CN114886421B - Near infrared-based high-precision noninvasive blood glucose concentration detection system and method - Google Patents
Near infrared-based high-precision noninvasive blood glucose concentration detection system and method Download PDFInfo
- Publication number
- CN114886421B CN114886421B CN202210523520.5A CN202210523520A CN114886421B CN 114886421 B CN114886421 B CN 114886421B CN 202210523520 A CN202210523520 A CN 202210523520A CN 114886421 B CN114886421 B CN 114886421B
- Authority
- CN
- China
- Prior art keywords
- blood
- spectrum data
- glucose concentration
- near infrared
- blood glucose
- 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.)
- Active
Links
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/145—Measuring characteristics of blood in vivo, e.g. gas concentration or pH-value ; Measuring characteristics of body fluids or tissues, e.g. interstitial fluid or cerebral tissue
- A61B5/14532—Measuring characteristics of blood in vivo, e.g. gas concentration or pH-value ; Measuring characteristics of body fluids or tissues, e.g. interstitial fluid or cerebral tissue for measuring glucose, e.g. by tissue impedance measurement
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/145—Measuring characteristics of blood in vivo, e.g. gas concentration or pH-value ; Measuring characteristics of body fluids or tissues, e.g. interstitial fluid or cerebral tissue
- A61B5/1455—Measuring characteristics of blood in vivo, e.g. gas concentration or pH-value ; Measuring characteristics of body fluids or tissues, e.g. interstitial fluid or cerebral tissue using optical sensors, e.g. spectral photometrical oximeters
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Physics & Mathematics (AREA)
- Biomedical Technology (AREA)
- Medical Informatics (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Engineering & Computer Science (AREA)
- Veterinary Medicine (AREA)
- Heart & Thoracic Surgery (AREA)
- Optics & Photonics (AREA)
- Molecular Biology (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Emergency Medicine (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The invention relates to a near infrared detection technology, and discloses a near infrared-based high-precision noninvasive blood glucose concentration detection system and method, which solve the problem that noninvasive acquisition of blood spectrum data is easily influenced by environmental factors and physiological parameters of detected personnel so as to cause low prediction precision. According to the invention, near infrared spectrum data of blood under two conditions of invasive collection and noninvasive collection of a tested person are firstly obtained respectively, wavelength fitting is carried out on the blood spectrum data of the invasive collection and the noninvasive collection, a wavelength fitting formula is calculated and obtained, then the blood glucose concentration of the same tested person is calibrated by adopting a traditional blood glucose concentration detection method, a spectrum model between the blood spectrum data of the invasive collection and a blood glucose concentration calibration value is established, finally, the blood spectrum data of unknown blood glucose concentration is obtained noninvasively, the blood spectrum data of the blood is converted into the blood spectrum data of the invasive collection through the wavelength fitting formula, and further the blood glucose spectrum model of the invasive collection is adopted to rapidly predict the blood spectrum data of the blood.
Description
Technical Field
The invention relates to a near infrared detection technology, in particular to a near infrared-based high-precision noninvasive blood glucose concentration detection system and method.
Background
Diabetes is a global and common endocrine disorder. According to the current global diabetes research report, the number of diabetics in the world exceeds 4 hundred million, and diabetes is one of the most serious public health problems in the world. The number of diabetics in China is up to about 1.5 hundred million, and the diabetes mellitus is the first major country of diabetes mellitus, and the diabetes mellitus presents an increasing trend and a younger trend, so that the diabetes mellitus prevention and treatment method has very positive significance. The current blood sugar detection method is to take blood samples from the fingertips of patients for detection, and the method is also called invasive detection, and has the defects that although the detection accuracy is high, the operation of the method causes pain to the patients, the fingertips take blood to risk infection of other diseases, and psychological stress is caused to the patients; the detection time is long, and the real-time performance is poor; if the blood glucose test is frequently used for a long time, the test cost is increased, which is unfavorable for frequent test of the blood glucose. To overcome these shortcomings, many organizations and organizations worldwide have developed studies of noninvasive blood glucose testing techniques.
Currently, noninvasive blood glucose detection methods mainly comprise an electrochemical method and an optical method. Among these, electrochemical methods are also classified as minimally invasive, because they produce a relatively strong skin irritation. In the optical method, the near infrared noninvasive blood glucose detection technology is considered to be the noninvasive blood glucose detection technology with the most promising application prospect due to the advantages of high precision, small pollution, low cost and the like.
In the existing noninvasive blood glucose near-infrared detection technology, a noninvasive acquisition prediction method is mostly adopted, noninvasive acquisition of near-infrared spectrum data is carried out on blood in a finger of a person to be detected by using finger-clamped near-infrared acquisition equipment at intervals of tissues such as human skin, and a spectrum model of the near-infrared spectrum data and blood glucose concentration is established to predict blood glucose concentration. Although the method can realize rapid and noninvasive detection, the method directly adopts the corresponding relation between noninvasively acquired blood spectrum data and blood glucose concentration for spectrum modeling, and the noninvasive acquired spectrum data is easily influenced by environmental factors and physiological parameters (including but not limited to environmental temperature, environmental humidity, systolic pressure of a detected person, diastolic pressure of the detected person, pulse rate of the detected person, body temperature of the detected person and the like) of the detected person, so that the problem of poor stability and low precision of the noninvasive acquired spectrum data is caused, the prediction capability of a spectrum model is further influenced, and the prediction precision of the noninvasive blood glucose detection technology is reduced.
Disclosure of Invention
The technical problems to be solved by the invention are as follows: the system and the method for detecting the high-precision noninvasive blood glucose concentration based on the near infrared solve the problem that the prediction precision is low due to the fact that noninvasive blood spectrum data are easily affected by environmental factors and physiological parameters of detected personnel, effectively improve the accuracy and the real-time of blood glucose concentration prediction, and realize the rapid detection of the high-precision noninvasive blood glucose concentration.
The technical scheme adopted for solving the technical problems is as follows:
in one aspect, the present invention provides a near infrared-based high-precision noninvasive blood glucose concentration detection system, comprising:
the spectrum data acquisition module is used for acquiring blood near infrared spectrum data under two conditions of invasive acquisition and noninvasive acquisition of a tested person respectively;
the wavelength fitting module is used for performing wavelength fitting on the blood near infrared spectrum data sample which is collected in a invasive way and the blood near infrared spectrum data sample which is collected in a non-invasive way, and calculating and obtaining a wavelength fitting formula;
The blood glucose concentration calibration module is used for calibrating the blood glucose concentration in the blood of the same person to be tested by adopting a traditional blood glucose concentration detection method;
the spectrum model building module is used for building a spectrum model between the invasively acquired blood spectrum data and the blood glucose concentration calibration value;
the spectrum data conversion module is used for converting the blood spectrum data of the unknown blood sugar concentration which is collected in a non-invasive way into the blood spectrum data which is collected in a invasive way through a wavelength fitting formula;
and the blood glucose concentration prediction module is used for predicting the blood glucose concentration by utilizing the spectrum model based on the blood spectrum data obtained through conversion and collected in an invasive way.
Further, the spectrum data acquisition module is specifically configured to acquire spectrum data by using a near infrared fiber probe to sample blood from fingertips of m tested persons, where the near infrared fiber probe is in direct contact with the blood from fingertips of the tested persons, and acquire m pieces of spectrum data; the same near infrared optical fiber probe is used for collecting spectral data of finger tip blood of the m tested persons through human finger skin tissues; and acquiring m pieces of spectrum data.
Further, the wavelength fitting module is specifically configured to calculate a fitting coefficient β i between the two wavelength points through mean value conversion of spectrum data between the two invasive acquisition and the non-invasive acquisition:
wherein, beta i is the fitting coefficient corresponding to the ith wavelength point in the band range of the near infrared spectrum acquisition equipment; p m is the light intensity value of the ith wavelength point in the mth invasive acquisition blood spectrum data; s m is the light intensity value of the ith wavelength point in the mth noninvasive collected blood spectrum data;
according to the invasive collected blood spectrum data and the noninvasive collected blood spectrum data, calculating a wavelength fitting formula by combining fitting coefficients corresponding to all the wavelength points:
(H1,H2,Hi……Hn)＝(β1*Z1,β2*Z2,βi*Zi……βn*Zn).
Wherein H i is the light intensity value of the ith wavelength point in the noninvasively acquired blood spectrum data, and Z i is the light intensity value of the ith wavelength point in the noninvasively acquired blood spectrum data.
Furthermore, the blood glucose concentration calibration module is specifically configured to calibrate the blood glucose concentrations in blood that are collected by the m tested persons in an invasive manner one by using a glucose oxidase electrode measurement method, so as to obtain m blood glucose concentration calibration values.
Furthermore, the spectrum model building module is specifically configured to perform linear fitting on m pieces of blood spectrum data that are collected in an invasive manner and m blood glucose concentration calibration values through a partial least square method, so as to obtain a spectrum model.
Further, the spectrum data conversion module is specifically configured to convert non-invasively collected blood spectrum data of unknown blood glucose concentration into invasively collected blood spectrum data through a wavelength fitting formula:
A＝(β1*X1,β2*X2,βi*Xi……βn*Xn)；
Wherein X i is the light intensity value corresponding to the ith wavelength point in the blood spectrum data of the non-invasively acquired unknown blood glucose concentration.
On the other hand, the invention also provides a near infrared-based high-precision noninvasive blood glucose concentration detection method, which comprises the following steps of:
s1, acquiring blood near infrared spectrum data of a tested person under two conditions of invasive acquisition and noninvasive acquisition respectively;
s2, performing wavelength fitting on the blood near infrared spectrum data acquired in a invasive way and the blood near infrared spectrum data acquired in a non-invasive way, and calculating to acquire a wavelength fitting formula;
S3, calibrating the blood glucose concentration in blood of the same tested person by adopting a traditional blood glucose concentration detection method;
S4, establishing a spectrum model between the invasively collected blood spectrum data and the blood glucose concentration calibration value;
S5, converting the blood spectrum data of the unknown blood sugar concentration which are collected in a non-invasive manner into blood spectrum data which are collected in a invasive manner through a wavelength fitting formula;
s6, based on the blood spectrum data obtained through conversion and collected in an invasive mode, predicting the blood sugar concentration by using the spectrum model.
Further, in step S1, the method for acquiring the near infrared spectrum data of the blood that is invasively acquired by the tested person is as follows:
The method comprises the steps that a near infrared optical fiber probe is used for collecting spectral data of finger tip blood samples of m tested persons, the near infrared optical fiber probe is in direct contact with the finger tip blood of the tested persons, and m pieces of spectral data are collected;
the method for acquiring the blood near infrared spectrum data of the tested person in a noninvasive way comprises the following steps:
the same near infrared optical fiber probe is used for collecting spectral data of finger tip blood of the m tested persons through human finger skin tissues; and acquiring m pieces of spectrum data.
Further, in step S2, the wavelength fitting is performed on the invasively collected near infrared spectrum data of the blood and the non-invasively collected near infrared spectrum data of the blood, and a wavelength fitting formula is calculated and obtained, which specifically includes:
Assuming that for the ith wavelength point in the band range of the near infrared spectrum acquisition device used, the light intensity value of M pieces of noninvasive blood spectrum data is M 1＝(P1,P2,……Pm), the light intensity value of M pieces of noninvasive blood spectrum data is N 1＝(S1,S2,……Sm), calculating a fitting coefficient beta i between the two wavelength points through mean value conversion of the spectrum data between the invasive acquisition and the noninvasive acquisition, wherein the fitting coefficient beta i is:
Then n fitting coefficients, β 1,β2,βi……βn respectively, can be obtained for n wavelength points in the near infrared spectrum acquisition device band range;
For spectral data of blood of a single tested person, the spectral data H= (H 1,H2,Hi……Hn) of the blood is set, the spectral data Z= (Z 1,Z2,Zi……Zn) of the blood is collected in a non-invasive way, and a wavelength fitting formula can be calculated by combining fitting coefficients (beta 1,β2,βi……βn) corresponding to n wavelength points:
(H1,H2,Hi……Hn)＝(β1*Z1,β2*Z2,βi*Zi……βn*Zn).
in step S3, the blood glucose concentrations in the blood collected by the m tested persons in a invasive manner are calibrated one by using a glucose oxidase electrode measurement method, so as to obtain m blood glucose concentration calibration values.
Further, in step S4, the spectral model is obtained by linearly fitting the m pieces of invasively collected blood spectral data with the m blood glucose concentration calibration values by the partial least square method.
Further, in step S5, the converting the non-invasively collected blood spectrum data of unknown blood glucose concentration into the invasively collected blood spectrum data by the wavelength fitting formula specifically includes:
blood spectrum data are acquired in a non-invasive acquisition mode of a tested person with blood glucose concentration to be tested, the blood spectrum data are set to be X= (X 1,X2,Xi……Xn), and the non-invasive acquisition blood spectrum data are converted into invasive acquisition blood spectrum data A= (beta 1*X1,β2*X2,βi*Xi……βn*Xn) through a wavelength fitting formula.
The beneficial effects of the invention are as follows:
According to the method, the corresponding relation between the invasive blood spectral data and the blood glucose concentration is adopted for spectral modeling, interference of external factors is avoided, the accuracy of a spectral model is improved, the wavelength fitting conversion relation is further built through the invasive collection and the noninvasive collection of the blood spectral data, the noninvasive collection of the blood spectral data after the wavelength fitting conversion is predicted through the spectral model built through the invasive blood spectral data and the blood glucose concentration, the problem that the noninvasive collection of the blood spectral data is easily influenced by environmental factors and physiological parameters of detected personnel, so that the prediction accuracy is low is solved, the blood glucose concentration prediction accuracy and the prediction instantaneity are effectively improved, and the high-accuracy noninvasive blood glucose concentration rapid detection is realized.
Drawings
FIG. 1 is a flow chart of a near infrared based high precision noninvasive blood glucose concentration detection method in an embodiment of the present invention;
Fig. 2 is a block diagram of a high-precision noninvasive blood glucose concentration detection system based on near infrared in an embodiment of the present invention.
Detailed Description
The invention aims to provide a near infrared-based high-precision noninvasive blood glucose concentration detection system and method, which solve the problem that noninvasive blood spectrum data are easily influenced by environmental factors and physiological parameters of detected personnel to further cause low prediction precision, effectively improve the accuracy and the real-time of blood glucose concentration prediction, and realize the rapid detection of high-precision noninvasive blood glucose concentration. According to the scheme, near infrared spectrum data of blood under two conditions of invasive collection and noninvasive collection of a tested person are firstly obtained respectively, wavelength fitting is carried out on the blood spectrum data of the invasive collection and the noninvasive collection, a wavelength fitting formula is calculated and obtained, then the blood glucose concentration of the same tested person is calibrated by adopting a traditional blood glucose concentration detection method, a spectrum model between the blood spectrum data of the invasive collection and a blood glucose concentration calibration value is built, finally, the blood spectrum data of unknown blood glucose concentration is obtained noninvasively, the blood spectrum data of the blood is converted into the blood spectrum data of the invasive collection through the wavelength fitting formula, and further the blood glucose spectrum model of the invasive collection is used for fast prediction.
Examples
As shown in fig. 1, the near-infrared-based high-precision noninvasive blood glucose concentration detection method provided in the present embodiment includes the following implementation steps:
step 101, blood near infrared spectrum data of a tested person under two conditions of invasive collection and noninvasive collection are respectively obtained;
In a specific example, the collecting part of the spectrum data is selected as a finger tip, the blood of the part is rich, other interfering substances are less, the instrument is convenient to collect, the design difficulty of the collecting instrument is reduced, the influence of physiological and psychological factors and external factors of the collecting instrument is not easy to influence, and the accuracy of the spectrum data can be effectively improved.
The method for invasively collecting the near infrared spectrum data of the blood comprises the following steps: the method comprises the steps that a near infrared optical fiber probe is used for collecting spectral data of a blood sample taken by a fingertip of a person to be tested, and the near infrared optical fiber probe is in direct contact with the blood of the fingertip of the person to be tested;
The method for noninvasively collecting the near infrared spectrum data of the blood comprises the following steps: the same near infrared optical fiber probe is used for collecting spectral data of the blood of the fingertip of the tested person through tissues such as the skin of the finger of the human body.
According to the research, in the practical application of the near infrared spectrum noninvasive blood glucose analysis technology, when the number of blood samples reaches 400, namely, the spectrum model established corresponding to 400 spectrum data has good stability and prediction capability, then in this embodiment, 400 blood samples of the tested person are selected as original reference samples to ensure the prediction capability of the blood glucose concentration spectrum model. Each blood sample of the tested person is divided into two parts, wherein one part is used for collecting invasive blood near infrared spectrum data, and the other part is used for subsequent blood glucose concentration calibration.
The wavelength range of the near infrared spectrum acquisition device adopted in the embodiment is 800 nm-1480 nm, a good fitting relation is formed between blood spectrum data acquired by a near infrared spectrometer in the wavelength range and blood glucose concentration, the resolution of the near infrared acquisition device is 10nm, the distribution mode of wavelength points is wavelength division, the number of the wavelength points is K=1+ (1480-800)/10=69, and then the fact that each piece of blood spectrum data is actually expressed as a light intensity value matrix on 69 wavelength points can be known.
Step 102, performing wavelength fitting on the blood near infrared spectrum data acquired in a invasive way and the blood near infrared spectrum data acquired in a noninvasive way, and calculating and acquiring a wavelength fitting formula:
Because the blood sample adopted by the traditional blood glucose concentration detection method is in a mode of invasive collection, the detection precision of the mode is high, in the actual non-invasive detection practical application, the collected and acquired spectral data is in a non-invasive mode, namely, the spectral data is collected on the measured human blood by being separated by human skin and other tissues, and the detection precision of the mode is low due to the influence of the environment or the factors of the measured human, therefore, the non-invasive collection and the invasive collection of the blood spectral data are required to be subjected to wavelength fitting, and a wavelength fitting formula is calculated and acquired for the conversion of the spectral data in two modes, so that the prediction precision of the non-invasive blood glucose is improved.
In this embodiment, step 101 indicates that the spectrum data of the blood collected in an invasive manner and the spectrum data of the blood collected in a non-invasive manner are both represented as a matrix of light intensity values at 69 wavelength points, and for a single wavelength point, a certain linear fitting relationship exists between the spectrum data of the two points, and the fitting coefficient can be calculated through the linear fitting relationship. For the whole spectrum data, a certain linear fitting relation exists between the two spectrum data, and a wavelength fitting formula can be further calculated and obtained by combining fitting coefficients of all single wavelength points so as to realize noninvasive acquisition and conversion between the noninvasively acquired blood spectrum data, so that the noninvasively acquired blood spectrum data are converted into the noninvasively acquired blood spectrum data for spectrum prediction, and the prediction accuracy of noninvasive blood glucose concentration is improved.
The specific wavelength fitting implementation mode is as follows:
When the wavelength point is set to 800nm, the light intensity value of 400 pieces of invasive blood spectrum data is M 1＝(P1,P2,……P400), the light intensity value of 400 pieces of noninvasive blood spectrum data is N 1＝(S1,S2,……S400), when linear fitting calculation is carried out on the spectrum data, the average value of the spectrum data can well reflect the interval range where the spectrum data is located, namely, the spectrum average value can be calculated by replacing the spectrum data approximately, and the fitting coefficient beta 1 between the two wavelength points is calculated through average value conversion of the spectrum data between invasive acquisition and noninvasive acquisition:
Similarly, when the wavelength points are (810 nm, 8230 nm and … … 1480 nm), the corresponding fitting coefficient is (beta 2,β3,……β69);
for the spectral data of a certain single tested human blood, the spectral data H= (H 1,H2,……H69) of the invasive collected blood is set, the spectral data Z= (Z 1,Z2,……Z69) of the noninvasive collected blood is combined with the fitting coefficient (beta 1,β2,……β69) corresponding to the wavelength points (800 nm,810nm and … … 1480 nm), and then the wavelength fitting formula can be calculated:
(H1,H2,……H69)＝(β1*Z1,β2*Z2,……β69*Z69).
step 103, calibrating the blood glucose concentration in blood of the same tested person by adopting a traditional blood glucose concentration detection method:
The traditional blood glucose concentration detection method adopts a glucose oxidase electrode measurement method, is the most commonly used blood glucose concentration measurement method, has the advantages of high detection precision, has the disadvantages of high cost and poor detection effect of a blood sampling tested human body, and can ensure the accuracy of a blood glucose concentration calibration value to the greatest extent by calibrating the blood glucose concentration of the same tested human body by adopting the traditional blood glucose concentration detection method.
In the practical application of the near infrared spectrum noninvasive blood glucose analysis technology in this embodiment, the accuracy of calibrating the blood glucose concentration of the blood sample for modeling will directly affect the prediction effect of the spectrum model, if the calibration deviation of the blood sample itself for modeling is large, the deviation is very easy to be introduced into the near infrared spectrum model, so that the deviation between the actual blood glucose concentration of the blood sample to be measured and the near infrared predicted blood glucose concentration is large, and the prediction accuracy of the near infrared noninvasive blood glucose analysis technology is greatly reduced. Therefore, the method adopts the traditional glucose oxidase electrode measurement method to calibrate the blood glucose concentration of the measured human blood, so that the calibration value accuracy can be greatly improved, and the prediction capacity of a spectrum model is further improved. The blood glucose concentration of 400 tested human blood is calibrated one by adopting a traditional blood glucose concentration detection method, and 400 blood glucose concentration calibration values are obtained.
Step 104, establishing a spectrum model between the invasively acquired blood spectrum data and the blood glucose concentration calibration value:
In the step, 400 pieces of invasive blood spectrum data are corresponding to 400 blood glucose concentration calibration values obtained by a traditional blood glucose concentration detection method one by one, and spectrum modeling is carried out by a partial least square method to obtain an invasive collected blood glucose spectrum model.
The partial least square method is the most commonly used spectral data modeling method and is also the linear fitting method which is most suitable for the characteristics of the blood spectral data, for the spectral data of the blood which is collected in an invasive way and the blood which is collected in a non-invasive way, the preprocessing mode of SG second order derivation is adopted to eliminate the influence of baseline drift and gentle background interference, and meanwhile, higher resolution and sensitivity than those of the original spectrum are provided, and the spectral model effect is improved. Therefore, in the embodiment, 400 pieces of invasive blood spectrum data and 400 blood glucose concentration calibration values obtained by a traditional blood glucose concentration detection method are linearly fitted through a partial least square method to obtain an invasive collected blood glucose spectrum model which is used for predicting blood glucose concentration of human blood to be detected.
Step 105, converting the blood spectrum data of the non-invasively acquired unknown blood glucose concentration into the blood spectrum data of the invasively acquired blood through a wavelength fitting formula, and further predicting the blood glucose concentration by using the spectrum model based on the obtained blood spectrum data of the invasively acquired blood through conversion:
In practical noninvasive blood glucose detection application, a noninvasive mode is adopted to collect blood spectrum data under skin tissues of a tested person, instead of a traditional invasive blood detection mode, in order to improve prediction accuracy, the blood spectrum data collected under the noninvasive mode needs to be converted into blood spectrum data under the invasive acquisition mode through a wavelength fitting formula, then a blood glucose spectrum model collected in a invasive mode is adopted to conduct real-time prediction, a predicted value of blood glucose concentration of the tested person is obtained, and real-time guidance data is provided for medical services.
In this embodiment, blood spectrum data is obtained by performing noninvasive collection on a person to be tested for blood glucose concentration, where the blood spectrum data is set to x= (X 1,X2,……X69), the noninvasive collected blood spectrum data is converted into invasive collected blood spectrum data a= (β 1*X1,β2*X2,……β69*X69) by using a wavelength fitting formula, and then the invasive collected blood glucose spectrum data a is further predicted by using an invasive collected blood glucose spectrum model to obtain a predicted blood glucose concentration value, and the predicted blood glucose concentration value is the blood glucose concentration of the person to be tested in the noninvasive collection mode, so as to provide real-time guidance data for medical services.
In addition, the embodiment also provides a near infrared-based high-precision noninvasive blood glucose concentration detection system, which comprises a spectrum data acquisition module, a wavelength fitting module, a blood glucose concentration calibration module, a spectrum model building module, a spectrum data conversion module and a blood glucose concentration prediction module, as shown in fig. 2; the specific functions of each module are introduced as follows:
the spectrum data acquisition module is used for acquiring blood near infrared spectrum data under two conditions of invasive acquisition and noninvasive acquisition of a tested person respectively;
the wavelength fitting module is used for performing wavelength fitting on the blood near infrared spectrum data sample which is collected in a invasive way and the blood near infrared spectrum data sample which is collected in a non-invasive way, and calculating and obtaining a wavelength fitting formula;
The blood glucose concentration calibration module is used for calibrating the blood glucose concentration in the blood of the same person to be tested by adopting a traditional blood glucose concentration detection method;
the spectrum model building module is used for building a spectrum model between the invasively acquired blood spectrum data and the blood glucose concentration calibration value;
the spectrum data conversion module is used for converting the blood spectrum data of the unknown blood sugar concentration which is collected in a non-invasive way into the blood spectrum data which is collected in a invasive way through a wavelength fitting formula;
and the blood glucose concentration prediction module is used for predicting the blood glucose concentration by utilizing the spectrum model based on the blood spectrum data obtained through conversion and collected in an invasive way.
Finally, it should be noted that the above examples are only preferred embodiments and are not intended to limit the invention. It should be noted that modifications, equivalents, improvements and others may be made by those skilled in the art without departing from the spirit of the invention and the scope of the claims, and are intended to be included within the scope of the invention.
Claims (12)
1. A near infrared-based high-precision noninvasive blood glucose concentration detection system, comprising:
the spectrum data acquisition module is used for acquiring blood near infrared spectrum data under two conditions of invasive acquisition and noninvasive acquisition of a tested person respectively;
the wavelength fitting module is used for performing wavelength fitting on the blood near infrared spectrum data sample which is collected in a invasive way and the blood near infrared spectrum data sample which is collected in a non-invasive way, and calculating and obtaining a wavelength fitting formula;
The blood glucose concentration calibration module is used for calibrating the blood glucose concentration in the blood of the same person to be tested by adopting a traditional blood glucose concentration detection method;
the spectrum model building module is used for building a spectrum model between the invasively acquired blood spectrum data and the blood glucose concentration calibration value;
the spectrum data conversion module is used for converting the blood spectrum data of the unknown blood sugar concentration which is collected in a non-invasive way into the blood spectrum data which is collected in a invasive way through a wavelength fitting formula;
and the blood glucose concentration prediction module is used for predicting the blood glucose concentration by utilizing the spectrum model based on the blood spectrum data obtained through conversion and collected in an invasive way.
2. A near infrared based high precision noninvasive blood glucose concentration detection system of claim 1, wherein,
The spectrum data acquisition module is specifically used for acquiring spectrum data of finger tip blood samples of m tested persons by using a near infrared optical fiber probe, wherein the near infrared optical fiber probe is in direct contact with the finger tip blood of the tested persons, and acquiring m pieces of spectrum data; the same near infrared optical fiber probe is used for collecting spectral data of finger tip blood of the m tested persons through human finger skin tissues; and acquiring m pieces of spectrum data.
3. A near infrared based high precision noninvasive blood glucose concentration detection system of claim 2, wherein,
The wavelength fitting module is specifically configured to calculate a fitting coefficient β i between the two wavelength points through mean value conversion of spectrum data between the two invasive acquisition and the non-invasive acquisition:
wherein, beta i is the fitting coefficient corresponding to the ith wavelength point in the band range of the near infrared spectrum acquisition equipment; p m is the light intensity value of the ith wavelength point in the mth invasive acquisition blood spectrum data; s m is the light intensity value of the ith wavelength point in the mth noninvasive collected blood spectrum data;
according to the invasive collected blood spectrum data and the noninvasive collected blood spectrum data, calculating a wavelength fitting formula by combining fitting coefficients corresponding to all the wavelength points:
(H1,H2,Hi……Hn)＝(β1*Z1,β2*Z2,βi*Zi……βn*Zn);
Wherein H i is the light intensity value of the ith wavelength point in the noninvasively acquired blood spectrum data, and Z i is the light intensity value of the ith wavelength point in the noninvasively acquired blood spectrum data.
4. A near infrared based high precision noninvasive blood glucose concentration detection system of claim 3, wherein,
The blood glucose concentration calibration module is specifically used for calibrating blood glucose concentrations in blood of the m tested persons in a invasive way one by adopting a glucose oxidase electrode measurement method, and obtaining m blood glucose concentration calibration values.
5. A near infrared based high precision noninvasive blood glucose concentration detection system of claim 4, wherein,
The spectrum model building module is specifically used for carrying out linear fitting on m pieces of blood spectrum data which are collected in a invasive way and m blood glucose concentration calibration values through a partial least square method to obtain a spectrum model.
6. A near infrared based high precision noninvasive blood glucose concentration detection system of claim 5, wherein,
The spectrum data conversion module is specifically used for converting blood spectrum data of unknown blood sugar concentration which is collected in a noninvasive manner into blood spectrum data which is collected in a noninvasive manner through a wavelength fitting formula:
A＝(β1*X1,β2*X2,βi*Xi……βn*Xn)；
Wherein X i is the light intensity value corresponding to the ith wavelength point in the blood spectrum data of the non-invasively acquired unknown blood glucose concentration.
7. The near infrared-based high-precision noninvasive blood glucose concentration detection method is characterized by comprising the following steps of:
s1, acquiring blood near infrared spectrum data of a tested person under two conditions of invasive acquisition and noninvasive acquisition respectively;
s2, performing wavelength fitting on the blood near infrared spectrum data acquired in a invasive way and the blood near infrared spectrum data acquired in a non-invasive way, and calculating to acquire a wavelength fitting formula;
S3, calibrating the blood glucose concentration in blood of the same tested person by adopting a traditional blood glucose concentration detection method;
S4, establishing a spectrum model between the invasively collected blood spectrum data and the blood glucose concentration calibration value;
S5, converting the blood spectrum data of the unknown blood sugar concentration which are collected in a non-invasive manner into blood spectrum data which are collected in a invasive manner through a wavelength fitting formula;
s6, based on the blood spectrum data obtained through conversion and collected in an invasive mode, predicting the blood sugar concentration by using the spectrum model.
8. The near infrared based high-precision noninvasive blood glucose concentration detection method of claim 7, wherein,
In step S1, the method for acquiring the near infrared spectrum data of blood collected by the tested person in a invasive way is as follows:
The method comprises the steps that a near infrared optical fiber probe is used for collecting spectral data of finger tip blood samples of m tested persons, the near infrared optical fiber probe is in direct contact with the finger tip blood of the tested persons, and m pieces of spectral data are collected;
the method for acquiring the blood near infrared spectrum data of the tested person in a noninvasive way comprises the following steps:
the same near infrared optical fiber probe is used for collecting spectral data of finger tip blood of the m tested persons through human finger skin tissues; and acquiring m pieces of spectrum data.
9. The near infrared based high-precision noninvasive blood glucose concentration detection method of claim 8, wherein,
In step S2, the wavelength fitting is performed on the blood near infrared spectrum data collected in an invasive manner and the blood near infrared spectrum data collected in a non-invasive manner, and a wavelength fitting formula is calculated and obtained, which specifically includes:
Assuming that for the ith wavelength point in the band range of the near infrared spectrum acquisition device used, the light intensity value of M pieces of noninvasive blood spectrum data is M 1＝(P1,P2,……Pm), the light intensity value of M pieces of noninvasive blood spectrum data is N 1＝(S1,S2,……Sm), calculating a fitting coefficient beta i between the two wavelength points through mean value conversion of the spectrum data between the invasive acquisition and the noninvasive acquisition, wherein the fitting coefficient beta i is:
Then n fitting coefficients, β 1,β2,βi……βn respectively, can be obtained for n wavelength points in the near infrared spectrum acquisition device band range;
For spectral data of blood of a single tested person, the invasive collected blood spectral data h=h 1,H2,Hi……Hn) is set, the noninvasive collected blood spectral data z= (Z 1,Z2,Zi……Zn) is combined with fitting coefficients (beta 1,β2,βi……βn) corresponding to n wavelength points, and then a wavelength fitting formula can be calculated:
(H1,H2,Hi……Hn)＝(β1*Z1,β2*Z2,βi*Zi……βn*Zn).
10. the near infrared based high-precision noninvasive blood glucose concentration detection method of claim 9, wherein,
In step S3, calibrating the blood glucose concentration in the blood of the m tested persons by using a glucose oxidase electrode measurement method one by one to obtain m blood glucose concentration calibration values.
11. The near infrared based high-precision noninvasive blood glucose concentration detection method of claim 10, wherein,
In step S4, linear fitting is carried out on m pieces of blood spectrum data which are collected in a invasive way and m blood glucose concentration calibration values through a partial least square method, so that a spectrum model is obtained.
12. The near infrared based high-precision noninvasive blood glucose concentration detection method of claim 11, wherein,
In step S5, the converting, by a wavelength fitting formula, the blood spectrum data of the non-invasively collected unknown blood glucose concentration into the blood spectrum data of the invasively collected blood specifically includes:
blood spectrum data are acquired in a non-invasive acquisition mode of a tested person with blood glucose concentration to be tested, the blood spectrum data are set to be X= (X 1,X2,Xi……Xn), and the non-invasive acquisition blood spectrum data are converted into invasive acquisition blood spectrum data A= (beta 1*X1,β2*X2,βi*Xi……βn*Xn) through a wavelength fitting formula.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210523520.5A CN114886421B (en) | 2022-05-13 | 2022-05-13 | Near infrared-based high-precision noninvasive blood glucose concentration detection system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210523520.5A CN114886421B (en) | 2022-05-13 | 2022-05-13 | Near infrared-based high-precision noninvasive blood glucose concentration detection system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114886421A CN114886421A (en) | 2022-08-12 |
| CN114886421B true CN114886421B (en) | 2024-10-18 |
Family
ID=82722552
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210523520.5A Active CN114886421B (en) | 2022-05-13 | 2022-05-13 | Near infrared-based high-precision noninvasive blood glucose concentration detection system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114886421B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116602668B (en) * | 2023-07-06 | 2023-10-31 | 深圳大学 | A fully automatic intelligent blood glucose detection system |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106473755A (en) * | 2016-11-30 | 2017-03-08 | 江西科技师范大学 | A kind of optical sound head for blood sugar monitoring |
| CN106551700A (en) * | 2015-09-28 | 2017-04-05 | 戴文招 | A kind of survey blood glucose method for measuring in several ways and demarcating elimination systematic error |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5054487A (en) * | 1990-02-02 | 1991-10-08 | Boston Advanced Technologies, Inc. | Laser systems for material analysis based on reflectance ratio detection |
| NZ527164A (en) * | 2001-01-26 | 2005-07-29 | Sensys Medical Inc | Noninvasive measurement of glucose through the optical properties of tissue |
| DE602005022388D1 (en) * | 2004-08-25 | 2010-09-02 | Panasonic Elec Works Co Ltd | Quantitative analyzer using a calibration curve |
| FR3038723B1 (en) * | 2015-07-07 | 2019-06-14 | Commissariat A L'energie Atomique Et Aux Energies Alternatives | METHOD OF ESTIMATING A QUANTITY OF ANALYTE IN A LIQUID |
| CN108937955A (en) * | 2017-05-23 | 2018-12-07 | 广州贝塔铁克医疗生物科技有限公司 | The adaptive wearable blood glucose bearing calibration of personalization and its means for correcting based on artificial intelligence |
| CN108634964A (en) * | 2018-05-04 | 2018-10-12 | 中国科学院遥感与数字地球研究所 | A kind of non-invasive blood sugar instrument based on spectrum |
| CN112568902A (en) * | 2020-12-15 | 2021-03-30 | 无锡轲虎医疗科技有限责任公司 | Noninvasive blood glucose calibration method based on blood glucose value |
-
2022
- 2022-05-13 CN CN202210523520.5A patent/CN114886421B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106551700A (en) * | 2015-09-28 | 2017-04-05 | 戴文招 | A kind of survey blood glucose method for measuring in several ways and demarcating elimination systematic error |
| CN106473755A (en) * | 2016-11-30 | 2017-03-08 | 江西科技师范大学 | A kind of optical sound head for blood sugar monitoring |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114886421A (en) | 2022-08-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110575181A (en) | Near-infrared spectroscopy non-invasive blood glucose detection network model training method | |
| CN108937955A (en) | The adaptive wearable blood glucose bearing calibration of personalization and its means for correcting based on artificial intelligence | |
| US6949070B2 (en) | Non-invasive blood glucose monitoring system | |
| Losoya-Leal et al. | State of the art and new perspectives in non-invasive glucose sensors | |
| JP2011520552A (en) | Method and system for non-invasive and optical detection of blood glucose using spectral data analysis | |
| CN111466921A (en) | Non-invasive blood glucose detector and detection method based on multi-source information perception and fusion | |
| CN111599470A (en) | Method for improving near-infrared noninvasive blood glucose detection precision | |
| CN113208586A (en) | Noninvasive blood glucose rapid diagnosis differential Raman spectroscopy system | |
| CN112022167A (en) | Noninvasive blood glucose detection method based on spectral sensor | |
| CN111227844A (en) | Noninvasive blood glucose detection device and detection method based on Raman scattering spectrum | |
| CN114886421B (en) | Near infrared-based high-precision noninvasive blood glucose concentration detection system and method | |
| CN112568902A (en) | Noninvasive blood glucose calibration method based on blood glucose value | |
| Yamakoshi et al. | A new non-invasive method for measuring blood glucose using instantaneous differential near infrared spectrophotometry | |
| Albalat et al. | Non-invasive blood glucose sensor: A feasibility study | |
| Ionescu et al. | Measuring and detecting blood glucose by methods non-invasive | |
| Shulei et al. | Non-invasive blood glucose measurement scheme based on near-infrared spectroscopy | |
| CN104983430B (en) | The blood-sugar detecting instrument of non-intrusion type | |
| CN105342627A (en) | Microwave-based glucose measuring system | |
| CN112587134A (en) | Noninvasive blood glucose detection method | |
| CN114739945A (en) | Method for bidirectionally correcting noninvasive detection blood glucose near infrared spectrum data | |
| CN105816186A (en) | Noninvasive type automated hand-held blood analyzer | |
| CN113974618B (en) | Noninvasive blood glucose testing method based on water peak blood glucose correction | |
| CN104921735A (en) | Microwave noninvasive blood glucose measurement system | |
| CN103876749A (en) | Built-in spectroscopy rapid blood glucose detection system | |
| CN205031268U (en) | Blood glucose measurement device based on microwave |
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 |