CN116530942A

CN116530942A - Monitoring device based on spectrum sensor

Info

Publication number: CN116530942A
Application number: CN202310809888.2A
Authority: CN
Inventors: 王洲; 朱岩
Original assignee: Tsinghua University
Current assignee: Tsinghua University
Priority date: 2023-07-04
Filing date: 2023-07-04
Publication date: 2023-08-04

Abstract

The application relates to a monitoring device based on a spectrum sensor, and particularly relates to the technical field of health monitoring. There is provided a monitoring device based on a spectrum sensor, the monitoring device comprising: a spectrum sensor system, a communication system; the spectrum sensor system is used for monitoring spectrum data of a target object and obtaining spectrum information of the target object; calculating based on the spectrum data to obtain other types of index information except the spectrum information; the communication system is configured to read the monitoring information of the target object, and transmit the monitoring information of the target object back to a designated terminal, where the monitoring information includes: the spectral information, the other types of index information. Based on the technical scheme, multi-dimensional monitoring of multiple indexes such as spectrum information, other index information and the like of the target object can be realized for a long time, and the accuracy of monitoring is improved.

Description

Monitoring device based on spectrum sensor

Technical Field

The application relates to the technical field of health monitoring, in particular to a monitoring device based on a spectrum sensor.

Background

With the aging or adverse lifestyle effects of the human body, the human tissue gradually degenerates, such as: the joint degeneration mainly refers to the phenomenon of hyperosteogeny and aging of joint parts, namely the phenomenon of bone aging.

In the prior art, the lack of an effective method for timely finding and preventing human tissues such as joint degeneration can only be realized through periodic physical examination, but the physical examination symptoms are mostly the subsequent phenomenon, so that the method has little significance for preventing diseases.

Disclosure of Invention

The application provides a monitoring device based on a spectrum sensor. The technical scheme is as follows.

In one aspect, there is provided a monitoring device based on a spectrum sensor, the monitoring device comprising: a spectrum sensor system, a communication system;

the spectrum sensor system is used for monitoring spectrum data of a target object and obtaining spectrum information of the target object; calculating based on the spectrum data to obtain other types of index information except the spectrum information;

the communication system is configured to read the monitoring information of the target object, and transmit the monitoring information of the target object back to a designated terminal, where the monitoring information includes: the spectral information, the other types of index information.

In one possible implementation, the spectrum sensor system is configured to analyze a first specified feature corresponding to the spectrum data; evaluating a first type of index information based on the first specified feature;

and/or the number of the groups of groups,

the spectrum sensor system is used for analyzing a second specified characteristic corresponding to the spectrum data; and inputting the second designated features into a prediction model to obtain second type index information.

In one possible implementation, the first specified feature is: spectral characteristics of hemoglobin and oxygenated hemoglobin, the first type of index information being: tissue oxygen and levels;

and/or the number of the groups of groups,

the first specified feature is: metabolite absorption characteristics, the first type of index information is: a metabolic state of the tissue;

and/or the number of the groups of groups,

the first specified feature is: optical characteristics including reflection and scattering, and the first type index information is: tissue structure and histological features;

and/or the number of the groups of groups,

the first specified feature is: biomarker signature, the first type of index information is: a biomarker.

In one possible implementation, the prediction model is: the flow prediction model, the second type index information is: liquid information;

and/or the number of the groups of groups,

the prediction model is as follows: the mechanical prediction model, the second type index information is: mechanical information.

In one possible implementation, the liquid information includes at least one of: the pH value, sugar content and solid solution content of the liquid; the mechanical information includes at least one of: acceleration, amount of shake, eccentricity.

In one possible implementation, the monitoring device further includes: a flow sensor system, a mechanical sensor system;

the flow sensor system is used for monitoring the target object to obtain liquid information in the target object;

the mechanical sensor system is used for monitoring the target object to obtain mechanical information of the target object.

In one possible implementation manner, the communication system is configured to perform data fusion on first liquid information and second liquid information, and take the fused result as final liquid information, where the first liquid information is liquid information read from the spectrum sensor, and the second liquid information is liquid information read from the flow sensor;

the communication system is used for carrying out data fusion on first mechanical information and second mechanical information, and taking the fused result as final mechanical information, wherein the first mechanical information is mechanical information read from the spectrum sensor, and the second mechanical information is mechanical information read from the mechanical sensor.

In one possible implementation, the second specified feature includes at least one of:

spectral reflectance; spectral absorptivity.

In one possible implementation, the number of the spectrum sensor systems is one, the spectrum sensor system includes: a spectrum sensor and a spectrum generator;

or alternatively, the first and second heat exchangers may be,

the number of the spectrum sensor systems is a plurality, and each spectrum sensor system comprises: a spectrum sensor and a spectrum generator;

or alternatively, the first and second heat exchangers may be,

the number of the spectrum sensor systems is a plurality, and each spectrum sensor system comprises: one spectral sensor, one spectral generator, one spectral reflector, and the spectral generators, spectral reflectors in different spectral sensor systems are common or independent.

In one possible implementation, in case the number of the spectrum sensor systems is 3, the spectrum sensors are arranged in a Y-array.

In one possible implementation, the spectrum sensor and the spectrum generator in the spectrum sensor system are two devices separately provided.

In one possible implementation, the spectral generator is operable to step up the spectral frequency range from low to high frequencies.

The technical scheme that this application provided can include following beneficial effect:

the spectrum sensor system is used for monitoring the monitoring information of the target object (such as human tissue) for a long time, the monitoring information comprises spectrum information and other types of index information of the target object obtained based on the spectrum information, and the data can be transmitted back to the appointed terminal through the communication system, so that multi-dimensional monitoring of multiple indexes such as the spectrum information, other index information and the like of the target object for a long time is realized, and the monitoring accuracy is improved.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings that are needed in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present application, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

FIG. 1 is a schematic diagram of a spectral sensor-based monitoring apparatus, according to an exemplary embodiment.

Fig. 2 is a schematic diagram of a spectral sensor based monitoring apparatus according to an exemplary embodiment.

Fig. 3 is a schematic diagram of a spectral sensor based monitoring apparatus according to an exemplary embodiment.

Fig. 4 is a schematic diagram of a spectral sensor based monitoring apparatus according to an exemplary embodiment.

Detailed Description

The following description of the embodiments of the present application will be made apparent and fully in view of the accompanying drawings, in which some, but not all embodiments of the invention are shown. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are within the scope of the present disclosure.

The terminology used in the embodiments of the application is for the purpose of describing particular embodiments only and is not intended to be limiting of embodiments of the application. As used in the embodiments of the present application and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used herein refers to and encompasses any or all possible combinations of one or more of the associated listed items.

It should be understood that although the terms first, second, third, etc. may be used in embodiments of the present application to describe various information, these information should not be limited to these terms. These terms are only used to distinguish one type of information from another. For example, the first information may also be referred to as second information, and similarly, the second information may also be referred to as first information, without departing from the scope of embodiments of the present application. The word "if" as used herein may be interpreted as "at … …" or "at … …" or "responsive to a determination", depending on the context.

In order to avoid the above drawbacks, in an embodiment of the present application, a monitoring device based on a spectrum sensor is provided. The technical scheme provided by the application is further described below with reference to the following examples.

As shown in fig. 1, the monitoring device includes: a spectrum sensor system, a communication system; the spectrum sensor system is used for monitoring spectrum data of the target object to obtain spectrum information of the target object; calculating based on the spectrum data to obtain other types of index information except the spectrum information; the communication system is used for reading the monitoring information of the target object and transmitting the monitoring information of the target object back to the designated terminal, and the monitoring information comprises: spectral information, other types of index information.

Wherein the spectrum sensor system is a system for monitoring by spectrum.

The communication system is a system for externally communicating through a wired network or a wireless network. The wireless network can be wifi, bluetooth, zigbee, 5G, 4G, 3G, 2G and other existing network systems, can also be other transmission modes for space communication through electromagnetic waves, and the wired network can be cable, optical fiber and other existing systems, or can be other transmission modes for communication through electric signals or wired media.

Wherein the target object is an object monitored by the monitoring device. The target object may be a human tissue such as a joint, bone, muscle, etc.

In the embodiment of the application, the monitoring device is a monitoring device with a communication function and based on a spectrum sensor, the monitoring information of a target object (such as human tissue) is monitored for a long time through a spectrum sensor system, and the data can be returned to a designated terminal through the communication system, so that the long-term retention and further analysis of the data are realized, and the long-term monitoring of the target object is realized. The communication system can periodically transmit the monitoring information of the target object back to the designated terminal, or can transmit the monitoring information of the target object back to the designated terminal under the reporting instruction of the designated terminal in a passive wake-up mode.

In the embodiment of the application, the spectrum information directly obtained by monitoring the spectrum sensor system can be further processed to obtain other types of index information of the target object obtained based on the spectrum information, so that multi-dimensional monitoring of multiple indexes of the spectrum information and other index information of the target object is realized for a long time.

In one possible implementation, the spectral information includes: waveform information, and optical flow information.

For example, the state of health of the target object may be obtained by identifying the spectrum information, and in the case that the spectrum information is identified as abnormal spectrum information, based on the abnormal spectrum information, whether an organic change occurs in the target object may be determined to identify whether an interventional therapy is required, where the abnormal spectrum information includes at least one of the following methods:

(1) Spectral morphological feature analysis: and judging whether the spectrum result is normal or not by analyzing the morphological characteristics of the spectrum curve corresponding to the spectrum information, such as peak position, peak width, peak shape and the like. Abnormal spectral information may be manifested as peak shifts, peak shape changes, peak width increases, etc.

(2) And (3) spectrum quantitative analysis: and quantitatively analyzing the spectrum data corresponding to the spectrum information, such as calculating indexes of absorptivity, reflectivity and the like, and comparing the indexes with a normal value to judge whether a spectrum result is normal. The abnormal spectrum information may be represented as an index value deviating from a normal range, or the like.

(3) Spectral comparison analysis: the measured spectrum information is compared with the known normal spectrum information to judge whether the spectrum result is normal or not. Abnormal spectral information may appear inconsistent or significantly different from normal spectral results, etc.

(4) Statistical analysis: and judging whether the spectrum result is normal or not by carrying out statistical analysis, such as mean, variance, correlation coefficient and the like, on spectrum data corresponding to the spectrum information. Abnormal spectral information may appear as statistical parameter anomalies, etc.

Thus, by performing various means such as morphological analysis, quantitative analysis, comparison analysis, and statistical analysis on the spectral data, abnormal spectral results can be more comprehensively identified and resolved. Meanwhile, the specific recognition mode and the refinement method can be adjusted and optimized according to the specific spectrum sensor, the monitoring object, the research purpose and the like.

In one possible implementation, a spectral sensor system is used to analyze a first specified feature corresponding to the spectral data; evaluating the first type of index information based on the first specified feature; and/or a spectral sensor system for analyzing a second specified feature corresponding to the spectral data; and inputting the second designated characteristic into a prediction model to obtain second type index information.

In this implementation manner, after the spectrum data is monitored, a first specified feature corresponding to the spectrum data may be analyzed, so as to directly obtain first type index information, or a second specified feature corresponding to the spectrum data may be analyzed, a trained prediction model may be input by using the second specified feature, and second type index information may be output, so that other types of index information of different types may be obtained by using multiple modes.

In one possible implementation, the first specified feature is: spectral characteristics of hemoglobin and oxygenated hemoglobin, the first type of index information is: tissue oxygen and levels; and/or, the first specified feature is: metabolite absorption characteristics, the first type of index information is: a metabolic state of the tissue; and/or, the first specified feature is: optical characteristics including reflection and scattering, and the first type of index information is: tissue structure and histological features; and/or, the first specified feature is: biomarker signature, first type index information is: a biomarker.

By way of example, tissue oxygenation level information can be obtained by analyzing spectral features of hemoglobin and oxygenated hemoglobin in spectral feedback, which is of great significance for assessing blood supply and oxygenation status of tissue; by analyzing the metabolite absorption characteristics in the spectroscopic feedback, information about the metabolic state of the tissue can be obtained, for example, the concentration changes of metabolites such as glucose, lactic acid, etc. can be monitored, thereby assessing the metabolic activity and health of the tissue; spectral feedback can provide information about tissue structure and histological features, and by analyzing optical properties such as reflection, scattering, etc., morphological features, fibrous structure, and possibly pathological changes of the tissue can be assessed; other relevant physiological and pathological indicators may be obtained by spectral feedback depending on the particular biomarker and its characteristic behavior in the spectrum, e.g., the spectral signature of a particular fluorescent marker may be used to detect a particular disease or pathological state.

In one possible implementation, the predictive model is: the flow prediction model, the second type index information is: liquid information; and/or, the predictive model is: the mechanical prediction model, the second type index information is: mechanical information.

Wherein the second specified feature input by the prediction model includes at least one of the following: spectral reflectance; spectral absorptivity.

Illustratively, the spectrum sensor may acquire a series of spectrum data by measuring an absorption spectrum or a scattering spectrum of a target object to be measured. These spectral data need to be processed to obtain the desired fluid information and mechanical information. The specific process will vary from spectral sensor to spectral sensor and application domain, but generally comprises the following steps:

(1) And (3) spectrum data acquisition: absorption spectrum or scattering spectrum data of the target object are acquired by a spectrum sensor.

(2) Data preprocessing: preprocessing the collected original spectrum data, including background correction, signal noise removal, wavelength calibration and the like.

(3) Spectral analysis: and analyzing the preprocessed spectrum data, and extracting the required characteristic information. In applications of liquid information and mechanical information, an index such as spectral reflectance or spectral absorptivity is generally used for analysis.

(4) And (3) establishing a model: and establishing a corresponding prediction model according to the extracted characteristic information. Common models include partial least squares regression (Partial Least Squares Regression, PLSR), support vector machine regression (Support Vector machine Regression, SVR), and the like.

(5) Prediction result: and predicting the new type of monitoring information by using the established model to obtain information such as liquid information, mechanical information and the like.

In one possible implementation, the liquid information includes at least one of: the pH value, sugar content and solid solution content of the liquid.

For example, the ph value, sugar content and solid solution content of the target object can be obtained based on the spectrum data, and the above parameters can be used for measuring the health state of the corresponding target object. In general, when the flow rate decreases, the pH value is abnormal, the sugar content is abnormal, and the solid solution is abnormal, which means that the corresponding target object is about to suffer from excessive strain and organic damage.

In one possible implementation manner, the monitoring device further includes: a flow sensor system; and the flow sensor system is used for monitoring the target object and obtaining the liquid information in the target object.

Illustratively, a flow sensor system is added to the monitoring device, and the flow sensor system is used for synchronously monitoring the liquid information in the target object, so that the monitoring accuracy of the liquid information is ensured.

In one possible implementation, the mechanical information includes at least one of: acceleration, amount of shake, eccentricity.

For example, acceleration, amount of shake, eccentricity may be obtained based on the spectral data, and the above parameters may measure the health status of the corresponding target subject. In general, as the force continues to decrease, the acceleration is abnormal, the amount of shake is abnormal, and the eccentricity is indicative of the impending overstrain and organic damage to the corresponding target subject.

In one possible implementation, the monitoring device further comprises a mechanical sensor system; and the mechanical sensor system is used for monitoring the target object and obtaining mechanical information of the target object.

Illustratively, a mechanics sensor system is added in the monitoring device, and the mechanics information of the target object is synchronously monitored by using the mechanics sensor system, so that the monitoring accuracy of the mechanics information is ensured.

In one possible implementation, the communication system is configured to perform data fusion on first liquid information and second liquid information, and take the fused result as final liquid information, where the first liquid information is liquid information read from the spectrum sensor, and the second liquid information is liquid information read from the flow sensor; and the communication system is used for carrying out data fusion on the first mechanical information and the second mechanical information, taking the fused result as final mechanical information, wherein the first mechanical information is the mechanical information read from the spectrum sensor, and the second mechanical information is the mechanical information read from the mechanical sensor.

For example, data fusion may be performed on information read from different sensors (e.g., results 1, 2) by:

(1) Data comparison and calibration: first, result 1 and result 2 are aligned and calibrated. This can be achieved by analyzing and correcting the difference between the two results. For example, the results may be calibrated based on known standards or reference values to eliminate possible measurement errors or deviations.

(2) Data fusion and synthesis: after calibration, results 1 and 2 may be data fused and integrated. This may be a simple weighted average, where results 1 and 2 are weighted according to their reliability and accuracy. More complex algorithms are also possible, such as methods based on machine learning or model prediction, combining the two results to obtain a more accurate blood flow result.

(3) Error detection and correction: error detection and correction mechanisms should also be considered when data fusion is performed. If there is a large difference or inconsistency between results 1 and 2, this may mean that one of the results is problematic. By setting reasonable thresholds or employing other algorithms, this discrepancy can be detected and corrected, or given lower weight in the data analysis.

Based on the technical scheme, the results of spectral feedback and sensor measurement are comprehensively considered, and more comprehensive and accurate liquid information and mechanical information can be provided, so that the health condition of human tissues can be accurately judged. The specific processing method and algorithm can be designed and optimized according to actual conditions and requirements.

In summary, according to the monitoring device based on the spectrum sensor provided in the embodiment, the spectrum sensor system monitors the monitoring information of the target object (such as human tissue) for a long time, the monitoring information includes the spectrum information, the liquid information in the target object obtained based on the spectrum information, and the mechanical information of the target object obtained based on the spectrum information, and the data can be transmitted back to the designated terminal through the communication system, so that the multi-dimensional monitoring of multiple indexes such as spectrum, liquid, mechanical and the like of the target object for a long time is realized, and the accuracy of the monitoring is improved.

In addition, a flow sensor system and a mechanical sensor system can be added in the monitoring device to synchronously monitor the liquid information and the mechanical information of the target object, so that the monitoring accuracy of the liquid information and the mechanical information is ensured.

In a specific application scenario, the analysis process of the spectrum data may be as follows:

and (3) data acquisition: first, spectroscopic data of the joint region is acquired using a spectrometer and other monitoring devices. The spectrometer is capable of collecting spectral reflectance or absorbance signals over different wavelength ranges. Such data may include optical characteristics of the joint tissue, such as spectral characteristics of various tissue components.

Data preprocessing: the data needs to be pre-processed before the spectroscopic data analysis can be performed. This includes steps of noise removal, background correction, wavelength calibration, etc., to ensure the quality and reliability of the data. The preprocessing step may involve techniques of signal filtering, peak alignment, baseline correction, etc., to improve the accuracy and consistency of the data.

Feature extraction: next, features are extracted from the preprocessed spectral data. Features may be peaks, valleys, wave shapes, etc. of the spectrum. By analyzing these features, quantitative and qualitative information about the spectral signal can be obtained. Feature extraction may be performed using mathematical algorithms, statistical methods, pattern recognition techniques, or the like.

Abnormality detection: by using the extracted features, abnormal detection, identification and discrimination of abnormal spectrum signals can be performed. The anomaly detection method may be based on statistical analysis, model building, or machine learning techniques. By comparing the difference between the current spectral data and the normal range or predefined pattern, it can be determined whether an abnormal situation exists.

Data interpretation and diagnosis: after an abnormal spectral signal is found, it needs to be interpreted and diagnosed with the relevant health condition. This may be based on previous studies or expertise to determine the correlation between abnormal signals and joint degeneration or other diseases. Meanwhile, other index data (such as blood flow and mechanical index) can be combined for comprehensive analysis so as to obtain more comprehensive health information.

Outcome output and reporting: and finally, according to the result of the spectrum data analysis, generating corresponding result output and report. These reports may provide a quantitative or qualitative assessment of joint health status.

In the training phase and the prediction phase of the prediction model, data processing is a critical step, which involves preprocessing of the input data and post-processing of the output data. The data processing of the training phase and the prediction phase will be described below, respectively.

Data processing in a training stage:

data cleaning: first, the raw data is cleaned, including noise removal, missing values processing, outliers processing, etc. This helps to improve the robustness and accuracy of the model.

Feature extraction and selection: features associated with the problem are extracted and selected from the cleaned data. This can be done by statistical methods, feature engineering techniques or domain knowledge. The selection of appropriate features helps to improve the model's effectiveness and generalization ability.

Data conversion and normalization: the data are transformed and normalized according to the requirements of the specific model. For example, normalizing features, encoding category features, vectorizing text data, and the like. This may improve the convergence speed and performance of the model.

Dividing data: the data set is divided into a training set, a validation set and a test set. The training set is used for training the model, the verification set is used for parameter adjustment and selection of the model, and the test set is used for evaluating the performance of the model. Reasonable data partitioning can ensure generalization capability and reliability of the model.

Model training: and training the model by using the processed training set. Various optimization algorithms and loss functions can be applied in the training process, and the model parameters are optimized through iteration, so that the model parameters gradually approach to the optimal solution.

Data processing in a prediction stage:

data preprocessing: the input data to be predicted is subjected to preprocessing steps similar to a training stage, including cleaning, feature extraction, data conversion and the like. And ensuring that the processing mode of the input data is consistent with that of the training data so as to maintain the consistency and accuracy of the model.

Model prediction: the pre-processed data is predicted using a trained model. The model is calculated according to the characteristics and parameters of the input data, and corresponding prediction results are generated.

Post-treatment: and carrying out post-processing on the output result of the model according to the requirements of specific problems. This may include decoding, de-normalization, post-processing filtering, etc. of the results to obtain final interpretable and usable prediction results.

In the data processing process, proper methods and techniques need to be selected according to the requirements of specific problems and models. At the same time, data processing is also an iterative process, requiring constant attempts and adjustments to obtain optimal data representation and model performance.

The following are some common examples of data processing formulas:

data cleaning:

removing noise or outliers: statistical methods such as mean, median, standard deviation, etc. may be used.

Processing the missing values: common methods include mean filling, median filling, interpolation, and the like.

Feature extraction and selection:

and (3) extracting statistical characteristics: such as average, maximum, minimum, variance, etc.

The characteristic engineering method comprises the following steps: such as polynomial features, cross features, principal Component Analysis (PCA), etc.

The characteristic selection method comprises the following steps: such as mutual information, correlation coefficients, L1 regularization, etc.

Data conversion and normalization:

normalization: for example scaling the data to the [0,1] range or the [ -1,1] range.

Standardization: for example, the data is converted to a standard normal distribution with a mean of 0 and a variance of 1.

Encoding: such as single-hot encoding, tag encoding, etc., of the category characteristics.

Model training:

loss function: depending on the task, different loss functions may be selected, such as Mean Square Error (MSE), cross entropy loss, etc.

Optimization algorithm: such as random gradient descent (SGD), adam optimizer, etc.

Model prediction:

and (3) calculating input characteristics: the representation of the input features is calculated from the input feature representation of the model, e.g., vectors, matrices, etc.

Model inference: and calculating a prediction result of the model on the input data according to the trained model parameters.

Note that the above formulas are merely common examples, and that the specific formulas may vary depending on the specific model and task. In practical application, the formula can be customized and adjusted according to specific requirements.

In the scheme, the fusion process of the liquid information aims at integrating and synthesizing the liquid information from different sources so as to obtain more comprehensive and accurate information. The following is a general procedure describing the liquid information fusion process:

and (3) data acquisition: a variety of liquid information, such as liquid composition, concentration, pH, etc., is obtained from the liquid sample using a liquid sensor or other monitoring device. Different sensors or devices may provide different types of information.

Data preprocessing: raw liquid information acquired from the sensor is preprocessed. This may include steps to remove noise, correct bias, supplement missing values, etc. to ensure quality and consistency of the data.

Feature extraction: useful features are extracted from the pre-processed liquid information. This may include statistical features (e.g., mean, variance), frequency domain features (e.g., spectral analysis), time domain features (e.g., waveform shape), time sequence features (e.g., trend analysis), etc.

Data fusion: the liquid information from the different sensors or devices is fused. The method of fusion may be a simple weighted average, feature level fusion, or a more complex model level fusion, such as fusing the predicted results of multiple models.

Decision and output: based on the fused liquid information, decision making is carried out or a final output result is generated. This may be classifying the liquid sample, predicting liquid properties, detecting abnormal conditions, etc.

In the liquid information fusion process, a key step is to select an appropriate fusion method and algorithm. Depending on factors such as the nature of the liquid information, the reliability of the data, the weight and reliability of the individual sensors or devices. Meanwhile, the aim of data fusion is to improve the accuracy, stability and reliability of the whole information and provide a more comprehensive liquid analysis result.

In this scheme, the fusion process of the liquid information involves a plurality of steps and methods, and the specific form of the formula will vary according to factors such as the data type, the fusion method and the model selection. The following are some common examples of formulas:

weighted average fusion:

post-fusion results = w1 data source 1+w2 data source 2+ & wn data source n

Where w1, w2,..wn is the weight of the corresponding data source, which may be determined based on data quality, reliability, or other factors.

Feature level fusion:

results after fusion= [ feature 1 source 1, feature 1 source 2, ], feature 1 source n, feature 2 source 1, feature 2 source 2, ], feature 2 source n ]

Wherein, feature 1, feature 2,..is a feature extracted from the liquid information, each column represents the value of one feature on a different data source.

Model level fusion:

results after fusion = f1 (data source 1) +f2 (data source 2) +.+ fn (data source n)

Where f1, f2,..fn is a predictive function for each model, which can be determined from the performance and weight of each model.

These formulas are just some common examples, and the form of formulas in a particular liquid information fusion process will vary from case to case. In practical application, according to specific tasks and data characteristics, a proper fusion method and a proper formula can be selected, and corresponding adjustment and optimization are performed.

In an exemplary embodiment, one or more spectral sensor systems are included in the monitoring device.

In one possible implementation, the number of the spectrum sensor systems is one, the spectrum sensor system includes: a spectrum sensor and a spectrum generator.

In this embodiment, the monitoring device includes one spectrum sensor system. Illustratively, referring to fig. 2 in combination, the monitoring device includes a spectrum sensor system, where the spectrum sensor system includes a spectrum sensor and a spectrum generator, and the spectrum generator emits light waves (either visible light or invisible light) with corresponding frequencies, and the spectrum sensor monitors spectrum information of the target object.

In one possible implementation, the number of spectral sensor systems is a plurality, each spectral sensor system comprising: a spectrum sensor and a spectrum generator.

In this embodiment, the monitoring device includes a plurality of spectrum sensor systems, each of which is composed of one spectrum sensor and one spectrum generator. Illustratively, referring to fig. 3 in combination, the monitoring device includes a plurality of spectrum sensor systems, each spectrum sensor system includes a spectrum sensor and a spectrum generator, and a spectrum receiving array is formed by the plurality of spectrum sensor systems, so that multi-dimensional monitoring of monitoring information is effectively realized.

In one possible implementation, the number of spectral sensor systems is a plurality, each spectral sensor system comprising: one spectral sensor, one spectral generator, one spectral reflector, and the spectral generators, spectral reflectors in different spectral sensor systems are common or independent.

In this embodiment, the monitoring device includes a plurality of spectrum sensor systems each including one spectrum sensor, one spectrum generator, and one spectrum reflector, and the entire monitoring device includes one or more spectrum generators, one or more spectrum sensors, and one or more spectrum reflectors. By way of example, referring to fig. 4 in combination, the monitoring device includes a spectrum generator, a spectrum reflector and two spectrum sensors, the spectrum generator emits light waves (either visible light or invisible light) with corresponding frequencies, the spectrum reflector can reflect the light waves, the spectrum sensor monitors spectrum information of the target object, and by arranging the spectrum transmitter, the spectrum generator can be reduced to save the cost of the device, and meanwhile, the spectrum sensor can obtain more signals.

In the implementation mode, when the monitoring device comprises a plurality of spectrum sensor systems, and therefore the monitoring device is provided with the plurality of spectrum sensors, a spectrum receiving array is formed by utilizing the feedback sensitivity of the spectrum sensors to the spectrum, and the Y-shaped array is preferentially protected, so that the multi-dimensional monitoring of the spectrum information is effectively realized, and the spectrum information can be effectively eliminated.

In this implementation, the spectrum generator and the spectrum sensor employ separate operating mechanisms. Because the spectrum generator may have a certain damage to the human body, the spectrum generator and the spectrum sensor can be set into two different devices, wherein the spectrum sensor is contacted with the target object for a long time, and the spectrum generator is externally arranged and then is placed at a corresponding point position when monitoring is needed.

In one possible implementation, the spectral generator is operative to step up the spectral frequency range from low to high frequencies.

In the implementation mode, the spectrum generator can adjust the spectrum frequency, gradually increase the spectrum frequency range from low frequency to high frequency, and collect spectrum information in a certain frequency range, so that simplified spectrum information full acquisition is realized.

In summary, the monitoring device based on the spectrum sensor provided in the embodiment has one or more spectrum generators, one or more spectrum sensors and one or more spectrum reflectors, so that the hardware composition mode of the monitoring device can be flexibly selected according to the needs of different scenes.

Other embodiments of the present application will be apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the application following, in general, the principles of the application and including such departures from the present disclosure as come within known or customary practice within the art to which the application pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the application being indicated by the following claims.

It is to be understood that the present application is not limited to the precise arrangements and instrumentalities shown in the drawings, which have been described above, and that various modifications and changes may be effected without departing from the scope thereof. The scope of the application is limited only by the appended claims.

Claims

1. A monitoring device based on a spectrum sensor, the monitoring device comprising: a spectrum sensor system, a communication system;

the communication system is configured to read the monitoring information of the target object, and transmit the monitoring information of the target object back to a designated terminal, where the monitoring information includes: the spectral information, the other types of index information;

the spectrum sensor system is specifically used for analyzing a first designated feature corresponding to the spectrum data; evaluating a first type of index information based on the first specified feature;

and/or the number of the groups of groups,

the spectrum sensor system is specifically used for analyzing a second specified characteristic corresponding to the spectrum data; inputting the second designated feature into a prediction model to obtain second type index information;

the first specified feature is: spectral characteristics of hemoglobin and oxygenated hemoglobin, the first type of index information being: tissue oxygen and levels;

and/or the number of the groups of groups,

2. The spectroscopic sensor-based monitoring apparatus as set forth in claim 1, wherein,

the prediction model is as follows: the flow prediction model, the second type index information is: liquid information;

and/or the number of the groups of groups,

3. The spectroscopic sensor-based monitoring apparatus as set forth in claim 2, wherein,

the liquid information includes at least one of: the pH value, sugar content and solid solution content of the liquid;

the mechanical information includes at least one of: acceleration, amount of shake, eccentricity.

4. The spectral sensor-based monitoring apparatus of claim 2, wherein the monitoring apparatus further comprises: a flow sensor system, a mechanical sensor system;

5. The spectroscopic sensor-based monitoring apparatus as set forth in claim 4, wherein,

the communication system is used for carrying out data fusion on first liquid information and second liquid information, and taking the fused result as final liquid information, wherein the first liquid information is liquid information read from the spectrum sensor, and the second liquid information is liquid information read from the flow sensor;

6. The spectral sensor-based monitoring apparatus of claim 1, wherein the second specified feature comprises at least one of:

spectral reflectance;

spectral absorptivity.

7. A monitoring device based on a spectroscopic sensor as claimed in any one of claims 1 to 6, characterized in that,

the number of the spectrum sensor systems is one, and the spectrum sensor system comprises: a spectrum sensor and a spectrum generator;

or alternatively, the first and second heat exchangers may be,

8. The spectroscopic sensor-based monitoring apparatus as set forth in claim 7, wherein,

in the case where the number of the spectrum sensor systems is 3, the spectrum sensors are arranged in a Y-type array.

9. The spectroscopic sensor-based monitoring apparatus as set forth in claim 7, wherein,

the spectrum sensor and the spectrum generator in the spectrum sensor system are two devices which are arranged separately.

10. The spectroscopic sensor-based monitoring apparatus as set forth in claim 7, wherein,

the spectral generator is operable to step up the spectral frequency range from low to high frequencies.