental state sequence/>An ith environmental state parameter; /(I)The method comprises the steps of presetting standard environmental state parameters; /(I)For the historical environmental state sequence/>The i-th historical environmental state parameter; m is a historical environmental state sequence/>Total number of elements in (a); /(I)For the historical environmental state sequence/>A first historical environmental status parameter; for the environmental state sequence/> The difference between the i-th environmental state parameter and the i-1-th environmental state parameter; /(I)For the historical environmental state sequence/>The difference between the first historical environmental state parameter and the first-1 environmental state parameter; /(I)For the environmental state sequence/>The difference between the i-th environmental state parameter and the i+1-th environmental state parameter; /(I)For a sequence of historical environmental statesThe difference between the first historical environmental state parameter and the first +1st environmental state parameter; i is the environmental state sequence/>A sequence number of the environmental status parameter; l is the historical environmental state sequence/>A serial number of a medium history environmental status parameter; u is a union symbol; max is the maximum value symbol.
4. The method for intelligently managing real-time data of a portable sterile surgical kit storage device according to claim 2, wherein the determining the preferred parameters of the historical environmental state sequence in combination with the stability and the frequent change of the historical environmental state sequence comprises:
The calculation formula of the preferred parameters is as follows:
；
wherein, Is a preferred parameter; /(I)For the current historical environmental state sequence/>First order difference of the first historical environmental state parameter; /(I)For the current historical environmental state sequence/>Second order difference of the first historical environmental state parameter; m is a historical environmental state sequence/>Total number of elements in (a); /(I)Is the number of historical environmental state sequences; /(I)To divide the historical environmental state sequence/>The number of other historical environmental state sequences; /(I)For the historical environmental state sequence/>And divide the historical environmental state sequence/>Other than the h-th historical environmental state sequence.
5. The method for intelligently managing real-time data of a portable sterile surgical kit storage device according to claim 2, wherein the step of adjusting the similar parameters according to the preferred parameters to determine a matching degree between an environmental state sequence at a current time and a historical environmental state sequence comprises the steps of:
and taking the normalized value of the product of the preferred parameter and the similar parameter as the matching degree of the environmental state sequence at the current moment and the historical environmental state sequence.
6. The method for intelligently managing real-time data of a portable sterile surgical kit storage device according to claim 1, wherein the inputting the complete prediction sequence into a regressor weighted by the matching degree to obtain a prediction anomaly coefficient comprises:
weights of weak regressors The acquisition method of (1) comprises the following steps: and carrying out weighted summation on the initial weight of the historical environmental state sequence in the weak regressor training set and the matching degree corresponding to the historical environmental state sequence, wherein the result value of the weighted summation is the weight of the weak regressor.
7. The method for intelligently managing real-time data of a portable sterile surgical kit storage device according to claim 6, wherein the inputting the complete prediction sequence into a regressor weighted by the matching degree to obtain a prediction anomaly coefficient, further comprises:
Determining a stored abnormal value of the historical environment state sequence according to the historical environment state sequence and the validity period duration;
Training a weak regressor by the historical environmental state sequence and the corresponding stored outlier; inputting the complete prediction sequence into each weak regressor to obtain initial abnormal values of each weak regressor;
and carrying out weighted summation on the initial abnormal value corresponding to each weak regressive and the weight of each weak regressive to obtain the prediction abnormal coefficient of the complete prediction sequence of the surgical instrument package.
8. The method for intelligently managing real-time data of a portable sterile surgical kit storage device according to claim 7, wherein the determining the stored outlier of the historical environmental status sequence according to the historical environmental status sequence and the validity period duration comprises:
the calculation formula of the stored abnormal value of the historical environment state sequence is as follows:
；
wherein, For the historical environmental state sequence/>Is stored with an outlier; /(I)Is a normalization function; /(I)For the historical environmental state sequence/>Total number of elements in (a); /(I)For the historical environmental state sequence/>A first historical environmental status parameter; /(I)The method comprises the steps of presetting standard environmental state parameters; /(I)For the historical environmental state sequence/>The standard validity period of the corresponding surgical instrument package; Historical environmental state sequence/>, for human tagging The actual period of validity of the corresponding surgical kit.
9. The method for intelligently managing real-time data of a portable sterile surgical kit storage device according to claim 1, wherein selecting a part of a historical environmental state sequence from the historical environmental state sequences according to the matching degree, and predicting the environmental state sequence to obtain a predicted environmental state sequence comprises:
Selecting historical environment state sequences with the matching degree of Top-k according to the sequence from large to small as a reference historical environment state sequence;
And carrying out weighted summation on the historical environmental state parameters in the reference historical environmental state sequence according to the normalized matching degree to obtain a predicted environmental state sequence.
10. The method for intelligently managing real-time data of a portable sterile surgical kit storage device according to claim 1, wherein the step of acquiring the environmental state sequence of the surgical kit at the current moment comprises the steps of:
And tracking the position of each surgical instrument package through an RFID technology, and determining the environmental state sequence of the surgical instrument package at the current moment under the current position.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410480647.2A CN118072929B (en) | 2024-04-22 | 2024-04-22 | Real-time data intelligent management method for portable sterile surgical instrument package storage equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410480647.2A CN118072929B (en) | 2024-04-22 | 2024-04-22 | Real-time data intelligent management method for portable sterile surgical instrument package storage equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118072929A true CN118072929A (en) | 2024-05-24 |
| CN118072929B CN118072929B (en) | 2024-06-21 |
Family
ID=91100638
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410480647.2A Active CN118072929B (en) | 2024-04-22 | 2024-04-22 | Real-time data intelligent management method for portable sterile surgical instrument package storage equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118072929B (en) |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP3520722A1 (en) * | 2018-02-05 | 2019-08-07 | Covidien LP | Methods of determining a relative state of charge of a battery for a surgical instrument and surgical instruments and computer-readable media implementing the same |
| WO2020119051A1 (en) * | 2018-12-10 | 2020-06-18 | 平安科技（深圳）有限公司 | Cloud platform resource usage prediction method and terminal device |
| CN114745337A (en) * | 2022-03-03 | 2022-07-12 | 武汉大学 | Real-time congestion control method based on deep reinforcement learning |
| CN116364295A (en) * | 2022-12-28 | 2023-06-30 | 北京谊安医疗系统股份有限公司 | Medical data processing method and system |
| CN116453673A (en) * | 2023-04-07 | 2023-07-18 | 南京邮电大学 | Motion estimation method for flexible cable driven minimally invasive surgical instrument end effector |
| CN116582597A (en) * | 2023-07-13 | 2023-08-11 | 湖北省林业科学研究院 | Intelligent monitoring method and system for olive seedling raising environment data |
| CN117011381A (en) * | 2023-08-08 | 2023-11-07 | 赫丽佰(合肥)智能科技有限公司 | Real-time surgical instrument pose estimation method and system based on deep learning and stereoscopic vision |
| US20230386074A1 (en) * | 2020-10-19 | 2023-11-30 | Stephen CANTON | Computer vision and machine learning to track surgical tools through a use cycle |
-
2024
- 2024-04-22 CN CN202410480647.2A patent/CN118072929B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP3520722A1 (en) * | 2018-02-05 | 2019-08-07 | Covidien LP | Methods of determining a relative state of charge of a battery for a surgical instrument and surgical instruments and computer-readable media implementing the same |
| WO2020119051A1 (en) * | 2018-12-10 | 2020-06-18 | 平安科技（深圳）有限公司 | Cloud platform resource usage prediction method and terminal device |
| US20230386074A1 (en) * | 2020-10-19 | 2023-11-30 | Stephen CANTON | Computer vision and machine learning to track surgical tools through a use cycle |
| CN114745337A (en) * | 2022-03-03 | 2022-07-12 | 武汉大学 | Real-time congestion control method based on deep reinforcement learning |
| CN116364295A (en) * | 2022-12-28 | 2023-06-30 | 北京谊安医疗系统股份有限公司 | Medical data processing method and system |
| CN116453673A (en) * | 2023-04-07 | 2023-07-18 | 南京邮电大学 | Motion estimation method for flexible cable driven minimally invasive surgical instrument end effector |
| CN116582597A (en) * | 2023-07-13 | 2023-08-11 | 湖北省林业科学研究院 | Intelligent monitoring method and system for olive seedling raising environment data |
| CN117011381A (en) * | 2023-08-08 | 2023-11-07 | 赫丽佰(合肥)智能科技有限公司 | Real-time surgical instrument pose estimation method and system based on deep learning and stereoscopic vision |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118072929B (en) | 2024-06-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Huang et al. | An enhanced deep learning-based fusion prognostic method for RUL prediction | |
| CN111931983B (en) | Precipitation prediction method and system | |
| CN109085469A (en) | A kind of method and system of the signal type of the signal of cable local discharge for identification | |
| CN110880369A (en) | Gas marker detection method based on radial basis function neural network and application | |
| CN110781595B (en) | Method, device, terminal and medium for predicting energy use efficiency (PUE) | |
| CN110633846A (en) | Gas load prediction method and device | |
| CN114925938B (en) | Electric energy meter running state prediction method and device based on self-adaptive SVM model | |
| Mo et al. | Few-shot RUL estimation based on model-agnostic meta-learning | |
| Çelik | Using machine learning algorithms to detect milk quality | |
| Kadakadiyavar et al. | Efficient mixture control chart pattern recognition using adaptive RBF neural network | |
| Lu et al. | Image classification and identification for rice leaf diseases based on improved WOACW_SimpleNet | |
| CN118072929B (en) | Real-time data intelligent management method for portable sterile surgical instrument package storage equipment | |
| Ardywibowo et al. | Varigrow: Variational architecture growing for task-agnostic continual learning based on bayesian novelty | |
| Gao et al. | Neural networks for event extraction from time series: a back propagation algorithm approach | |
| Lobo et al. | Exploiting the stimuli encoding scheme of evolving spiking neural networks for stream learning | |
| CN117371608A (en) | Pig house multi-point temperature and humidity prediction method and system based on deep learning | |
| CN115392594B (en) | Electrical load model training method based on neural network and feature screening | |
| Ancona et al. | On the importance of sorting in" Neural Gas" training of vector quantizers | |
| Shirali et al. | Pruning the Way to Reliable Policies: A Multi-Objective Deep Q-Learning Approach to Critical Care | |
| CN114328921A (en) | Small sample entity relation extraction method based on distribution calibration | |
| Efitorov et al. | Adaptive neuro-fuzzy inference system used to classify the measurements of chemical sensors | |
| Haque et al. | Improved adam-based feedforward deep neural network model for personalized asthma predictions | |
| CN113804833A (en) | Universal electronic nose drift calibration method based on convex set projection and extreme learning machine | |
| Li et al. | Multiscale Feature Extension Enhanced Deep Global-Local Attention Network for Remaining Useful Life Prediction | |
| Chang et al. | ANODE-GAN: Incomplete Time Series Imputation by Augmented Neural ODE-Based Generative Adversarial Networks |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant |