CN116959696A - Data processing method and device based on laser therapeutic instrument - Google Patents
Data processing method and device based on laser therapeutic instrument Download PDFInfo
- Publication number
- CN116959696A CN116959696A CN202311210927.3A CN202311210927A CN116959696A CN 116959696 A CN116959696 A CN 116959696A CN 202311210927 A CN202311210927 A CN 202311210927A CN 116959696 A CN116959696 A CN 116959696A
- Authority
- CN
- China
- Prior art keywords
- data
- model
- training
- user
- target user
- 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
- 230000001225 therapeutic effect Effects 0.000 title claims abstract description 45
- 238000003672 processing method Methods 0.000 title claims abstract description 19
- 238000012549 training Methods 0.000 claims abstract description 76
- 238000011269 treatment regimen Methods 0.000 claims abstract description 29
- 238000000034 method Methods 0.000 claims abstract description 16
- 238000012545 processing Methods 0.000 claims abstract description 13
- 238000013136 deep learning model Methods 0.000 claims abstract description 11
- 238000004458 analytical method Methods 0.000 claims description 16
- 230000015654 memory Effects 0.000 claims description 13
- 238000012360 testing method Methods 0.000 claims description 7
- 238000004590 computer program Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 4
- 238000002647 laser therapy Methods 0.000 claims 3
- 230000000694 effects Effects 0.000 abstract description 6
- 238000011282 treatment Methods 0.000 description 21
- 238000001514 detection method Methods 0.000 description 11
- 239000008280 blood Substances 0.000 description 10
- 210000004369 blood Anatomy 0.000 description 10
- 230000002159 abnormal effect Effects 0.000 description 6
- 238000004891 communication Methods 0.000 description 5
- 230000036772 blood pressure Effects 0.000 description 4
- 239000000523 sample Substances 0.000 description 4
- 201000010099 disease Diseases 0.000 description 3
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 238000013532 laser treatment Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- WQZGKKKJIJFFOK-GASJEMHNSA-N Glucose Natural products OC[C@H]1OC(O)[C@H](O)[C@@H](O)[C@@H]1O WQZGKKKJIJFFOK-GASJEMHNSA-N 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000013500 data storage Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 239000008103 glucose Substances 0.000 description 2
- 238000011221 initial treatment Methods 0.000 description 2
- 150000002632 lipids Chemical class 0.000 description 2
- 210000001331 nose Anatomy 0.000 description 2
- 238000001126 phototherapy Methods 0.000 description 2
- 230000003068 static effect Effects 0.000 description 2
- 208000024891 symptom Diseases 0.000 description 2
- 210000000707 wrist Anatomy 0.000 description 2
- 206010033078 Otitis media Diseases 0.000 description 1
- 201000007100 Pharyngitis Diseases 0.000 description 1
- 230000003213 activating effect Effects 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 210000004204 blood vessel Anatomy 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013499 data model Methods 0.000 description 1
- 210000000613 ear canal Anatomy 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000035800 maturation Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000013508 migration Methods 0.000 description 1
- 230000005012 migration Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000003928 nasal cavity Anatomy 0.000 description 1
- 238000011369 optimal treatment Methods 0.000 description 1
- 239000000047 product Substances 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000013468 resource allocation Methods 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
- 230000009469 supplementation Effects 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
- 238000002560 therapeutic procedure Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- 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
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61N—ELECTROTHERAPY; MAGNETOTHERAPY; RADIATION THERAPY; ULTRASOUND THERAPY
- A61N5/00—Radiation therapy
- A61N5/06—Radiation therapy using light
- A61N5/0613—Apparatus adapted for a specific treatment
- A61N5/0625—Warming the body, e.g. hyperthermia treatment
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61N—ELECTROTHERAPY; MAGNETOTHERAPY; RADIATION THERAPY; ULTRASOUND THERAPY
- A61N5/00—Radiation therapy
- A61N5/06—Radiation therapy using light
- A61N5/067—Radiation therapy using light using laser light
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/2433—Single-class perspective, e.g. one-against-all classification; Novelty detection; Outlier detection
-
- 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
- G06N3/044—Recurrent networks, e.g. Hopfield networks
- G06N3/0442—Recurrent networks, e.g. Hopfield networks characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
-
- 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
- G06N3/08—Learning methods
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61N—ELECTROTHERAPY; MAGNETOTHERAPY; RADIATION THERAPY; ULTRASOUND THERAPY
- A61N5/00—Radiation therapy
- A61N5/06—Radiation therapy using light
- A61N2005/0626—Monitoring, verifying, controlling systems and methods
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Public Health (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Molecular Biology (AREA)
- Mathematical Physics (AREA)
- Veterinary Medicine (AREA)
- Animal Behavior & Ethology (AREA)
- General Business, Economics & Management (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Business, Economics & Management (AREA)
- Computing Systems (AREA)
- Radiology & Medical Imaging (AREA)
- Software Systems (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Pathology (AREA)
- Optics & Photonics (AREA)
- Primary Health Care (AREA)
- Epidemiology (AREA)
- Medical Informatics (AREA)
- Laser Surgery Devices (AREA)
Abstract
The application relates to the field of data processing, in particular to a data processing method and device based on a laser therapeutic instrument, wherein the method comprises the following steps: combining and classifying a treatment strategy of a preset laser therapeutic instrument with various parameters to obtain various treatment strategy classification results, and constructing a time sequence medical database containing the classification results, wherein the parameters comprise an operation mode, use intensity and use time; constructing a first model, wherein the first model is a pre-trained deep learning model, training the first model by using a database to obtain a second model, and the second model is a pre-trained deep learning model classified by a treatment strategy; creating a data queue of the target user, and distributing a model storage space to store historical data of the target user and a second model of the target user; and performing incremental training on the second model to generate and output a suggestion list containing treatment strategy classification results. The application has the effect of improving the accuracy and reliability of the strategy guidance of the appointed therapeutic apparatus.
Description
Technical Field
The present application relates to the field of data processing, and in particular, to a data processing method and apparatus based on a laser therapeutic apparatus.
Background
With the continuous development and maturation of emerging technologies such as big data, internet of things, artificial intelligence and the like, the fusion of the traditional medical industry with the emerging technologies is accelerated. Among them, new medical state represented by big data of health medical treatment is continuously activating the development of the medical industry. The medical big data has huge potential value, and the application of the medical big data is beneficial to improving medical service quality, reducing resource waste, optimizing resource allocation, improving self-health management and the like.
The traditional laser therapeutic apparatus only adopts a manual experience guiding mode of a doctor to guide the user to select a therapeutic mode, and the user only collects relevant physiological parameter data before primary treatment so as to determine a therapeutic strategy.
However, since the medical data has time sequence and single detection contingency, statistics and analysis of measurement data of patient changed in the treatment process in the prior art are not performed, and the treatment strategy is determined by only collecting relevant physiological parameter data before primary treatment, so that the accuracy and reliability are low.
Disclosure of Invention
In order to improve the accuracy and reliability of strategy guidance of a specified therapeutic instrument, the application provides a data processing method and device based on a laser therapeutic instrument.
In a first aspect, the present application provides a data processing method based on a laser therapeutic apparatus, which adopts the following technical scheme:
a data processing method based on a laser therapeutic instrument comprises the following steps: combining and classifying a treatment strategy of a preset laser therapeutic instrument with various parameters to obtain various treatment strategy classification results, and constructing a time sequence medical database containing the classification results, wherein the parameters comprise an operation mode, use intensity and use time; constructing a first model, wherein the first model is a pre-trained deep learning model, training the first model by using the database to obtain a second model, and the second model is a pre-trained deep learning model classified by a treatment strategy; creating a data queue of the target user, and distributing a model storage space to store historical data of the target user and the second model of the target user; performing incremental training on the second model, generating and outputting a suggestion list containing treatment strategy classification results, wherein the latest data in the data queue and the second model after the latest training are used for performing incremental training, and the sequence value required to be input by the model incremental training is a preset value of the data volume of the data queue.
By adopting the technical scheme, the user history data and the current detection data are combined, and the deep learning model capable of being updated in an increment is used for analysis, so that a reliable treatment strategy classification result can be obtained more quickly and accurately, the accuracy and the reliability of formulating the current treatment instrument strategy can be better ensured by analyzing the time history detection data, the correct guidance of the user on using the laser treatment instrument can be realized, and the treatment effect can be improved to the greatest extent.
Optionally, the method further comprises the following steps: and carrying out dynamic weight adjustment on the second model and target user data, wherein the target user data comprises historical data of a target user and acquisition data of the user.
By adopting the technical scheme, the problem that the correlation degree between the pre-training model data and the target user data in the data queue is not high is solved.
Optionally, the step of dynamically adjusting the weight of the second model and the target user data includes:
setting a pre-training data set and a user history data set, wherein the pre-training data set and the user history data set meet the following calculation formula:, wherein ,/>For the pre-training data set, +.>For the user history data set, +.>For the latest collected data in the target user data, Q is the value of the data amount of the data queue,/for the latest collected data in the target user data>=1；
Initializing and updating weights of the pre-training data set and the user history data set, wherein the expression of weight updating is as follows:
wherein ,weight value representing ith pre-training data, < +.>Represents the j-th user history data weight value,for the Euclidean distance between the ith pre-training data and the test data,/for the test data>Weight for pre-training data, +.>Weights for the user history dataset;
merging training data sets, wherein the relation of the merged training data sets is as follows:, wherein ,/>For the pre-training data set, < >>For user history data set->Weight for pre-training data, +.>Weights for the user history dataset. />For weighting the pre-training dataset +.>To weight the user history data set.
By adopting the technical scheme, the combined training data set is the final training set, as the user history data is continuously added in the queue, the gap between the pre-training data and the latest acquired data is continuously increased, the weight corresponding to the pre-training data and the latest acquired data is continuously reduced, and the proportion of the target user data in the training set is correspondingly increased.
Optionally, performing exception analysis on the target user data, and determining an exception data type, wherein the exception data type comprises a missing value and an outlier, if the missing value is determined to exist, sending a first signal, if the outlier is determined to exist, sending a second signal, and if the exception is determined to not exist, sending a normal signal; responding to a first signal confirmed instruction, sending a first alarm signal, and reminding a user of detecting that an interrupt exists; and responding to the confirmed instruction of the second signal, sending a second alarm signal to remind the user of detecting the abnormal wearing of the equipment, and improving the association degree of the model data and the target user data.
By adopting the technical scheme, through detection and processing of the abnormal data, the possibility of occurrence of the abnormal data and interference generated by operation of an analysis model are reduced to the greatest extent, and a user can know the body condition of the user through the acquired current physiological parameter data before treatment and provide data support for analysis decisions carried out later.
Optionally, the first model and the second model adopt a long-and-short-time memory network model.
In a second aspect, the present application provides a data processing device based on a laser therapeutic apparatus, which adopts the following technical scheme: a data processing apparatus based on a laser therapeutic apparatus, comprising: the laser therapeutic apparatus comprises a processor and a memory, wherein the memory stores computer program instructions which when executed by the processor realize the data processing method based on the laser therapeutic apparatus.
By adopting the technical scheme, the data processing method based on the laser therapeutic instrument generates a computer program, and the computer program is stored in the memory to be loaded and executed by the processor, so that the terminal equipment is manufactured according to the memory and the processor, and the use is convenient.
The application has the following technical effects:
1. the user history data and the current detection data are combined, and the incremental updatable LSTM model is used for analysis, so that a reliable treatment strategy classification result can be obtained more quickly and accurately, the accuracy and the reliability of formulating the current treatment instrument strategy can be better ensured through the analysis of the time history detection data, the correct guidance of the user on using the laser treatment instrument can be realized, and the treatment effect can be improved to the greatest extent.
2. Through detection and processing of the abnormal data, the possibility of occurrence of the abnormal data and interference generated by operation of an analysis model are reduced to the greatest extent, and a user can know own physical condition through the acquired current physiological parameter data before treatment and provide data support for analysis decisions carried out later.
3. The conventional LSTM model must prepare all data before learning, and the model cannot be updated as time series data grows. However, in the practical application scenario, new training data is continuously generated by the data stream over time, and if all the obtained data are to be stored and trained together, the data stream has extremely high requirements on memory resources, and the time required for training is continuously increased. Therefore, the application provides the LSTM model capable of being updated in an increment and self-adapting to dynamic weight adjustment in combination with the current application scene, and the model parameters can be updated after each new data uploading by combining with the model increment training and the migration learning method so as to improve the model classification accuracy without consuming a large amount of operation time.
Drawings
The above, as well as additional purposes, features, and advantages of exemplary embodiments of the present application will become readily apparent from the following detailed description when read in conjunction with the accompanying drawings. In the drawings, embodiments of the application are illustrated by way of example and not by way of limitation, and like reference numerals refer to similar or corresponding parts and in which:
fig. 1 is a flowchart of a method for processing steps S1 to S6 in a data processing method based on a laser therapeutic apparatus according to an embodiment of the present application.
Fig. 2 is a flowchart of a method of steps S60-S62 in a data processing method based on a laser therapeutic apparatus according to an embodiment of the present application.
Fig. 3 is a logic block diagram of a data processing system based on a laser therapeutic apparatus according to an embodiment of the present application.
Detailed Description
The following description of the embodiments of the present application will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are some, but not all embodiments of the application. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
It should be understood that when the terms "first," "second," and the like are used in the claims, the specification and the drawings of the present application, they are used merely for distinguishing between different objects and not for describing a particular sequential order. The terms "comprises" and "comprising" when used in the specification and claims of the present application are taken to specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
The embodiment of the application discloses a data processing method based on a laser therapeutic instrument, which is applied to an Internet of things communication architecture of an edge device and a cloud server, and is used for realizing user acquisition data transmission, namely cloud analysis, wherein the edge device is the laser therapeutic instrument, and the cloud server is a cloud platform for medical big data. In the process of constructing the framework, communication needs to be established between the user equipment and the cloud platform, so that the data storage function and calculation force support historical data storage and model operation provided by the cloud platform provide necessary support for a user treatment strategy.
The data acquisition module and the data analysis module are arranged in the laser therapeutic instrument, and physiological parameters which possibly influence the therapeutic strategy of the laser therapeutic instrument are detected before the user selects the operation mode to start the therapy. For example, the physiological parameters to be detected in the data processing method of the laser therapeutic apparatus of the present application may be blood pressure, blood fat, blood sugar, blood viscosity, skin thickness, blood vessel wall thickness, etc. The user can also manually input parameters, and the data acquisition module prompts the user to input the parameters by himself, such as gender, age, diseases to be treated, and the like, wherein the diseases can be diseases such as three highs, otitis media, pharyngitis and the like.
Referring to fig. 1, the data processing method based on the laser therapeutic apparatus includes steps S1 to S6, specifically as follows:
s1: and carrying out exception analysis on the target user data, judging the type of the exception data, wherein the type of the exception data comprises a missing value and an outlier, sending a first signal if the missing value is judged to exist, sending a second signal if the outlier is judged to exist, and sending a normal signal if the exception value is judged to not exist.
In particular, a missing value refers to a situation where the value of some parameter feature is missing or not recorded, and the cause of the missing value may be that the user accidentally interrupts the detection process.
Outliers refer to data that differ significantly from other data acquisition values, and in the present application are represented as data that deviate significantly from the normal physiological parameter range of the human body, which may be due to the fact that the user does not wear the data acquisition device correctly or does not clean the surface of the acquisition device, resulting in the insensitivity of the detection device. The outlier can be judged by adopting a common sense analysis method, a standard deviation method, a box diagram (quartile range-IQR) method and the like.
For example, an age value can be directly set to a range of [0,120] by common sense analysis, and an age value outside the range is considered as an outlier; the data with a large amount of statistical researches such as blood pressure and blood fat can be set into the upper and lower limit ranges of the data by a standard difference method, and the data detected to exceed the set upper and lower limit ranges is regarded as an abnormal value, and the calculation formula used by the standard difference method is as follows: upper limit: ll=μ+3σ, lower limit: UL = μ -3σ, where μ represents the data population mean and σ represents the data population standard deviation.
And responding to the confirmed instruction of the first signal, sending a first alarm signal to remind a user to detect that the interruption exists.
And responding to the confirmed instruction of the second signal, sending a second alarm signal to remind a user of detecting that the equipment is wearing abnormally.
S2: the treatment strategy of the preset laser therapeutic instrument is combined and classified with various parameters, various treatment strategy classification results are obtained, a time sequence medical database containing the classification results is constructed, and the parameters comprise an operation mode, use intensity and use time.
And carrying out combination classification on the treatment strategies of the laser therapeutic instrument according to the operation mode, the use intensity, the use time and the like of the therapeutic instrument, and obtaining different classification results of the treatment strategies. For example, the treatment apparatuses are classified into three kinds according to the treatment symptoms, i.e., a nasal cavity phototherapy probe, an ear canal laser probe, a single outputter of three parts of a pharyngeal laser probe, and a wrist phototherapy probe for simultaneous treatment with the three outputters. The operation mode can be that one device or a plurality of devices simultaneously perform treatment, and the wrist type and the ear/nose/throat type output devices simultaneously perform treatment (three types) or only the ear/nose/throat type output devices perform treatment (three types); the use intensity can be divided into four gears of 1/2/3/4 according to the power; the using time length can be divided into 15/30/45/60 minutes and four gears. Then the overall classification result of the treatment strategy is sharedA kind of module is assembled in the module and the module is assembled in the module.
And constructing a time sequence medical database according to the physiological parameters of the crowd treated by the laser therapeutic instrument in history and the adaptive parameters of the laser therapeutic instrument or the data obtained by a traditional meter on a user using the product in history. Examples of fields contained in the temporal medical database are as follows: [ timestamp, blood pressure, blood lipid, blood glucose, blood viscosity, ], the desired treatment symptoms, treatment policy classification ], wherein timestamp is the time of uploading data by the user, blood pressure, blood lipid, blood glucose, blood viscosity, ], etc. is the user-related physiological data, and the treatment policy classification is expressed as a number encoding 0-95.
S3: and constructing a first model, wherein the first model is a pre-trained deep learning model, training the first model by using a database to obtain a second model, and the second model is a pre-trained deep learning model classified by a treatment strategy.
In the application, a deep learning model uses an LSTM (Long Short-Term Memory network) model, and the LSTM model is trained by using data in a database in the step S2, so that a pre-trained laser therapeutic instrument treatment strategy classification LSTM model is obtained.
S4: creating a data queue of the target user, and distributing a model storage space to store historical data of the target user and a second model of the target user.
Specifically, after the user uploads data for the first time, an independent model storage space is opened for the user on the cloud platform, the data required by the current training of the user is stored in a queue, and the value of the data amount of the data queue is set to be Q, for example, in the application, q=30 is set, namely, 30 data which are the latest is taken each time, the data change amount is about one month and is used as the sequence data required to be input for the incremental training of the model, the adjacent historical data value can be referenced to the greatest extent, and the data which have less influence on the training of the model and are earlier can be discarded.
S5: and performing incremental training on the second model to generate and output a suggestion list containing treatment strategy classification results.
Specifically, the latest data in the data queue and the second model after the latest training are used for incremental training, and the sequence value required by the model incremental training is the value of the preset data amount of the data queue.
The model divided for each user individually will be updated with the incrementally trained parameters to accommodate the individual situation of each user to ensure more accurate classification results are derived. For example, when the value Q of the data amount of the data queue takes a value of 30, the model training classification can be classified into the following three cases:
when the initial user data amount in the data queue is 1, namely the user uploads the data for the first time, the uploaded data is directly input into the second model, and a classification result is given. The data will be kept in the queue as user history data after the classification result is obtained.
When the data quantity of the user in the data queue is less than 30, namely the data uploaded by the user is less than 30, the current data is subjected to data enhancement supplementation according to the classification result obtained before. For example, data that is randomly selected from a user database for a deep learning model of a pre-trained treatment strategy classification to be consistent with that user classification is filled at the beginning of the queue to supplement 30 pieces of data, and these enhancement data will be discarded bit-by-bit forward as the user data in the queue is updated.
When the user data amount in the data queue is equal to 30, that is, the total user uploading data amount is greater than 30, the latest data will enter the last bit of the queue, the first bit of the queue, that is, the earliest historical data, will be released, then the model is further updated by combining the latest data in the queue with the model updated by the last training, and the classification result, that is, the current optimal treatment strategy, is output.
After the suggestion list including the treatment strategy classification result is obtained in step S5, the medical big data platform sends the suggestion list to the equipment terminal of the user therapeutic apparatus, and the user can adjust the mode and parameters of the laser therapeutic apparatus which are required to be used for the current treatment according to the treatment strategy indicated by the classification result, so as to maximize the treatment effect of each use.
Through communication and data updating exchange between the laser therapeutic instrument and the medical big data platform, a user can check historical body data and change trend of the user in the set cloud platform of the medical big data at any time.
S6: and carrying out dynamic weight adjustment on the second model and target user data, wherein the target user data comprises historical data of a target user and acquisition data of the user. Referring to fig. 2, steps S60-S62 are included:
s60: a pre-training dataset and a user history dataset are set.
Wherein the pre-training data set isUser history data set is +.>，/>For the test data, the test data refers to the latest collected data in the target user data, ++>Wherein Q is a queue value,>=1；
s61: the weights of the pre-training dataset and the user history dataset are initialized and updated.
Specifically, the expression of the weight update is:
wherein ,weight value representing ith pre-training data, < +.>Represents the j-th user history data weight value,weight for pre-training data, +.>Weights for user history dataset +.>The Euclidean distance between the ith pre-training data and the test data can be expressed as: />D constitutes a vector dimension for the data.
The Euclidean distance between the data in the pre-training data set and the latest acquired data in the target user data represents the difference between the pre-training data and the latest acquired data in the target user data, and the larger the difference is, the smaller the corresponding weight is, namely the influence of the pre-training data on model training is reduced. In the present method an exponential function is usedProcessing the weight, and mapping the value range to [0,1 ]]And (3) inner part.
S62: the training data sets are merged.
The relationship for merging training datasets is:. wherein ,/>For the pre-training data set, < >>For user history data set->Weight for pre-training data, +.>Weights for the user history dataset.For weighting the pre-training dataset +.>To weight the user history data set.
The combined training data set is the mostThe training set used by the terminal is a weighted pre-training data set obtained by multiplying the weightAnd weighted user history dataset->Is a union of (a) and (b). As user history data is continuously added into the queue, the gap between the pre-training data and the latest acquired data is continuously increased, the corresponding weight is continuously reduced, and the proportion of the target user data in the training set is correspondingly increased.
After the dynamic weight adjustment in the step S6, the problem that the correlation degree between the pre-training model data and the target user data in the data queue is not high is solved, and the cloud platform of the medical large data can update and perfect the used second model through the data detected for a long time so as to continuously adjust the classification accuracy of the treatment strategy.
The implementation principle of the data processing method based on the laser therapeutic instrument in the embodiment of the application is as follows: the user history data and the current detection data are combined, and the incremental updatable LSTM model is used for analysis, so that a reliable treatment strategy classification result can be obtained more quickly and accurately, the accuracy and the reliability of formulating the current treatment instrument strategy can be better ensured through the analysis of the time history detection data, the correct guidance of the user on using the laser treatment instrument can be realized, and the treatment effect can be improved to the greatest extent.
The embodiment of the application also discloses a data processing system based on the laser therapeutic instrument, referring to fig. 3, comprising a processor and a memory, wherein the memory stores computer program instructions, and the computer program instructions realize the data processing method based on the laser therapeutic instrument according to the application when being executed by the processor.
The above system further comprises other components well known to those skilled in the art, such as a communication bus and a communication interface, the arrangement and function of which are known in the art and therefore are not described in detail herein.
In the context of this patent, the foregoing memory may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. For example, the computer readable storage medium may be any suitable magnetic or magneto-optical storage medium, such as, for example, resistive random access memory RRAM (ResistiveRandomAccessMemory), dynamic random access memory DRAM (DynamicRandomAccessMemory), static random access memory SRAM (static random access memory), enhanced dynamic random access memory EDRAM (EnhancedDynamicRandomAccessMemory), high-bandwidth memory HBM (High-bandwidth memory), hybrid storage cube HMC (HybridMemoryCube), etc., or any other medium that may be used to store the desired information and that may be accessed by an application, a module, or both. Any such computer storage media may be part of, or accessible by, or connectable to, the device. Any of the applications or modules described herein may be implemented using computer-readable/executable instructions that may be stored or otherwise maintained by such computer-readable media.
While various embodiments of the present application have been shown and described herein, it will be obvious to those skilled in the art that such embodiments are provided by way of example only. Many modifications, changes, and substitutions will now occur to those skilled in the art without departing from the spirit and scope of the application. It should be understood that various alternatives to the embodiments of the application described herein may be employed in practicing the application.
The above embodiments are not intended to limit the scope of the present application, so: all equivalent changes in structure, shape and principle of the application should be covered in the scope of protection of the application.
Claims (6)
1. A data processing method based on a laser therapeutic apparatus, comprising the steps of:
combining and classifying a treatment strategy of a preset laser therapeutic instrument with various parameters to obtain various treatment strategy classification results, and constructing a time sequence medical database containing the classification results, wherein the parameters comprise an operation mode, use intensity and use time;
constructing a first model, wherein the first model is a pre-trained deep learning model, training the first model by using the database to obtain a second model, and the second model is a pre-trained deep learning model classified by a treatment strategy;
creating a data queue of the target user, and distributing a model storage space to store historical data of the target user and the second model of the target user;
performing incremental training on the second model, generating and outputting a suggestion list containing treatment strategy classification results, wherein the latest data in the data queue and the second model after the latest training are used for performing incremental training, and the sequence value required to be input by the model incremental training is a preset value of the data volume of the data queue.
2. The laser therapy apparatus-based data processing method according to claim 1, further comprising the steps of:
and carrying out dynamic weight adjustment on the second model and target user data, wherein the target user data comprises historical data of a target user and acquisition data of the user.
3. The method for processing data based on a laser therapeutic apparatus according to claim 2, wherein the step of dynamically weighting the second model and the target user data comprises:
setting a pre-training data set and a user history data set, wherein the pre-training data set and the user history data set meet the following calculation formula:
wherein ,for the pre-training data set, +.>For the user history data set, +.>For the latest collected data in the target user data, Q is the value of the data amount of the data queue,/for the latest collected data in the target user data>=1；
Initializing and updating weights of the pre-training data set and the user history data set, wherein the expression of weight updating is as follows:
wherein ,weight value representing ith pre-training data, < +.>Represents the j-th user history data weight value, < ->For the Euclidean distance between the ith pre-training data and the test data,/for the test data>Weight for pre-training data, +.>Weights for the user history dataset;
merging training data sets, wherein the relation of the merged training data sets is as follows:, wherein ,for the pre-training data set, < >>For user history data set->Weight for pre-training data, +.>Weights for user history dataset +.>For weighting the pre-training dataset +.>To weight the user history data set.
4. A method of laser therapy apparatus based data processing according to claim 2 or 3, further comprising the steps of:
performing exception analysis on the target user data, and judging an exception data type, wherein the exception data type comprises a missing value and an outlier, if the missing value is judged to exist, a first signal is sent, if the outlier is judged to exist, a second signal is sent, and if the exception is judged to not exist, a normal signal is sent;
responding to a first signal confirmed instruction, sending a first alarm signal, and reminding a user of detecting that an interrupt exists;
and responding to the confirmed instruction of the second signal, sending a second alarm signal to remind a user of detecting that the equipment is wearing abnormally.
5. The method of claim 1, wherein the first model and the second model are based on a long-short-time memory network model.
6. A data processing apparatus based on a laser therapeutic apparatus, comprising: a processor and a memory storing computer program instructions which, when executed by the processor, implement the laser therapy apparatus based data processing method according to any one of claims 1-5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311210927.3A CN116959696B (en) | 2023-09-20 | 2023-09-20 | Data processing method and device based on laser therapeutic instrument |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311210927.3A CN116959696B (en) | 2023-09-20 | 2023-09-20 | Data processing method and device based on laser therapeutic instrument |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116959696A true CN116959696A (en) | 2023-10-27 |
| CN116959696B CN116959696B (en) | 2023-12-08 |
Family
ID=88454938
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311210927.3A Active CN116959696B (en) | 2023-09-20 | 2023-09-20 | Data processing method and device based on laser therapeutic instrument |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116959696B (en) |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109143199A (en) * | 2018-11-09 | 2019-01-04 | 大连东软信息学院 | Sea clutter small target detecting method based on transfer learning |
| CN110415119A (en) * | 2019-07-30 | 2019-11-05 | 中国工商银行股份有限公司 | Model training, bill business prediction technique, device, storage medium and equipment |
| WO2021169577A1 (en) * | 2020-02-27 | 2021-09-02 | 山东大学 | Wireless service traffic prediction method based on weighted federated learning |
| US20220051467A1 (en) * | 2020-08-17 | 2022-02-17 | Intel Corpoation | Apparatus and method for efficient graphics processing including ray tracing |
| CN114241779A (en) * | 2022-02-24 | 2022-03-25 | 深圳市城市交通规划设计研究中心股份有限公司 | Short-time prediction method, computer and storage medium for urban expressway traffic flow |
| CN114613492A (en) * | 2022-03-14 | 2022-06-10 | 重庆大学 | Wavelength control method of multimode laser xerophthalmia therapeutic instrument |
| CN114618090A (en) * | 2022-03-14 | 2022-06-14 | 重庆大学 | Energy control method of strong pulse laser xerophthalmia therapeutic instrument |
| CN114668980A (en) * | 2022-04-20 | 2022-06-28 | 北京思蒂尔科技发展有限责任公司 | Laser physiotherapy system and control method thereof |
| CN115130679A (en) * | 2022-02-11 | 2022-09-30 | 希望知舟技术(深圳)有限公司 | Data management and control method, related device and medium program product |
| US20230123322A1 (en) * | 2021-04-16 | 2023-04-20 | Strong Force Vcn Portfolio 2019, Llc | Predictive Model Data Stream Prioritization |
| CN116271563A (en) * | 2023-04-18 | 2023-06-23 | 广州市五合医疗器械有限公司 | Semiconductor laser therapeutic apparatus control method and system |
| CN116432780A (en) * | 2023-04-11 | 2023-07-14 | 济南博观智能科技有限公司 | Model increment learning method, device, equipment and storage medium |
-
2023
- 2023-09-20 CN CN202311210927.3A patent/CN116959696B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109143199A (en) * | 2018-11-09 | 2019-01-04 | 大连东软信息学院 | Sea clutter small target detecting method based on transfer learning |
| CN110415119A (en) * | 2019-07-30 | 2019-11-05 | 中国工商银行股份有限公司 | Model training, bill business prediction technique, device, storage medium and equipment |
| WO2021169577A1 (en) * | 2020-02-27 | 2021-09-02 | 山东大学 | Wireless service traffic prediction method based on weighted federated learning |
| US20220051467A1 (en) * | 2020-08-17 | 2022-02-17 | Intel Corpoation | Apparatus and method for efficient graphics processing including ray tracing |
| US20230123322A1 (en) * | 2021-04-16 | 2023-04-20 | Strong Force Vcn Portfolio 2019, Llc | Predictive Model Data Stream Prioritization |
| CN115130679A (en) * | 2022-02-11 | 2022-09-30 | 希望知舟技术(深圳)有限公司 | Data management and control method, related device and medium program product |
| CN114241779A (en) * | 2022-02-24 | 2022-03-25 | 深圳市城市交通规划设计研究中心股份有限公司 | Short-time prediction method, computer and storage medium for urban expressway traffic flow |
| CN114613492A (en) * | 2022-03-14 | 2022-06-10 | 重庆大学 | Wavelength control method of multimode laser xerophthalmia therapeutic instrument |
| CN114618090A (en) * | 2022-03-14 | 2022-06-14 | 重庆大学 | Energy control method of strong pulse laser xerophthalmia therapeutic instrument |
| CN114668980A (en) * | 2022-04-20 | 2022-06-28 | 北京思蒂尔科技发展有限责任公司 | Laser physiotherapy system and control method thereof |
| CN116432780A (en) * | 2023-04-11 | 2023-07-14 | 济南博观智能科技有限公司 | Model increment learning method, device, equipment and storage medium |
| CN116271563A (en) * | 2023-04-18 | 2023-06-23 | 广州市五合医疗器械有限公司 | Semiconductor laser therapeutic apparatus control method and system |
Non-Patent Citations (1)
| Title |
|---|
| 陈翔;王莉萍;顾庆;王赞;倪超;刘望舒;王秋萍;: "跨项目软件缺陷预测方法研究综述", 计算机学报, no. 01 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116959696B (en) | 2023-12-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10957451B2 (en) | Patient healthcare interaction device and methods for implementing the same | |
| US20200118685A1 (en) | Method and apparatus for providing user health status | |
| US11529105B2 (en) | Digital twin updating | |
| US11710572B2 (en) | Experience engine-method and apparatus of learning from similar patients | |
| EP3796226A1 (en) | Data conversion/symptom scoring | |
| US20200111576A1 (en) | Producing a multidimensional space data structure to perform survival analysis | |
| US20200234829A1 (en) | Systems and methods for facilitating response prediction for a condition | |
| WO2022151943A1 (en) | Health management method, device and system | |
| WO2020206172A1 (en) | Confidence evaluation to measure trust in behavioral health survey results | |
| CN116959696B (en) | Data processing method and device based on laser therapeutic instrument | |
| CN113436738A (en) | Method, device, equipment and storage medium for managing risk users | |
| Arya et al. | Heart disease prediction with machine learning and virtual reality: from future perspective | |
| US11457863B1 (en) | Virtual health coach based insulin titration platform | |
| Manikandan et al. | Improving the performance of classifiers by ensemble techniques for the premature finding of unusual birth outcomes from cardiotocography | |
| US20240006052A1 (en) | System implementing generative adversarial network adapted to prediction in behavioral and/or physiological contexts | |
| CN112957017B (en) | Real-time physical sign monitoring method and system for coma patient | |
| Hupalo et al. | Acquisition and Processing of Data in CPS for Remote Monitoring of the Human functional State | |
| US11610675B1 (en) | Dynamic and targeted allocation of resources for coaching service | |
| US20220270716A1 (en) | Confidence evaluation to measure trust in behavioral health survey results | |
| CN117594227B (en) | Health state monitoring method, device, medium and equipment based on wearable equipment | |
| EP4261842A1 (en) | Apparatus and method for generating an alert | |
| US20230238145A1 (en) | Health score and predictive analysis | |
| CN116759059B (en) | Female reproduction monitoring intelligent management system based on Internet | |
| Hajrulla et al. | Numerical Method Evaluation on Leveraging Machine-Learning Innovations for Improved Patient Care. Comparison of Numerical Results | |
| Ashcroft | Does machine learning improve the accuracy of clinical risk predictions? An exemplar examining risk of cardiovascular disease |
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 | ||
| GR01 | Patent grant |