CN108898238A - Medical equipment failure forecasting system and correlation technique, device and equipment - Google Patents
Medical equipment failure forecasting system and correlation technique, device and equipment Download PDFInfo
- Publication number
- CN108898238A CN108898238A CN201810510490.8A CN201810510490A CN108898238A CN 108898238 A CN108898238 A CN 108898238A CN 201810510490 A CN201810510490 A CN 201810510490A CN 108898238 A CN108898238 A CN 108898238A
- Authority
- CN
- China
- Prior art keywords
- training
- medical equipment
- status data
- equipment failure
- real
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H40/00—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices
- G16H40/40—ICT specially adapted for the management or administration of healthcare resources or facilities; ICT specially adapted for the management or operation of medical equipment or devices for the management of medical equipment or devices, e.g. scheduling maintenance or upgrades
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Health & Medical Sciences (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- Biomedical Technology (AREA)
- Human Resources & Organizations (AREA)
- General Physics & Mathematics (AREA)
- Strategic Management (AREA)
- General Health & Medical Sciences (AREA)
- Economics (AREA)
- Primary Health Care (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Artificial Intelligence (AREA)
- Epidemiology (AREA)
- Medical Informatics (AREA)
- Data Mining & Analysis (AREA)
- Public Health (AREA)
- Development Economics (AREA)
- General Engineering & Computer Science (AREA)
- Game Theory and Decision Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Medical Treatment And Welfare Office Work (AREA)
- Measuring And Recording Apparatus For Diagnosis (AREA)
Abstract
This application discloses medical equipment failure forecasting systems and correlation technique, device and equipment, this method to include：Acquire real-time status data, real time fail status data and the achievement data of Medical Devices；According to the real-time status data, real time fail status data and achievement data, current signature vector set S is obtainedn；By the current signature vector set SnNewest medical equipment failure prediction network is inputted, Fisrt fault prediction probability P is exportedn；It is searched for by Monte Carlo tree, exports the second failure predication probability Nn；With the Fisrt fault prediction probability PnWith the second failure predication probability NnMatching degree reach the first preset value and the Fisrt fault prediction probability PnWith next set of eigenvectors S of inputn+1It is training objective that the matching degree of middle real time fail status data, which reaches the second preset value,.The unsatisfactory technical problem of the predictablity rate of network is predicted with the medical equipment failure for solving that the mode training of supervised learning is caused to obtain due to exceptional sample limited amount.
Description
Technical field
This application involves medical equipment failures to predict field, in particular to medical equipment failure forecasting system and related side
Method, device and equipment.
Background technique
Medical Devices are the most basic elements in medical treatment, scientific research, mechanism and clinical speciality work, such as positron emission is broken
Layer imaging/x-ray computer tomography instrument (positron emission tomography/computedtomography,
PET/CT) equipment.Once but Medical Devices failure, the workflow that will lead to entire dependence Medical Devices are terminated, are caused tight
The economic loss of weight, it is often more important that may constitute a threat to the safety of patient, therefore, extremely to the failure predications of Medical Devices
It closes important.
Currently, the main training medical equipment failure in the way of supervised learning predicts network, above-mentioned training is reused
Obtained medical equipment failure prediction network carries out the failure predication of Medical Devices.But supervised learning is needed in training rank
Section inputs enough labeled exceptional samples, it is known that the quantity of the exceptional sample of Medical Devices is very limited, because
This, predicts that the predictablity rate of network is unsatisfactory using the medical equipment failure that the mode training of supervised learning obtains.
Summary of the invention
In view of this, the application proposes medical equipment failure forecasting system and correlation technique, device and equipment, it is existing to solve
Have due to exceptional sample limited amount in technology, the pre- survey grid of medical equipment failure for causing the mode training of supervised learning to obtain
The unsatisfactory technical problem of the predictablity rate of network.
In order to achieve the above object, technical solution used by the application is：
According to the embodiment of the present application in a first aspect, propose a kind of method of trained medical equipment failure forecasting system,
The method includes the steps：
Acquire real-time status data, real time fail status data and the achievement data of Medical Devices；
By carrying out feature extraction to the real-time status data, real time fail status data and achievement data, acquisition is worked as
Preceding set of eigenvectors Sn；
To the current signature vector set SnThe training medical equipment failure predicts network, wherein
By the current signature vector set SnInput newest medical equipment failure prediction network, output Fisrt fault prediction
Probability Pn；
It is searched for by Monte Carlo tree, exports the second failure predication probability Nn；
With the Fisrt fault prediction probability PnWith the second failure predication probability NnMatching degree reach the first preset value, with
And the Fisrt fault prediction probability PnWith next set of eigenvectors S of inputn+1The matching degree of middle real time fail status data reaches
It is training objective to the second preset value, the training medical equipment failure predicts network.
In some instances, the medical equipment failure prediction network is deep neural network；
The deep neural network includes：Residual error module based on convolutional neural networks.
In some instances, the medical equipment failure predicts network, also according to other Medical Devices training result into
Row training.
In some instances, the method also includes steps：Training result is uploaded to cloud, so that other equipment make
With.
In some instances, the real-time status data includes following at least any：Each electronic component operation of Medical Devices
Status data；
The running environment data of Medical Devices；
Correction log, maintenance log and the running log of Medical Devices.
According to the second aspect of the embodiment of the present application, a kind of medical equipment failure forecasting system, the system packet are proposed
It includes：
Data acquisition module, for acquiring real-time status data, real time fail status data and the index number of Medical Devices
According to；
Characteristic extracting module, for by the real-time status data, real time fail status data and achievement data into
Row feature extraction obtains current signature vector set Sn；
Machine learning module, for the current signature vector set SnThe training medical equipment failure predicts network,
Wherein,
By the current signature vector set SnInput newest medical equipment failure prediction network, output Fisrt fault prediction
Probability Pn；
It is searched for by Monte Carlo tree, exports the second failure predication probability Nn；
With the Fisrt fault prediction probability PnWith the second failure predication probability NnMatching degree reach the first preset value, with
And the Fisrt fault prediction probability PnWith next set of eigenvectors S of inputn+1The matching degree of middle real time fail status data reaches
It is training objective to the second preset value, the training medical equipment failure predicts network.
In some instances, the system also includes：Cooperative Study module,
The Cooperative Study module is used for the training result of itself to be uploaded to cloud for other Medical Devices；
Or/and
The training result of other Medical Devices is received, the training of the medical equipment failure prediction network is carried out.
According to the third aspect of the embodiment of the present application, a kind of medical equipment failure prediction technique is proposed, the method makes
Failure predication is carried out with the medical equipment failure forecasting system that such as above-mentioned first aspect training obtains, the method includes：
Acquire the real-time status data of Medical Devices；
Feature extraction is carried out to the real-time status data, obtains real-time vector set；
The medical equipment failure that the real-time vector set input training obtains is predicted into network, it is pre- to obtain medical equipment failure
The output result of survey grid network；
The failure predication result of Medical Devices is obtained according to output result obtained.
According to the fourth aspect of the embodiment of the present application, a kind of equipment of trained medical equipment failure forecasting system is proposed,
Including：
The memory of storage processor executable instruction；Wherein, the processor is coupled in the memory, for reading
The program instruction of the memory storage, and in response, it executes the training medical treatment as described in above-mentioned first aspect any one and sets
The operation of the method for standby failure prediction system.
According to the 5th of the embodiment of the present application the aspect, a kind of console device is proposed, including：Internal bus, Yi Jitong
Cross memory, processor and the external interface of internal bus connection；Wherein,
The external interface, for connecting Medical Devices；
The processor is coupled in the memory, for reading the program instruction of the memory storage, and as sound
It answers, the method for executing described instruction to realize the training medical equipment failure forecasting system as described in above-mentioned first aspect any one
Operation.
According to the 5th of the embodiment of the present application the aspect, a kind of storage medium is proposed, program is stored thereon with, the program quilt
Processor executes the operation of the method for training medical equipment failure forecasting system as described in above-mentioned first aspect any one.
The scheme that the application proposes is with current signature vector set SnFor input, the current signature vector set SnBy to doctor
Real-time status data, real time fail status data and the achievement data for treating equipment carry out feature extraction acquisition, and above-mentioned medical treatment is set
Standby real-time status data, real time fail status data and achievement data is the master data of Medical Devices, is very easy to obtain.
By the current signature vector set SnNewest medical equipment failure prediction network is inputted, Fisrt fault prediction probability P is exportedn, lead to
The search of Monte Carlo tree is crossed, the second failure predication probability N is exportedn；With the Fisrt fault prediction probability PnWith the second failure predication
Probability NnMatching degree reach the first preset value and the Fisrt fault prediction probability PnWith next set of eigenvectors of input
Sn+1It is training objective that the matching degree of middle real time fail status data, which reaches the second preset value, is trained and establishes learning strategy,
So that training obtains the higher medical equipment failure prediction network of accuracy rate in the case where not enough exceptional samples.
Detailed description of the invention
Fig. 1 is a kind of block diagram for trained medical equipment failure forecasting system that the embodiment of the present application illustrates；
Fig. 2 is a kind of machine learning module training medical equipment failure prediction shown in the embodiment of the present application exemplary
The flow chart of network；
Fig. 3 is a kind of medical equipment failure prediction technique that the embodiment of the present application illustrates；
Fig. 4 is the frame diagram for another medical equipment failure forecasting system that the embodiment of the present application illustrates；
Fig. 5 is the logic diagram of the equipment of the application training medical equipment failure forecasting system；
Fig. 6 is a kind of logic diagram of console device of the application.
Specific embodiment
The application is described in detail below with reference to specific embodiment shown in the drawings.But these embodiments are simultaneously
The application is not limited, structure that those skilled in the art are made according to these embodiments, method or functionally
Transformation is all contained in the protection scope of the application.
It is only to be not intended to be limiting the application merely for for the purpose of describing particular embodiments in term used in this application.
It is also intended in the application and the "an" of singular used in the attached claims, " described " and "the" including majority
Form, unless the context clearly indicates other meaning.It is also understood that term "and/or" used herein refers to and wraps
It may be combined containing one or more associated any or all of project listed.
Medical equipment failure is predicted to the safety of the service life of Medical Devices, the reasonable arrangement of medical resource and tested patient extremely
It closes important.With flourishing for artificial intelligence field, the mode of supervised learning be used to train medical equipment failure prediction
Network, the medical equipment failure prediction network obtained by using above-mentioned training carry out the failure predication of Medical Devices.We know
Supervised learning needs in road input enough labeled exceptional samples in the training stage, but the exceptional sample of Medical Devices
Quantity it is very limited, therefore, use supervised learning mode training obtain medical equipment failure prediction network prediction
Accuracy rate is unsatisfactory.
In order to solve the above-mentioned medical equipment failure prediction network for due to exceptional sample limited amount, training being caused to obtain
The unsatisfactory problem of predictablity rate, the application provide a kind of medical equipment failure forecasting system and correlation technique, device and
Equipment.
Referring to Fig.1, a kind of block diagram of the trained medical equipment failure forecasting system illustrated for the embodiment of the present application,
The system comprises：Data acquisition module 110, characteristic extracting module 120 and machine learning module 130.
Data acquisition module 110, for acquiring real-time status data, real time fail status data and the index of Medical Devices
Data；
The embodiment of the present application propose real-time status data may include each electronic component running state data of Medical Devices,
The running environment data of Medical Devices, the correction log of Medical Devices, maintenance log and running log etc..
Each electronic component running state data may include pressure, electric current and the temperature of each electronic component, and each
Revolving speed, position and times of exercise of mechanical part etc.；
The running environment data of the Medical Devices include：Climatic data and season data etc. are run, such as running environment temperature
The information such as degree, humidity.
The real time fail status data that the embodiment of the present application proposes is used to characterize the real time fail state status of Medical Devices,
For example, each component (including hardware and software etc., the hardware may include electronic component and mechanical part etc.) of Medical Devices
Fault condition.
The achievement data that the embodiment of the present application proposes may include the temporal resolution, spatial resolution, spirit of Medical Devices
The performance indicators such as sensitivity, counting rate and the data transmission objective of detector data acquisition channel etc..
Characteristic extracting module 120, for by the real-time status data, real time fail status data and achievement data
Feature extraction is carried out, current signature vector set S is obtainedn；
Specifically, can be the real-time status data for acquiring data acquisition module 110, real time fail status data and refer to
Data input features extraction module 120 is marked, the data of above-mentioned input are subjected to tagsort and feature extraction, obtain current signature
Vector set Sn, wherein n belongs to positive integer, the tagsort refer to by real-time status data, real time fail status data with
Data in achievement data are classified；The feature extraction refer to from real-time status data, real time fail status data with
Specified data is extracted in achievement data, such as：It is many to correct log content, feature extraction only extracts correction log middle finger
Fixed data.
Machine learning module 130, for the current signature vector set SnThe training pre- survey grid of medical equipment failure
Network.It is a kind of machine learning module training medical equipment failure prediction shown in the embodiment of the present application exemplary referring to Fig. 2
The flow chart of network, part training step are as follows：
S210：By the current signature vector set SnInput newest medical equipment failure prediction network, the first event of output
Hinder prediction probability Pn；
S220：It is searched for by Monte Carlo tree, exports the second failure predication probability Nn；
The Fisrt fault prediction probability PnAnd the second failure predication probability NnCurrent signature vector is inputted for characterizing
Collect SnSubsequent time, the probability of each component malfunction of Medical Devices.
S230：With the Fisrt fault prediction probability PnWith the second failure predication probability NnMatching degree to reach first default
Value and the Fisrt fault prediction probability PnWith next set of eigenvectors S of inputn+1Of middle real time fail status data
Reaching the second preset value with degree is training objective, and the training medical equipment failure predicts network.
The specific can be that in medical equipment failure prediction network inputs set of eigenvectors SnAfterwards, utilization is newest
Medical equipment failure predict network, carry out Monte Carlo tree search, output the second failure predication probability Nn, by set of eigenvectors
SnIt is input to the convolutional layer of medical equipment failure prediction network, output Fisrt fault predicts PnAnd characterization prediction it is accurate whether
Cost function Vn, to maximize the PnWith NnSimilarity and the cost function VnBe assigned a value of 1 for training objective, more
The parameter of the new medical equipment failure prediction network.By constantly iterating to calculate, until the VnIt is 1, and Pn and Nn
When matching degree reaches the first preset value, the training terminates.Wherein, the cost function VnAssignment it is pre- by the Fisrt fault
Survey probability PnWith next set of eigenvectors S of inputn+1The matching degree of middle real time fail status data determines, specifically can be：Doctor
Treat the next set of eigenvectors S of equipment fault prediction network inputsn+1Afterwards, due to next set of eigenvectors Sn+1In real time fail
Status data characterizes current signature vector set SnSubsequent time Medical Devices true fault situation, so by next feature
Vector set Sn+1In real time fail status data and PnIt is compared, the accuracy that can be predicted with characterization failure works as real time fail
Status data and PnMatching degree reach the second preset value, by the VnIt is assigned a value of 1, as real time fail status data and Pn?
It is not up to the second preset value with degree, by VnIt is assigned a value of 0.
After the training of medical equipment failure forecasting system, it can use above-mentioned trained system and carry out Medical Devices
Failure predication.Referring to Fig. 3, for a kind of medical equipment failure prediction technique that the embodiment of the present application illustrates, part is walked
It is rapid as follows：
S310：Acquire the real-time status data of Medical Devices.
S320：Feature extraction is carried out to the real-time status data, obtains real-time vector set.
S330：The medical equipment failure that the real-time vector set input training obtains is predicted into network, obtains Medical Devices
The output result of failure predication network.
In the application, " output result " can be each component of Medical Devices (including hardware and software etc., the hardware
May include electronic component and mechanical part etc.) probability that breaks down in subsequent time, such as：Output result can be：Figure
As failure (such as artifact occurs in image) probability：60%, cooling-water machine probability of malfunction：20% etc..
S340：The failure predication result of Medical Devices is obtained according to output result obtained.
In this step, failure predication result is in the output result, and probability of malfunction is more than the failure of the component of preset value
As a result, for example：Preset value is 80%, and it is 98% that output result, which is that image breaks down, and the probability that cooling-water machine breaks down is
60% ..., then failure predication result is that subsequent time image will break down.
The scheme that the application proposes is with current signature vector set SnFor input, the current signature vector set SnBy to doctor
Real-time status data, real time fail status data and the achievement data for treating equipment carry out feature extraction acquisition, and above-mentioned medical treatment is set
Standby real-time status data, real time fail status data and achievement data is the master data of Medical Devices, is very easy to obtain.
By the current signature vector set SnNewest medical equipment failure prediction network is inputted, Fisrt fault prediction probability P is exportedn, lead to
The search of Monte Carlo tree is crossed, the second failure predication probability N is exportedn；With the Fisrt fault prediction probability PnWith the second failure predication
Probability NnMatching degree reach the first preset value and the Fisrt fault prediction probability PnWith next set of eigenvectors of input
Sn+1It is training objective that the matching degree of middle real time fail status data, which reaches the second preset value, is trained and establishes learning strategy,
So that training obtains the higher medical equipment failure prediction network of accuracy rate in the case where being not necessarily to a large amount of exceptional samples.
In some instances, referring to Fig. 4, another medical equipment failure prediction illustrated for the embodiment of the present application
The frame diagram of system, the system also includes Cooperative Study module 140, the Cooperative Study module 140, for by the instruction of itself
Practice result and be uploaded to cloud, is used for other Medical Devices；Or/and the training result of other Medical Devices is received, described in progress
The training of medical equipment failure prediction network.
Specifically, the medical equipment failure forecasting system training when a Medical Devices finishes, it can be by the training of itself
As a result it is uploaded to cloud, it is pre- for medical equipment failure of other Medical Devices in training according to training result training itself
Examining system, so that the training result of other Medical Devices is more accurate.Certainly, when machine learning module 130 is trained, association
The training result is sent to machine learning module by the training result that can receive other Medical Devices with study module 140
130, so that machine learning module 130 predicts network according to the training result training medical equipment failure.
" training result " that the embodiment of the present application proposes can refer to the medical equipment failure prediction network that training obtains,
In some instances, described " training result " can also refer to the parameter for the medical equipment failure prediction network that training obtains
Value.
In some instances, the medical equipment failure prediction network is deep neural network；The deep neural network
Including：Residual error module based on convolutional neural networks uses batch regularization and non-thread in some instances in above-mentioned residual error module
Property integrates function.In some instances, value network is incorporated into tactful network in the medical equipment failure prediction network
In one framework.
Method and system described in the embodiment of the present application specification etc., can be in personal computer, desktop computer, work
It stands, executed on the electronic equipments such as industrial computer, Medical Devices and the table apparatus that is connected with Medical Devices.The electronics
Equipment can use embedded system, select SoC chip, handle by multicore ARM, FPGA and the DSP integrated in SoC chip
Device parallel computation, the calculating for carrying out the training of above-mentioned medical equipment failure forecasting system and using.
In addition, the description of each step, can be implemented as software, hardware or its form combined, for example, this field skill
Art personnel can implement these as the form of software code, can be the calculating that can be realized the corresponding logic function of the step
Machine executable instruction.When it is realized in the form of software, the executable instruction be can store in memory, and be set
Processor in standby executes.
Referring to Fig. 5, for equipment one embodiment schematic diagram of the application training medical equipment failure forecasting system, the equipment
500 may include：The memory of storage processor executable instruction；Wherein, the processor is coupled in the memory, is used for
The program instruction of the memory storage is read, and in response, is performed the following operations：
Acquire real-time status data, real time fail status data and the achievement data of Medical Devices；
By carrying out feature extraction to the real-time status data, real time fail status data and achievement data, acquisition is worked as
Preceding set of eigenvectors Sn；
To the current signature vector set SnThe training medical equipment failure predicts network, wherein
By the current signature vector set SnInput newest medical equipment failure prediction network, output Fisrt fault prediction
Probability Pn；
It is searched for by Monte Carlo tree, exports the second failure predication probability Nn；
With the Fisrt fault prediction probability PnWith the second failure predication probability NnMatching degree reach the first preset value, with
And the Fisrt fault prediction probability PnWith next set of eigenvectors S of inputn+1The matching degree of middle real time fail status data reaches
It is training objective to the second preset value, the training medical equipment failure predicts network.
Referring to Fig. 6, for equipment one embodiment schematic diagram of the application training medical equipment failure forecasting system, a kind of control
Platform equipment processed, including：Internal bus, and the memory, processor and the external interface that are connected by internal bus；Wherein,
The external interface, for connecting Medical Devices；
The processor is coupled in the memory, for reading the program instruction of the memory storage, and as sound
It answers, executes described instruction or less and operate：
Acquire real-time status data, real time fail status data and the achievement data of Medical Devices；
By carrying out feature extraction to the real-time status data, real time fail status data and achievement data, acquisition is worked as
Preceding set of eigenvectors Sn；
To the current signature vector set SnThe training medical equipment failure predicts network, wherein
By the current signature vector set SnInput newest medical equipment failure prediction network, output Fisrt fault prediction
Probability Pn；
It is searched for by Monte Carlo tree, exports the second failure predication probability Nn；
With the Fisrt fault prediction probability PnWith the second failure predication probability NnMatching degree reach the first preset value, with
And the Fisrt fault prediction probability PnWith next set of eigenvectors S of inputn+1The matching degree of middle real time fail status data reaches
It is training objective to the second preset value, the training medical equipment failure predicts network.
In the embodiment of the present application, computer readable storage medium can be diversified forms, for example, in different examples
In, the machine readable storage medium can be：RAM (Radom Access Memory, random access memory), it volatile deposits
Reservoir, nonvolatile memory, flash memory, memory driver (such as hard disk drive), solid state hard disk, any kind of storage dish
(such as CD, dvd) perhaps similar storage medium or their combination.Special, described computer-readable medium
Can also be paper or other be suitably capable of the medium of print routine.Using these media, these programs can be passed through
The mode of electricity gets (for example, optical scanner), can be compiled, be explained and processing in an appropriate manner, then can be by
It stores in computer media.
Those skilled in the art after considering the specification and implementing the invention disclosed here, will readily occur to its of the application
Its embodiment.This application is intended to cover any variations, uses, or adaptations of the application, these modifications, purposes or
Person's adaptive change follows the general principle of the application and including the undocumented common knowledge in the art of the application
Or conventional techniques.The description and examples are only to be considered as illustrative, and the true scope and spirit of the application are by the application
Claim point out.
It should be understood that the application is not limited to the precise structure that has been described above and shown in the drawings, and
And various modifications and changes may be made without departing from the scope thereof.Scope of the present application is only limited by the accompanying claims.
Claims (11)
1. a kind of method of trained medical equipment failure forecasting system, which is characterized in that the method includes the steps：
Acquire real-time status data, real time fail status data and the achievement data of Medical Devices；
By carrying out feature extraction to the real-time status data, real time fail status data and achievement data, obtain current special
Levy vector set Sn；
To the current signature vector set SnThe training medical equipment failure predicts network, wherein
By the current signature vector set SnNewest medical equipment failure prediction network is inputted, Fisrt fault prediction probability is exported
Pn；
It is searched for by Monte Carlo tree, exports the second failure predication probability Nn；
With the Fisrt fault prediction probability PnWith the second failure predication probability NnMatching degree reach the first preset value, Yi Jisuo
State Fisrt fault prediction probability PnWith next set of eigenvectors S of inputn+1The matching degree of middle real time fail status data reaches
Two preset values are training objective, and the training medical equipment failure predicts network.
2. the method according to claim 1, wherein medical equipment failure prediction network is depth nerve net
Network；
The deep neural network includes：Residual error module based on convolutional neural networks.
3. the method according to claim 1, wherein the medical equipment failure predicts network, also according to other
The training result of Medical Devices is trained.
4. the method according to claim 1, wherein the method also includes steps：Training result is uploaded to
Cloud, for other equipment use.
5. the method according to claim 1, wherein the real-time status data includes following at least any：Doctor
Treat each electronic component running state data of equipment；
The running environment data of Medical Devices；
Correction log, maintenance log and the running log of Medical Devices.
6. a kind of medical equipment failure forecasting system, which is characterized in that the system comprises：
Data acquisition module, for acquiring real-time status data, real time fail status data and the achievement data of Medical Devices；
Characteristic extracting module, for special by being carried out to the real-time status data, real time fail status data and achievement data
Sign is extracted, and current signature vector set S is obtainedn；
Machine learning module, for the current signature vector set SnThe training medical equipment failure predicts network, wherein
By the current signature vector set SnNewest medical equipment failure prediction network is inputted, Fisrt fault prediction probability is exported
Pn；
It is searched for by Monte Carlo tree, exports the second failure predication probability Nn；
With the Fisrt fault prediction probability PnWith the second failure predication probability NnMatching degree reach the first preset value, Yi Jisuo
State Fisrt fault prediction probability PnWith next set of eigenvectors S of inputn+1The matching degree of middle real time fail status data reaches
Two preset values are training objective, and the training medical equipment failure predicts network.
7. system according to claim 6, which is characterized in that the system also includes：Cooperative Study module,
The Cooperative Study module is used for the training result of itself to be uploaded to cloud for other Medical Devices；Or/and
The training result of other Medical Devices is received, the training of the medical equipment failure prediction network is carried out.
8. a kind of medical equipment failure prediction technique, which is characterized in that the method uses the doctor obtained such as claim 1 training
It treats equipment fault forecasting system and carries out failure predication, the method includes：
Acquire the real-time status data of Medical Devices；
Feature extraction is carried out to the real-time status data, obtains real-time vector set；
The medical equipment failure that the real-time vector set input training obtains is predicted into network, obtains the pre- survey grid of medical equipment failure
The output result of network；
The failure predication result of Medical Devices is obtained according to output result obtained.
9. a kind of equipment of trained medical equipment failure forecasting system, which is characterized in that including：
The memory of storage processor executable instruction；Wherein, the processor is coupled in the memory, described for reading
The program instruction of memory storage, and in response, execute the training Medical Devices event as described in claim 1 to 5 any one
Hinder the operation of the method for forecasting system.
10. a kind of console device, which is characterized in that including：Internal bus, and by internal bus connect memory,
Processor and external interface；Wherein,
The external interface, for connecting Medical Devices；
The processor is coupled in the memory, for reading the program instruction of the memory storage, and in response, holds
Row described instruction is to realize as described in claim 1 to 5 any one the behaviour of the method for training medical equipment failure forecasting system
Make.
11. a kind of storage medium, is stored thereon with program, which is characterized in that the program be executed by processor as claim 1 to
The operation of the method for training medical equipment failure forecasting system described in 5 any one.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810510490.8A CN108898238B (en) | 2018-05-24 | 2018-05-24 | Medical equipment fault prediction system and related method, device and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810510490.8A CN108898238B (en) | 2018-05-24 | 2018-05-24 | Medical equipment fault prediction system and related method, device and equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108898238A true CN108898238A (en) | 2018-11-27 |
| CN108898238B CN108898238B (en) | 2022-02-01 |
Family
ID=64343362
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810510490.8A Active CN108898238B (en) | 2018-05-24 | 2018-05-24 | Medical equipment fault prediction system and related method, device and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108898238B (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110175659A (en) * | 2018-12-04 | 2019-08-27 | 深圳码隆科技有限公司 | Spinning machine fault monitoring method |
| CN111371893A (en) * | 2020-03-06 | 2020-07-03 | 江苏赛诺格兰医疗科技有限公司 | Medical image equipment state remote monitoring system and method based on cloud platform |
| CN111489816A (en) * | 2020-04-19 | 2020-08-04 | 赵芳 | Intelligent medical fault processing method based on visit data transmission and cloud server |
| CN111598322A (en) * | 2020-05-09 | 2020-08-28 | 东软医疗系统股份有限公司 | Medical equipment fault prediction method, device, medium and electronic equipment |
| CN113284600A (en) * | 2021-05-08 | 2021-08-20 | 武汉联影医疗科技有限公司 | Fault prediction method, device, computer equipment and storage medium |
| CN113660137A (en) * | 2021-08-13 | 2021-11-16 | 杭州安恒信息技术股份有限公司 | Vehicle-mounted network fault detection method and device, readable storage medium and electronic equipment |
| CN116153483A (en) * | 2023-01-03 | 2023-05-23 | 武汉博科国泰信息技术有限公司 | Medical data analysis processing method and system based on machine learning |
| CN116705271A (en) * | 2023-08-09 | 2023-09-05 | 山东博达医疗用品股份有限公司 | Big data medical treatment flushing equipment operation monitoring system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130204412A1 (en) * | 2012-02-02 | 2013-08-08 | International Business Machines Corporation | Optimal policy determination using repeated stackelberg games with unknown player preferences |
| CN104915552A (en) * | 2015-05-27 | 2015-09-16 | 百度在线网络技术（北京）有限公司 | Method and device for predicting system faults |
| CN107341509A (en) * | 2017-06-29 | 2017-11-10 | 北京小米移动软件有限公司 | The training method and device of convolutional neural networks |
| CN107679451A (en) * | 2017-08-25 | 2018-02-09 | 百度在线网络技术（北京）有限公司 | Establish the method, apparatus, equipment and computer-readable storage medium of human face recognition model |
-
2018
- 2018-05-24 CN CN201810510490.8A patent/CN108898238B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130204412A1 (en) * | 2012-02-02 | 2013-08-08 | International Business Machines Corporation | Optimal policy determination using repeated stackelberg games with unknown player preferences |
| CN104915552A (en) * | 2015-05-27 | 2015-09-16 | 百度在线网络技术（北京）有限公司 | Method and device for predicting system faults |
| CN107341509A (en) * | 2017-06-29 | 2017-11-10 | 北京小米移动软件有限公司 | The training method and device of convolutional neural networks |
| CN107679451A (en) * | 2017-08-25 | 2018-02-09 | 百度在线网络技术（北京）有限公司 | Establish the method, apparatus, equipment and computer-readable storage medium of human face recognition model |
Non-Patent Citations (2)
| Title |
|---|
| 李晴: "基于优化机器学习算法的模拟电路故障诊断研究", 《中国博士学位论文全文数据库》 * |
| 郑幸: "基于蒙特卡洛法的配电网可靠性评估", 《中国优秀硕士学位论文全文数据库》 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110175659B (en) * | 2018-12-04 | 2022-04-19 | 深圳码隆科技有限公司 | Spinning machine fault monitoring method |
| CN110175659A (en) * | 2018-12-04 | 2019-08-27 | 深圳码隆科技有限公司 | Spinning machine fault monitoring method |
| CN111371893A (en) * | 2020-03-06 | 2020-07-03 | 江苏赛诺格兰医疗科技有限公司 | Medical image equipment state remote monitoring system and method based on cloud platform |
| CN111371893B (en) * | 2020-03-06 | 2023-03-10 | 江苏赛诺格兰医疗科技有限公司 | Medical imaging equipment state remote monitoring system and method based on cloud platform |
| CN111489816A (en) * | 2020-04-19 | 2020-08-04 | 赵芳 | Intelligent medical fault processing method based on visit data transmission and cloud server |
| CN111598322A (en) * | 2020-05-09 | 2020-08-28 | 东软医疗系统股份有限公司 | Medical equipment fault prediction method, device, medium and electronic equipment |
| CN113284600A (en) * | 2021-05-08 | 2021-08-20 | 武汉联影医疗科技有限公司 | Fault prediction method, device, computer equipment and storage medium |
| CN113660137A (en) * | 2021-08-13 | 2021-11-16 | 杭州安恒信息技术股份有限公司 | Vehicle-mounted network fault detection method and device, readable storage medium and electronic equipment |
| CN113660137B (en) * | 2021-08-13 | 2024-03-22 | 杭州安恒信息技术股份有限公司 | Vehicle-mounted network fault detection method and device, readable storage medium and electronic equipment |
| CN116153483A (en) * | 2023-01-03 | 2023-05-23 | 武汉博科国泰信息技术有限公司 | Medical data analysis processing method and system based on machine learning |
| CN116153483B (en) * | 2023-01-03 | 2023-11-07 | 武汉博科国泰信息技术有限公司 | Medical data analysis processing method and system based on machine learning |
| CN116705271A (en) * | 2023-08-09 | 2023-09-05 | 山东博达医疗用品股份有限公司 | Big data medical treatment flushing equipment operation monitoring system |
| CN116705271B (en) * | 2023-08-09 | 2023-11-14 | 山东博达医疗用品股份有限公司 | Medical flushing equipment operation monitoring system based on big data |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108898238B (en) | 2022-02-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108898238A (en) | Medical equipment failure forecasting system and correlation technique, device and equipment | |
| Schlegel et al. | Towards a rigorous evaluation of XAI methods on time series | |
| US10679751B2 (en) | Cell abnormality diagnosis system using DNN learning, and diagnosis management method of the same | |
| US11921566B2 (en) | Abnormality detection system, abnormality detection method, abnormality detection program, and method for generating learned model | |
| US9265441B2 (en) | Assessment of traumatic brain injury | |
| CN112766342A (en) | Abnormity detection method for electrical equipment | |
| CN111314173B (en) | Monitoring information abnormity positioning method and device, computer equipment and storage medium | |
| CN113722985B (en) | Method and system for evaluating health state and predicting residual life of aero-engine | |
| CN110147878B (en) | Data processing method, device and equipment | |
| CN107730040A (en) | Power information system log information comprehensive characteristics extracting method and device based on RBM | |
| CN110111885B (en) | Attribute prediction method, attribute prediction device, computer equipment and computer readable storage medium | |
| CN113705685B (en) | Disease feature recognition model training, disease feature recognition method, device and equipment | |
| CN108647707A (en) | Probabilistic neural network creation method, method for diagnosing faults and device, storage medium | |
| CN108231189B (en) | Data processing method and disease diagnosis device based on double-deep neural learning network | |
| CN114662386A (en) | Bearing fault diagnosis method and system | |
| CN113255735A (en) | Method and device for determining medication scheme of patient | |
| CN113760670A (en) | Cable joint abnormity early warning method and device, electronic equipment and storage medium | |
| CN111783242B (en) | RVM-KF-based rolling bearing residual life prediction method and device | |
| CN114329022A (en) | Method for training erotic classification model, method for detecting image and related device | |
| CN117220283A (en) | Photovoltaic power generation power prediction method, device and equipment | |
| JP7315181B2 (en) | Search method and information processing system | |
| CN110704614A (en) | Information processing method and device for predicting user group type in application | |
| CN113112463A (en) | Medical image quality evaluation method, electronic device, and storage medium | |
| CN115409050A (en) | Bearing fault diagnosis method, device and system | |
| EP4374333A1 (en) | Systems and methods for processing images for image matching |
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 | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: 110167 No. 177-1 Innovation Road, Hunnan District, Shenyang City, Liaoning Province Applicant after: Shenyang Neusoft Medical Systems Co.,Ltd. Address before: 110167 No. 177-1 Innovation Road, Hunnan District, Shenyang City, Liaoning Province Applicant before: SHENYANG NEUSOFT MEDICAL SYSTEMS Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20230426 Address after: Room 308, No. 177-2 Chuangxin Road, Hunnan District, Shenyang City, Liaoning Province, 110167 Patentee after: Shenyang Zhihe Medical Technology Co.,Ltd. Address before: 110167 No. 177-1 Innovation Road, Hunnan District, Shenyang City, Liaoning Province Patentee before: Shenyang Neusoft Medical Systems Co.,Ltd. |