WO2023000530A1 - Method for improving accuracy of real-time concentration change trend during continuous monitoring of analyte concentration in animal body - Google Patents

Method for improving accuracy of real-time concentration change trend during continuous monitoring of analyte concentration in animal body Download PDF

Info

Publication number
WO2023000530A1
WO2023000530A1 PCT/CN2021/126936 CN2021126936W WO2023000530A1 WO 2023000530 A1 WO2023000530 A1 WO 2023000530A1 CN 2021126936 W CN2021126936 W CN 2021126936W WO 2023000530 A1 WO2023000530 A1 WO 2023000530A1
Authority
WO
WIPO (PCT)
Prior art keywords
time
analyte concentration
real
analyte
concentration
Prior art date
Application number
PCT/CN2021/126936
Other languages
French (fr)
Chinese (zh)
Inventor
韩洋
Original Assignee
苏州百孝医疗科技有限公司
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 苏州百孝医疗科技有限公司 filed Critical 苏州百孝医疗科技有限公司
Publication of WO2023000530A1 publication Critical patent/WO2023000530A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
    • 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, pH value; Measuring characteristics of body fluids or tissues, e.g. interstitial fluid, cerebral tissue
    • A61B5/14532—Measuring characteristics of blood in vivo, e.g. gas concentration, pH value; Measuring characteristics of body fluids or tissues, e.g. interstitial fluid, cerebral tissue for measuring glucose, e.g. by tissue impedance measurement
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/11—Complex mathematical operations for solving equations, e.g. nonlinear equations, general mathematical optimization problems
    • G06F17/12—Simultaneous equations, e.g. systems of linear equations
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B2503/00—Evaluating a particular growth phase or type of persons or animals
    • A61B2503/40—Animals
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B2503/00—Evaluating a particular growth phase or type of persons or animals
    • A61B2503/42—Evaluating a particular growth phase or type of persons or animals for laboratory research

Definitions

  • This application relates to the research field of the change of analyte concentration data and the real-time trend of analyte concentration change in the continuous monitoring process of animal body analyte concentration, especially involves judging whether the analyte concentration data is abnormal or suspicious, and judging whether to output analyte
  • the real-time trend of concentration change is a method for improving the accuracy of the real-time trend of analyte concentration change.
  • the aforementioned analytes include, but are not limited to: acetylcholine, amylase, bilirubin, cholesterol, chorionic gonadotropin, creatine kinase, sarcosine, creatinine, DNA, fructosamine, glucose, glutamate, growth hormone Classes, hormones, ketone bodies, lactic acid, peroxides, prostate-specific antigen, prothrombin, RNA, thyroid-stimulating hormone, and troponin, etc.
  • Analytes can also include drugs such as antibiotics (eg, gentamicin, vancomycin, etc.), digitoxin, digoxin, drugs of abuse, theophylline, and warfarin, among others.
  • the analyte concentration data is usually obtained by an analyte sensor, wherein the analyte sensor may be a glucose sensor, a ketone sensor or other sensors corresponding to the analyte.
  • the continuous analyte concentration monitoring system will output the analyte concentration data and concentration change trend according to a certain (collection) period (as shown in Figure 1, where the concentration change trend adopts representation of an arrow trend). It is relatively easy to obtain more accurate analyte concentration data through the analyte sensor, but to obtain relatively accurate analyte concentration change trends, especially the real-time trend of analyte concentration changes (referring to the current acquisition time given the analyte concentration data At the same time, the changing state of the analyte concentration) becomes relatively difficult.
  • the amount of change in the analyte concentration data D 2 (D 2 C(T)-C(T-n ⁇ t), n is a positive integer greater than 1)
  • the direction of concentration change can be obtained by the sign and size of D 1 and D 2 and the amount of change, according to the threshold interval classification of dD 1 and D 2 , the real-time arrow trend of concentration change is obtained.
  • one collection period with the current collection time T as the end point refers to 11:57 ⁇ 12:00, which contains 2 analyte concentration data;
  • the three acquisition periods with time T as the end point refer to 11:51-12:00, which include 4 analyte concentration data.
  • the longest time is determined according to different analyte types, analyte concentration change events, medical knowledge or characteristics of physiological characteristics.
  • a continuous 15-minute blood sugar trend is a blood sugar event; Under normal circumstances, it is generally considered that a certain period of time within three hours or less after a meal is a blood glucose event.
  • the change D 2 can be calculated, and the real-time arrow trend of the concentration change is given according to the threshold interval corresponding to dD 1 and D 2 (Table 1).
  • the upward angle and quantity of the arrow represent the rate of increase in the concentration of the analyte.
  • the method of judging the real-time trend of the above-mentioned analyte concentration change is no problem; however, when the analyte concentration data suddenly appears abnormal in a certain continuous and stable trend (for example, the analyte concentration data changes from rising to falling or falling to rising, the abnormality of the analyte sensor, the difference in individual metabolic function, the abnormal rise or fall of the concentration data caused by human intervention, etc.), which makes the analyte concentration data doubtful at this moment (
  • the data is also referred to as untrustworthy data or untrustworthy points hereinafter
  • the above method for judging the real-time trend of analyte concentration changes cannot accurately output the real-time trend of analyte concentration changes (according to the above method, the speed and direction of blood sugar changes can be calculated , can output the real-time trend of analyte concentration changes, but the accuracy rate is problematic), and even classify based on
  • the blood glucose concentration drops suddenly. According to the above judgment method, the blood glucose concentration is slowly declining at this moment. Because the concentration is close to the edge of the threshold interval, it may mislead the patient to take measures to avoid the blood glucose concentration from continuing to drop; or At the first two acquisition moments (when the acquisition period is 102 and 103), the blood glucose concentration is rising and falling, and at this moment (when the acquisition period is 104), the blood glucose concentration is rising, that is to say, at three consecutive acquisition moments (acquisition The period is 102 to 104 hours) respectively give the real-time trends of blood glucose concentration as rising, falling and rising, which will lead to poor user experience for patients. Therefore, the accuracy of the above method for judging the real-time trend of analyte concentration changes needs to be improved, and it is hoped to obtain a real-time trend summary chart as shown in FIG. 4 .
  • the present application introduces a method based on least square polynomial fitting (such as Equation 1) to fit the analyte concentration data C(t) with the current collection time T as the end point.
  • the above-mentioned first-order linear equation is the linear equation of the relationship between time t and analyte concentration C(t), so that at the acquisition time t, the analyte concentration data calculated by fitting and the analyte concentration output by the continuous analyte concentration monitoring system The cumulative difference between the concentration data satisfies the minimum.
  • the positive and negative sign of P corresponds to whether the concentration of the analyte is rising or falling
  • the absolute value of P corresponds to the fast or slow rate of change of the concentration of the analyte
  • the signs of P1 and P2 it can be quickly judged whether to output the real - time trend of the change of the analyte concentration. Specifically, if the signs are inconsistent (any one of which is 0 is also considered as inconsistent signs), it means that the analyte concentration data obtained at the collection time T is abnormal, or at least doubtful, and is untrustworthy data, and it is judged that the analyte is not output Real-time trend of concentration changes.
  • the sign of P 2 is positive, it means that the overall trend has been on the rise for a long period of time in the past (the duration of Tm 2 ⁇ t to T); and if the sign of P 1 is negative or equal to 0 at this moment , it means that in a short period of time in the past (the duration from Tm 1 ⁇ t to T), it is in a sudden decline or a sudden steady trend.
  • the analyte concentration data at the future collection time (T+ ⁇ t) is unknown, it is impossible to determine whether the current data belongs to the turning point from rising to falling, or the abrupt point of the overall rising stage. Therefore, no real-time trend of analyte concentration change is output at the acquisition time T.
  • the following method is further used to judge whether to output the real-time trend of the change of the analyte concentration.
  • FIG. 5 shows a schematic flow chart of the method for improving the accuracy of the real-time trend of the concentration change in the continuous monitoring process of the concentration of the analyte in the animal body.
  • P 3 represents the accumulation trend of the analyte concentration data in the m x collection periods before the current time T (that is, excluding the partial data including the current data). If the current data point is abnormally large (that is, the absolute value of P 1 is large) and does not have clinical significance at all, the previous period (T-(m 3 +m x ) ⁇ t to Tm x ⁇ t)m 3 acquisitions can be used Cumulative trending of analyte concentration data with high confidence within a period analyzes normal trends over past time periods. In addition, because the concentration of analytes in animals has a certain continuity, the cumulative trend of data in the previous period of m3 collection cycles can reflect the current trend to a certain extent.
  • m 2 is based on the analytes judged by medical research and experience The maximum time for which the concentration event lasts is determined.
  • a continuous 15-minute blood sugar trend is a blood sugar event, so preferably, m 2 ⁇ t ⁇ 15min.
  • the analyte is glucose in the blood
  • a certain period of time within three hours or less after a meal is One glycemic event, therefore preferably, m 2 ⁇ t ⁇ 180 min. .
  • the real-time trend of analyte concentration change is an important part of the information obtained by the user.
  • the real-time trend output without concentration change for a long time will reduce the user's experience and the monitoring effect, and the real-time trend output without concentration change for a long time will make itself Correct peak error judgment. Therefore, a reasonable choice of m2 can eliminate the real-time trend output without concentration changes to the greatest possible extent without affecting the user's experience.
  • m 2 can be obtained by machine learning, that is, using machine learning to obtain different m 2 through the sample size, and comparing the real-time trend of the analyte concentration change based on the m 2
  • the number of continuous cycles the less the number of continuous cycles, the more suitable m2 is.
  • the same m2 can correspond to different acquisition cycles that do not output the real-time trend of analyte concentration changes.
  • the retrospective method it is possible to traverse all the sets A that do not output the real-time trend of the analyte concentration change within a period of data under the artificially set threshold.
  • methods such as supervised machine learning or K-means clustering algorithm, the results that do not output real-time trends of analyte concentration changes are classified as set B.
  • the optimal solution of m2 can be obtained by comparing A and B.
  • the difference between P 1 and P 3 (P 1 -P 3 ) or the ratio of P 1 and P 3 (P 1 /P 3 ) exceeds the set threshold interval can be obtained from medical common sense, that is, through Medical research and experience determine the theoretical maximum value per unit time for the change in analyte concentration.
  • a reasonable threshold interval of P 1 /P 3 is ⁇ 10.
  • the method for improving the accuracy of the real-time trend of concentration changes disclosed in this application does not need to obtain a large amount of previous analyte concentration data for modeling like a retrospective analysis method, but only needs to call a small amount of previous analyte concentration data , which greatly simplifies the workload and complexity, and improves the accuracy of the real-time trend of concentration changes.
  • FIG. 1 is a schematic diagram of analyte concentration data and concentration change arrow trend output.
  • Figure 2 is a schematic diagram of the trend output with analyte concentration data but without concentration change arrows.
  • Fig. 3 is a real-time trend summary diagram obtained by simulating blood glucose concentration changes based on the glucose content data (a certain period of time) in the blood of normal healthy people by using the change rate dD 1 and the change amount D 2 judgment method.
  • Fig. 4 is a real-time trend summary diagram obtained by simulating blood glucose concentration changes based on the glucose content data (a certain period of time) in normal healthy human blood using the scheme disclosed in the present application.
  • Fig. 5 is a schematic flowchart of a method for improving the accuracy of the real-time trend of concentration changes in the process of continuous monitoring of animal body analyte concentrations disclosed in the present application.
  • Fig. 6 is another real-time trend summary diagram obtained by simulating blood glucose concentration changes based on the glucose content data (a certain period of time) in normal healthy human blood using the scheme disclosed in the present application.
  • FIG. 7 is another real-time trend summary diagram obtained by simulating blood glucose concentration changes based on the glucose content data (a certain period of time) in normal healthy human blood using the scheme disclosed in the present application.
  • the cited drawings are based on the blood sugar concentration data (a certain period of time) of normal healthy people, simulating the scheme disclosed in the application and outputting a real-time trend summary of the real-time trend of concentration changes picture.
  • the abscissa is the collection period, which can be understood as the signal collection at that time and output the concentration data, and the ordinate is the concentration of the degree of analysis (blood sugar), where the interval between the signal collection period and the output concentration data is 3 minutes.
  • the blood glucose concentration data is the data within a certain period of time within three hours after the meal
  • the current collection time is the 1796th collection cycle time
  • the blood sugar concentration at this time has slightly increased .
  • the sign of P2 is negative. If the signs of P 1 and P 2 are different, it is determined that the blood glucose concentration at the collection time (the 1796th collection cycle time) is suspicious, and only the concentration data is output, and the real-time trend of the analyte concentration change is not output.
  • the blood glucose concentration at this time has slightly increased compared with the 1796th and 1795th collection cycle time points.
  • the sign of P2 is negative. If the signs of P 1 and P 2 are different, it is determined that the blood glucose concentration at the collection time (the 1797th collection cycle time) is suspicious, and only the concentration data is output, and the real-time trend of the analyte concentration change is not output.
  • the blood glucose concentration still rises slightly compared with the 1796th collection period at this time.
  • the sign of P2 is negative. If the signs of P 1 and P 2 are different, it is determined that the blood glucose concentration at the collection time (the 1797th collection cycle time) is suspicious, and only the concentration data is output, and the real-time trend of the analyte concentration change is not output.
  • the obtained P 1 and P 2 have the same sign and are both negative.
  • the blood glucose concentration it is preferable to compare the proportional relationship between P 1 and P 3 , and if P 1 /P 3 >10, it is determined that the blood glucose concentration at the collection time (273th collection cycle number) is suspicious , only the concentration data is output, and the real-time trend of the analyte concentration change is not output.
  • the threshold value of the difference between P 1 and P 3 or the threshold value of the ratio of P 1 to P 3 can be obtained from medical common sense, that is, the theoretical maximum value per unit time of analyte concentration change determined through medical research and experience. Specifically, for the blood glucose concentration, if the ratio between P 1 and P 3 exceeds 10, the blood glucose concentration at the collection moment is considered to be suspicious.
  • This kind of judging method is mainly aimed at data abnormality in the case of short circuit (small data) or open circuit (large data) of signal sensor equipment.

Abstract

Disclosed in the present application is a method for improving the accuracy of a real-time concentration change trend during continuous monitoring of the analyte concentration in an animal body. The method comprises: performing, by means of polynomial fitting, straight line fitting on analyte concentration data C within different acquisition periods before the current acquisition moment, so as to obtain a first-order linear fitting equation C(t) = P·t + X (P being the slope of a fitted straight line, and X being a constant); and finally, by comparing symbols, differences or ratios of P within different acquisition periods, not outputting a real-time analyte concentration change trend, thereby reducing the possibility of introducing a wrong therapeutic plan. By means of the present application, compared with existing retrospective analysis methods, there is no need to acquire a large amount of previous analyte concentration data before modeling, and only a small amount of previous analyte concentration data needs to be called, thereby greatly reducing the workload and complexity, and improving the accuracy of a real-time concentration change trend.

Description

一种提高动物体分析物浓度连续监测过程中浓度变化实时趋势准确率的方法A method for improving the accuracy of the real-time trend of concentration changes in the process of continuous monitoring of animal analyte concentrations 技术领域technical field
本申请涉及动物体分析物浓度连续监测过程中，分析物浓度数据的变化、以及分析物浓度变化实时趋势的研究领域，特别是涉及判断分析物浓度数据是否为异常或存疑、判断是否输出分析物浓度变化实时趋势，以提高分析物浓度变化实时趋势准确率的方法。This application relates to the research field of the change of analyte concentration data and the real-time trend of analyte concentration change in the continuous monitoring process of animal body analyte concentration, especially involves judging whether the analyte concentration data is abnormal or suspicious, and judging whether to output analyte The real-time trend of concentration change is a method for improving the accuracy of the real-time trend of analyte concentration change.
背景技术Background technique
连续性分析物浓度监控系统(具有一定采集周期Δt的采集时刻t输出与采集时刻t对应的分析物浓度数据C(t))的出现，给慢性病患者提供了更好地了解该分析物浓度的变化水平，进而实现更好地控制该分析物的浓度。在临床上，对分析物浓度数据的分析(不管是实时的还是回顾性的)，对于慢性病的管理而言是非常有用的。The emergence of a continuous analyte concentration monitoring system (with a certain collection period Δt at the collection time t outputting the analyte concentration data C(t) corresponding to the collection time t) provides patients with chronic diseases with a better understanding of the analyte concentration. varying levels, thereby achieving better control over the concentration of that analyte. Clinically, the analysis of analyte concentration data, whether real-time or retrospective, is very useful for the management of chronic diseases.
上述分析物包括但不限于：乙酰胆碱、淀粉酶、胆红素、胆固醇、绒毛膜促性腺激素、肌酸激酶、肌氨酸、肌酸酐、DNA、果糖胺、葡萄糖、谷氨酸盐、生长激素类、激素类、酮体、乳酸、过氧化物、前列腺特异性抗原、凝血素、RNA、促甲状腺激素和肌钙蛋白等。分析物还可包括药物，如抗生素(比如，庆大霉素、万古霉素等)、洋地黄毒苷、地高辛、滥用药物、茶碱和华法令阻凝剂等。分析物浓度数据通常情况下是通过分析物传感器获得，其中，分析物传感器可以是葡萄糖传感器、酮传感器或其他与分析物对应的传感器等。The aforementioned analytes include, but are not limited to: acetylcholine, amylase, bilirubin, cholesterol, chorionic gonadotropin, creatine kinase, sarcosine, creatinine, DNA, fructosamine, glucose, glutamate, growth hormone Classes, hormones, ketone bodies, lactic acid, peroxides, prostate-specific antigen, prothrombin, RNA, thyroid-stimulating hormone, and troponin, etc. Analytes can also include drugs such as antibiotics (eg, gentamicin, vancomycin, etc.), digitoxin, digoxin, drugs of abuse, theophylline, and warfarin, among others. The analyte concentration data is usually obtained by an analyte sensor, wherein the analyte sensor may be a glucose sensor, a ketone sensor or other sensors corresponding to the analyte.
通常情况下，为便于慢性病患者更好地进行自我管理，连续性分析物浓度监控系统会按一定(采集)周期输出分析物浓度数据及浓度变化趋势(如图1所示，其中浓度变化趋势采用箭头趋势的表现形式)。通过分析物传感器获得较为准确的分析物浓度数据是相对容易的，而要获得相对准确的分析物浓度变化趋势，尤其是分析物浓度变化实时趋势(指在当前采集时刻给出分析物浓度数据的同时，分析物浓度正在发生的变化状态)则变得相对困难。Usually, in order to facilitate the self-management of patients with chronic diseases, the continuous analyte concentration monitoring system will output the analyte concentration data and concentration change trend according to a certain (collection) period (as shown in Figure 1, where the concentration change trend adopts representation of an arrow trend). It is relatively easy to obtain more accurate analyte concentration data through the analyte sensor, but to obtain relatively accurate analyte concentration change trends, especially the real-time trend of analyte concentration changes (referring to the current acquisition time given the analyte concentration data At the same time, the changing state of the analyte concentration) becomes relatively difficult.
针对分析物浓度变化实时趋势的判断，可计算以当前采集时刻T为终点的1个采集周期内的分析物浓度数据变化量D 1(D 1＝C(T)-C(T-Δt)，其中当前采集时 刻为T，相应时刻输出的分析物浓度为C(T)，下同)、变化率dD 1(dD 1＝D 1/Δt)和以当前采集时刻T为终点的n个采集周期内的分析物浓度数据变化量D 2(D 2＝C(T)-C(T-nΔt)，n为大于1的正整数)，通过D 1与D 2的符号和大小可以得到浓度变化方向和变化量，根据dD 1和D 2所处的阈值区间分类得到浓度变化实时箭头趋势。 For the judgment of the real-time trend of the change of the analyte concentration, the change amount D 1 of the analyte concentration data within a collection cycle with the current collection time T as the end point can be calculated (D 1 =C(T)-C(T-Δt), Where the current collection time is T, the analyte concentration output at the corresponding time is C(T), the same below), the rate of change dD 1 (dD 1 =D 1 /Δt), and n collection cycles ending at the current collection time T The amount of change in the analyte concentration data D 2 (D 2 = C(T)-C(T-nΔt), n is a positive integer greater than 1), the direction of concentration change can be obtained by the sign and size of D 1 and D 2 and the amount of change, according to the threshold interval classification of dD 1 and D 2 , the real-time arrow trend of concentration change is obtained.
以Δt＝3min、当前采集时刻是12:00为例，以当前采集时刻T为终点的1个采集周期指的是11:57～12:00，其包含2个分析物浓度数据；以当前采集时刻T为终点的3个采集周期指的是11:51～12:00，其包含4个分析物浓度数据。Taking Δt=3min and the current collection time is 12:00 as an example, one collection period with the current collection time T as the end point refers to 11:57～12:00, which contains 2 analyte concentration data; The three acquisition periods with time T as the end point refer to 11:51-12:00, which include 4 analyte concentration data.
关于D 2中n个采集周期(nΔt)的选择，则是根据不同的分析物种类、分析物浓度变化事件、医学知识或生理特征的特点来决定其最长时间。 Regarding the selection of n acquisition periods (nΔt) in D 2 , the longest time is determined according to different analyte types, analyte concentration change events, medical knowledge or characteristics of physiological characteristics.
例如，以分析血糖浓度为例，在无人为干预(无饮食摄入、用药、注射胰岛素或运动等)情况下，一般认为连续15分钟持续的血糖趋势为一个血糖事件；在有饮食摄入的情况下，一般认为餐后三小时或以内的某一时间段为一个血糖事件。依据上述方法，以15分钟(即nΔt＝15min)为例可以计算变化量D 2，并根据dD 1和D 2所对应的阈值区间(表1)给出浓度变化实时箭头趋势。 For example, taking the analysis of blood sugar concentration as an example, in the absence of human intervention (no food intake, medication, insulin injection or exercise, etc.), it is generally considered that a continuous 15-minute blood sugar trend is a blood sugar event; Under normal circumstances, it is generally considered that a certain period of time within three hours or less after a meal is a blood glucose event. According to the above method, taking 15 minutes (nΔt=15min) as an example, the change D 2 can be calculated, and the real-time arrow trend of the concentration change is given according to the threshold interval corresponding to dD 1 and D 2 (Table 1).
表1 血糖浓度变化实时箭头趋势判断表Table 1 Judgment table of real-time arrow trend of blood glucose concentration change
Figure PCTCN2021126936-appb-000001
Figure PCTCN2021126936-appb-000001
箭头向上的角度及数量的多少表示分析物浓度上升的速率，向上角度越大、数量越多则表示上升的速率越快；采用箭头向下的角度及数量的多少表示分析物浓度下降的速率，向下角度越大、数量越多则表示下降的速率越快；采用水平的箭头表示分析物浓度处于平稳的趋势。在无法计算出血糖的变化的速度和方向时则不输出实时趋势。The upward angle and quantity of the arrow represent the rate of increase in the concentration of the analyte. The larger the upward angle and the greater the quantity, the faster the rate of increase; the downward angle and quantity of the arrow represent the rate of decrease in the concentration of the analyte. The larger the downward angle and the larger the number, the faster the rate of decline; the horizontal arrow indicates that the analyte concentration is in a stable trend. When the speed and direction of blood sugar changes cannot be calculated, no real-time trend is output.
上述方法存在一个问题，比如，在某一个持续稳定的趋势中上述分析物浓度变化实时趋势的判断方法是没问题的；但是，当在某一个持续稳定的趋势中分析物浓度数据突然出现异常(如，分析物浓度数据由上升变为下降或下降变 为上升、分析物传感器异常、个体代谢功能的差异、人为干预等导致的浓度数据异常上升或下降)而导致该时刻分析物浓度数据存疑(下文也称该数据为不可信任数据或不可信任点)时，上述分析物浓度变化实时趋势的判断方法则无法准确输出分析物浓度变化实时趋势(依据上述方法，可以计算出血糖变化的速度和方向，可以输出分析物浓度变化实时趋势，但是准确率存在问题)，甚至基于上述阈值区间进行归类而导致错误的治疗方案。There is a problem in the above method, for example, in a certain continuous and stable trend, the method of judging the real-time trend of the above-mentioned analyte concentration change is no problem; however, when the analyte concentration data suddenly appears abnormal in a certain continuous and stable trend ( For example, the analyte concentration data changes from rising to falling or falling to rising, the abnormality of the analyte sensor, the difference in individual metabolic function, the abnormal rise or fall of the concentration data caused by human intervention, etc.), which makes the analyte concentration data doubtful at this moment ( When the data is also referred to as untrustworthy data or untrustworthy points hereinafter), the above method for judging the real-time trend of analyte concentration changes cannot accurately output the real-time trend of analyte concentration changes (according to the above method, the speed and direction of blood sugar changes can be calculated , can output the real-time trend of analyte concentration changes, but the accuracy rate is problematic), and even classify based on the above threshold intervals, resulting in wrong treatment plans.
此外，还存在dD 1和D 2处于不同阈值区间的情况，这同样会导致无法准确输出分析物浓度变化实时趋势。 In addition, there are situations where dD 1 and D 2 are in different threshold intervals, which also leads to the inability to accurately output the real-time trend of analyte concentration changes.
如图3实时趋势汇总图所示，以正常健康人体血液中的葡萄糖含量数据为例，采用上述判断方法来模拟血糖浓度变化实时趋势，可见在一个持续稳定的趋势中(如采集周期在104～112之间)，每个采集时刻均给出了缓慢上升或上升的实时趋势，这些趋势也是正确的。As shown in the real-time trend summary diagram in Figure 3, taking the glucose content data in normal healthy human blood as an example, using the above-mentioned judgment method to simulate the real-time trend of blood glucose concentration changes, it can be seen that in a continuous and stable trend (for example, the collection period is between 104~ 112), each acquisition moment gives a real-time trend of slowly rising or rising, and these trends are also correct.
而在采集周期为103时，血糖浓度突然下降，按照上述判断方法则给出了此刻血糖浓度正在缓慢下降，因该浓度接近阈值区间边缘，可能会误导患者采取措施以避免血糖浓度继续下降；或者在前两个采集时刻(采集周期为102和103时)给出血糖浓度正在上升和下降，在此刻(采集周期为104时)血糖浓度正在上升，也就是说在连续的三个采集时刻(采集周期为102～104时)分别给出了血糖浓度的实时趋势为上升、下降和上升，这将导致患者的使用体验感不佳。因此，上述分析物浓度变化实时趋势判断方法的准确性还有待提高，并希望获得如图4所示的实时趋势汇总图。However, when the collection period is 103, the blood glucose concentration drops suddenly. According to the above judgment method, the blood glucose concentration is slowly declining at this moment. Because the concentration is close to the edge of the threshold interval, it may mislead the patient to take measures to avoid the blood glucose concentration from continuing to drop; or At the first two acquisition moments (when the acquisition period is 102 and 103), the blood glucose concentration is rising and falling, and at this moment (when the acquisition period is 104), the blood glucose concentration is rising, that is to say, at three consecutive acquisition moments (acquisition The period is 102 to 104 hours) respectively give the real-time trends of blood glucose concentration as rising, falling and rising, which will lead to poor user experience for patients. Therefore, the accuracy of the above method for judging the real-time trend of analyte concentration changes needs to be improved, and it is hoped to obtain a real-time trend summary chart as shown in FIG. 4 .
需要指出的是，上述背景技术内容部分仅代表申请人对相关技术的理解，并不构成现有技术。It should be pointed out that the content of the above background technology only represents the applicant's understanding of related technologies, and does not constitute prior art.
发明内容Contents of the invention
针对上述问题，需要一种能够提供分析物浓度变化实时趋势的方法，尤其是提高分析物浓度变化实时趋势准确率的方法，在必要的时候仅有分析物浓度数据但无浓度变化实时箭头趋势输出(如图2所示)，同时希望这种无浓度变化实时箭头趋势输出情况越少越好，并且能够给予慢性病患者正确的治疗指导。In view of the above problems, there is a need for a method that can provide a real-time trend of analyte concentration changes, especially a method that improves the accuracy of the real-time trend of analyte concentration changes. When necessary, there is only analyte concentration data but no real-time arrow trend output of concentration changes (As shown in Figure 2), and at the same time, it is hoped that the less the real-time arrow trend output without concentration changes, the better, and the correct treatment guidance can be given to patients with chronic diseases.
更具体的，通过本申请公开的技术方案，可以识别出通过分析物浓度算法获得的分析物浓度数据是否异常或存疑、判断是否输出分析物浓度变化实时趋 势。More specifically, through the technical solution disclosed in this application, it is possible to identify whether the analyte concentration data obtained by the analyte concentration algorithm is abnormal or suspicious, and judge whether to output the real-time trend of the analyte concentration change.
本申请引入了基于最小二乘法多项式(如式1)拟合的方法对以当前采集时刻T为终点分析物浓度数据C(t)进行拟合。The present application introduces a method based on least square polynomial fitting (such as Equation 1) to fit the analyte concentration data C(t) with the current collection time T as the end point.
C(t)＝Pt a+P′t a-1+…+P″¨t 0 t∈(T-mΔt，T)，式1。 C(t)=Pt a +P′t a-1 +…+P″¨t 0 t∈(T-mΔt, T), Formula 1.
对于变化的动物体分析物浓度，过去的长时间数据(如72h，24h，12h，6h等)对分析其短期变化趋势没有实际意义，且从医学角度考虑在一定短的时间内分析物浓度变化不会出现剧烈的波动，因此进行拟合时涉及到的分析物浓度数据的数量有限，为简化运算模型，式1中取a＝1，得到拟合直线的一阶线性方程：For the changing animal body analyte concentration, the past long-term data (such as 72h, 24h, 12h, 6h, etc.) has no practical significance for analyzing the short-term change trend, and from the medical point of view, the change of analyte concentration in a certain short period of time is considered There will not be drastic fluctuations, so the number of analyte concentration data involved in the fitting is limited. In order to simplify the calculation model, a=1 is taken in formula 1 to obtain the first-order linear equation of the fitting line:
C(t)＝P·t+X(t∈(T-mΔt,T)，P为拟合直线的斜率，X为常量)。C(t)=P·t+X(t∈(T-mΔt,T), P is the slope of the fitted line, X is a constant).
上述一阶线性方程即为时间t与分析物浓度C(t)关系的线性方程，使得在采集时刻t时，经拟合计算的分析物浓度数据与连续性分析物浓度监控系统输出的分析物浓度数据之间的累积差值满足最小。The above-mentioned first-order linear equation is the linear equation of the relationship between time t and analyte concentration C(t), so that at the acquisition time t, the analyte concentration data calculated by fitting and the analyte concentration output by the continuous analyte concentration monitoring system The cumulative difference between the concentration data satisfies the minimum.
更具体的，采用多项式拟合：对以当前采集时刻T为终点的m 1个采集周期内的分析物浓度数据C进行直线拟合，得到一阶线性拟合方程C(t 1)＝P 1·t 1+X 1(t 1∈(T-m 1Δt,T)，1≤m 1，P 1为拟合直线的斜率，X 1为常量)。 More specifically, polynomial fitting is used: a straight line fitting is performed on the analyte concentration data C within m 1 acquisition periods with the current acquisition time T as the end point, and the first-order linear fitting equation C(t 1 )=P 1 is obtained ·t 1 +X 1 (t 1 ∈(Tm 1 Δt,T), 1≤m 1 , P 1 is the slope of the fitting line, and X 1 is a constant).
更具体的，采用多项式拟合：对以当前采集时刻T为终点的m 2个采集周期内的分析物浓度数据C进行直线拟合，得到一阶线性拟合方程C(t 2)＝P 2·t 2+X 2(t 2∈(T-m 2Δt,T)，m 1<m 2，P 2为拟合直线的斜率，X 2为常量)。 More specifically, polynomial fitting is used: a straight line fitting is performed on the analyte concentration data C within m 2 collection periods with the current collection time T as the end point, and the first-order linear fitting equation C(t 2 )=P 2 is obtained ·t 2 +X 2 (t 2 ∈(Tm 2 Δt,T), m 1 <m 2 , P 2 is the slope of the fitting line, and X 2 is a constant).
根据上述思路可知，P的符号正负则对应分析物浓度是上升还是下降，P的绝对值大小则对应分析物浓度变化速率的快和慢。According to the above ideas, the positive and negative sign of P corresponds to whether the concentration of the analyte is rising or falling, and the absolute value of P corresponds to the fast or slow rate of change of the concentration of the analyte.
作为本申请的一个方面，通过比较P 1和P 2的符号，即可快速判断是否输出分析物浓度变化实时趋势。具体的，若符号不一致(其中任意一个为0也视为符号不一致)，则说明在采集时刻T获得的分析物浓度数据异常，或至少是存疑的，为不可信任数据，判断为不输出分析物浓度变化实时趋势。 As an aspect of the present application, by comparing the signs of P1 and P2, it can be quickly judged whether to output the real - time trend of the change of the analyte concentration. Specifically, if the signs are inconsistent (any one of which is 0 is also considered as inconsistent signs), it means that the analyte concentration data obtained at the collection time T is abnormal, or at least doubtful, and is untrustworthy data, and it is judged that the analyte is not output Real-time trend of concentration changes.
更具体的，假设P 2符号为正，则表示在过去较长的一段时间段内(T-m 2Δt至T所持续的时间)整体处于上升的趋势；而如果此刻P 1符号为负或者等于0，则表示在过去较短的一段时间段内(T-m 1Δt至T所持续的时间)处于突然的下降或突然的平稳趋势。在这种情况下，由于未知未来采集时刻(T+Δt)的分析物浓度数据情况，无法确定当前数据是属于上升变为下降的转折点，还是整体 上升阶段的突变点。因此在采集时刻T不输出分析物浓度变化实时趋势。 More specifically, assuming that the sign of P 2 is positive, it means that the overall trend has been on the rise for a long period of time in the past (the duration of Tm 2 Δt to T); and if the sign of P 1 is negative or equal to 0 at this moment , it means that in a short period of time in the past (the duration from Tm 1 Δt to T), it is in a sudden decline or a sudden steady trend. In this case, since the analyte concentration data at the future collection time (T+Δt) is unknown, it is impossible to determine whether the current data belongs to the turning point from rising to falling, or the abrupt point of the overall rising stage. Therefore, no real-time trend of analyte concentration change is output at the acquisition time T.
作为本申请的另一个方面，若P 1和P 2符号一致，则进一步采用如下方法判断是否输出分析物浓度变化实时趋势。 As another aspect of the present application, if the signs of P 1 and P 2 are consistent, the following method is further used to judge whether to output the real-time trend of the change of the analyte concentration.
采用多项式拟合：对以采集时刻T-m xΔt为终点的m 3个采集周期内的分析物浓度数据C进行直线拟合，得到一阶线性拟合方程C(t 3)＝P 3·t 3+X 3(t 3∈(T-(m 3+m x)Δt,T-m xΔt)，1≤m x，m 1<m 3，P 3为拟合直线的斜率，X 3为常量)。 Using polynomial fitting: linear fitting is performed on the analyte concentration data C within m 3 acquisition periods with the acquisition time Tm x Δt as the end point, and the first-order linear fitting equation C(t 3 )=P 3 ·t 3 is obtained +X 3 (t 3 ∈(T-(m 3 +m x )Δt,Tm x Δt), 1≤m x , m 1 <m 3 , P 3 is the slope of the fitted line, X 3 is a constant).
再比较P 1与P 3的大小关系或比例关系，若P 1与P 3差值(P 1-P 3)或P 1与P 3比值(P 1/P 3)处于设定阈值区间之外，则不输出分析物浓度变化实时趋势。 Then compare the size relationship or proportional relationship between P 1 and P 3 , if the difference between P 1 and P 3 (P 1 -P 3 ) or the ratio between P 1 and P 3 (P 1 /P 3 ) is outside the set threshold range , the real-time trend of analyte concentration change will not be output.
图5给出了上述提高动物体分析物浓度连续监测过程中浓度变化实时趋势准确率方法的流程示意图。FIG. 5 shows a schematic flow chart of the method for improving the accuracy of the real-time trend of the concentration change in the continuous monitoring process of the concentration of the analyte in the animal body.
更具体的，P 3表示除去当前时刻T之前m x个采集周期内分析物浓度数据(即排除包含当前数据的部分数据)的累积趋势。如果当前数据点异常很大(即P 1的绝对值较大)完全不具备临床意义的情况时，可以采用前一段(T-(m 3+m x)Δt至T-m xΔt)m 3个采集周期内具有较高信任度的分析物浓度数据的累积变化趋势分析过去时间段内的正常趋势。另外，由于动物体内分析物浓度具有一定的连续性，前一段m 3个采集周期内数据的累积趋势可以一定程度上反映当前的趋势。 More specifically, P 3 represents the accumulation trend of the analyte concentration data in the m x collection periods before the current time T (that is, excluding the partial data including the current data). If the current data point is abnormally large (that is, the absolute value of P 1 is large) and does not have clinical significance at all, the previous period (T-(m 3 +m x )Δt to Tm x Δt)m 3 acquisitions can be used Cumulative trending of analyte concentration data with high confidence within a period analyzes normal trends over past time periods. In addition, because the concentration of analytes in animals has a certain continuity, the cumulative trend of data in the previous period of m3 collection cycles can reflect the current trend to a certain extent.
因P 1和P 2符号一致，由于采样周期较短，P 1和P 3的符号也一致，比较P 1与P 3的大小关系或比例关系：若P 1与P 3差值(P 1-P 3)或P 1与P 3比值(P 1/P 3)处于设定阈值区间之外，则说明在当前采集时刻T获得的分析物浓度数据存疑，为不可信任数据，判断为不输出分析物浓度变化实时趋势。 Because the signs of P 1 and P 2 are consistent, and the signs of P 1 and P 3 are also consistent due to the short sampling period, compare the size relationship or proportional relationship between P 1 and P 3 : If the difference between P 1 and P 3 (P 1 - P 3 ) or the ratio of P 1 to P 3 (P 1 /P 3 ) is outside the set threshold range, indicating that the analyte concentration data obtained at the current collection time T is suspicious, unreliable data, and judged not to output analysis Real-time trend of concentration changes.
作为本申请的优选方案，为使得一阶线性拟合方程C(t 1)＝P 1·t 1+X 1中P 1更具有代表性，优选地m 1≤3，更优选地m 1＝1，可直接反映在临近当前时刻的上一个周期内分析物浓度变化趋势。 As a preferred solution of this application, in order to make P 1 in the first-order linear fitting equation C(t 1 )=P 1 ·t 1 +X 1 more representative, preferably m 1 ≤ 3, more preferably m 1 = 1. It can directly reflect the change trend of the analyte concentration in the previous period close to the current moment.
作为本申请的另一优选方案，为使得一阶线性拟合方程C(t 2)＝P 2·t 2+X 2更具有代表性，优选地，m 2依据医学研究与经验判断的分析物浓度事件所持续的最长时间确定。 As another preferred solution of the present application, in order to make the first-order linear fitting equation C(t 2 )=P 2 ·t 2 +X 2 more representative, preferably, m 2 is based on the analytes judged by medical research and experience The maximum time for which the concentration event lasts is determined.
作为本申请的另一优选方案，分析物为血液中的葡萄糖时，根据目前的医学研究与经验判断，在无人为干预的情况下，连续15min持续的血糖趋势为一个血糖事件，因此优选地，m 2Δt≤15min。 As another preferred solution of the present application, when the analyte is glucose in the blood, according to current medical research and experience, without human intervention, a continuous 15-minute blood sugar trend is a blood sugar event, so preferably, m 2 Δt≤15min.
作为本申请的另一优选方案，分析物为血液中的葡萄糖时，根据目前的医 学研究与经验判断，在有饮食摄入的情况下，一般认为餐后三小时或以内的某一时间段为一个血糖事件，因此优选地，m 2Δt≤180min。。 As another preferred solution of the present application, when the analyte is glucose in the blood, according to current medical research and empirical judgment, in the case of dietary intake, it is generally believed that a certain period of time within three hours or less after a meal is One glycemic event, therefore preferably, m 2 Δt≦180 min. .
分析物浓度变化实时趋势作为用户得到的信息的重要组成部分，长时间的无浓度变化实时趋势输出会降低用户的体验感进而降低监测的效果，且长时间的无浓度变化实时趋势输出会将本身正确的峰值错误判断。因此合理地选择m 2可以最大可能的消除无浓度变化实时趋势输出，且不影响用户的体验感。 The real-time trend of analyte concentration change is an important part of the information obtained by the user. The real-time trend output without concentration change for a long time will reduce the user's experience and the monitoring effect, and the real-time trend output without concentration change for a long time will make itself Correct peak error judgment. Therefore, a reasonable choice of m2 can eliminate the real-time trend output without concentration changes to the greatest possible extent without affecting the user's experience.
作为本申请的另一优选方案，m 2可通过机器学习的方法来获得，即利用机器学习的方法通过样本量来获得不同m 2，比较基于该m 2判定为不输出分析物浓度变化实时趋势所持续周期的数量，持续周期数越少，则m 2越合适。对于不同的分析物浓度数据(不同的变化速率和异常点前后的数据趋势不同)，相同的m 2可以对应不同的不输出分析物浓度变化实时趋势所持续的采集周期数。通过回顾式的方法，可以遍历人为设定的阈值下的一段数据内的所有确定不输出分析物浓度变化实时趋势的集合A。通过有监督的机器学习或K均值聚类算法等方法，将不输出分析物浓度变化实时趋势的结果归类为集合B。通过A与B的比较可以得到m 2的最优解。 As another preferred solution of the present application, m 2 can be obtained by machine learning, that is, using machine learning to obtain different m 2 through the sample size, and comparing the real-time trend of the analyte concentration change based on the m 2 The number of continuous cycles, the less the number of continuous cycles, the more suitable m2 is. For different analyte concentration data (different change rates and different data trends before and after the abnormal point), the same m2 can correspond to different acquisition cycles that do not output the real-time trend of analyte concentration changes. Through the retrospective method, it is possible to traverse all the sets A that do not output the real-time trend of the analyte concentration change within a period of data under the artificially set threshold. Through methods such as supervised machine learning or K-means clustering algorithm, the results that do not output real-time trends of analyte concentration changes are classified as set B. The optimal solution of m2 can be obtained by comparing A and B.
作为本申请的另一优选方案，排除分析物浓度数据应尽可能的少，从而尽可能地保留分析物浓度变化的特征，优选地，m x≤3；更优选地m x＝1，即只排除当前时刻的浓度数据。 As another preferred solution of the present application, the exclusion of analyte concentration data should be as little as possible, thereby retaining the characteristics of analyte concentration changes as much as possible, preferably, m x ≤ 3; more preferably m x = 1, that is, only Concentration data at the current moment are excluded.
作为本申请的另一优选方案，P 1与P 3差值(P 1-P 3)或P 1与P 3比值(P 1/P 3)超出设定阈值区间可以由医学常识得到，即通过医学研究与经验确定分析物浓度变化的单位时间的理论最大值。优选地，针对血糖浓度，合理的P 1/P 3的阈值区间为≤10。 As another preferred solution of this application, the difference between P 1 and P 3 (P 1 -P 3 ) or the ratio of P 1 and P 3 (P 1 /P 3 ) exceeds the set threshold interval can be obtained from medical common sense, that is, through Medical research and experience determine the theoretical maximum value per unit time for the change in analyte concentration. Preferably, for the blood glucose concentration, a reasonable threshold interval of P 1 /P 3 is ≤10.
与现有技术相比，本申请公开的方案具有如下优势。Compared with the prior art, the solution disclosed in this application has the following advantages.
(1)本申请所公开的提高浓度变化实时趋势准确率的方法，无需像回顾式的分析方法需要获取之前大量的分析物浓度数据后进行建模，而只需要调用之前少量的分析物浓度数据，大大简化了工作量和复杂度，提高了浓度变化实时趋势的准确率。(1) The method for improving the accuracy of the real-time trend of concentration changes disclosed in this application does not need to obtain a large amount of previous analyte concentration data for modeling like a retrospective analysis method, but only needs to call a small amount of previous analyte concentration data , which greatly simplifies the workload and complexity, and improves the accuracy of the real-time trend of concentration changes.
(2)本申请所公开的方法可以避免在出现不可信任数据时错误提供实时趋势，降低引入错误治疗方案的可能性。(2) The method disclosed in this application can avoid wrongly providing real-time trends when untrustworthy data appears, and reduce the possibility of introducing wrong treatment plans.
(3)本申请所公开的方法可适用于无预测功能的实时检测，具有更好地适 应性。(3) The method disclosed in this application can be applied to real-time detection without predictive function, and has better adaptability.
附图说明Description of drawings
图1是有分析物浓度数据及浓度变化箭头趋势输出的示意图。FIG. 1 is a schematic diagram of analyte concentration data and concentration change arrow trend output.
图2是有分析物浓度数据无浓度变化箭头趋势输出的示意图。Figure 2 is a schematic diagram of the trend output with analyte concentration data but without concentration change arrows.
图3是基于正常健康人体血液中的葡萄糖含量数据(某一时间段)，采用变化率dD 1和变化量D 2判断方法来模拟血糖浓度变化获得的实时趋势汇总图。 Fig. 3 is a real-time trend summary diagram obtained by simulating blood glucose concentration changes based on the glucose content data (a certain period of time) in the blood of normal healthy people by using the change rate dD 1 and the change amount D 2 judgment method.
图4是基于正常健康人体血液中的葡萄糖含量数据(某一时间段)，采用本申请公开的方案来模拟血糖浓度变化获得的实时趋势汇总图。Fig. 4 is a real-time trend summary diagram obtained by simulating blood glucose concentration changes based on the glucose content data (a certain period of time) in normal healthy human blood using the scheme disclosed in the present application.
图5是本申请公开的提高动物体分析物浓度连续监测过程中浓度变化实时趋势准确率的方法流程示意图。Fig. 5 is a schematic flowchart of a method for improving the accuracy of the real-time trend of concentration changes in the process of continuous monitoring of animal body analyte concentrations disclosed in the present application.
图6是基于正常健康人体血液中的葡萄糖含量数据(某一时间段)，采用本申请公开的方案来模拟血糖浓度变化获得的另一实时趋势汇总图。Fig. 6 is another real-time trend summary diagram obtained by simulating blood glucose concentration changes based on the glucose content data (a certain period of time) in normal healthy human blood using the scheme disclosed in the present application.
图7是基于正常健康人体血液中的葡萄糖含量数据(某一时间段)，采用本申请公开的方案来模拟血糖浓度变化获得的另一实时趋势汇总图。FIG. 7 is another real-time trend summary diagram obtained by simulating blood glucose concentration changes based on the glucose content data (a certain period of time) in normal healthy human blood using the scheme disclosed in the present application.
图8是基于正常健康人体血液中的葡萄糖含量数据(某一时间段)，采用本申请公开的方案(m 2＝5)来模拟血糖浓度变化获得的另一实时趋势汇总图。 Fig. 8 is another real-time trend summary diagram obtained by simulating changes in blood glucose concentration using the scheme (m 2 =5) disclosed in the present application based on glucose content data in normal healthy human blood (a certain period of time).
图9是基于正常健康人体血液中的葡萄糖含量数据(某一时间段)，采用本申请公开的方案(m 2＝10)来模拟血糖浓度变化获得的另一实时趋势汇总图。 Fig. 9 is another real-time trend summary graph obtained by simulating changes in blood glucose concentration using the scheme disclosed in the present application (m 2 =10) based on the glucose content data in normal healthy human blood (a certain period of time).
具体实施方式detailed description
为便于本领域技术人员更好地理解本申请的技术方案，以下结合附图及各种示例性实施例来对本申请的方案做进一步阐述。需要提醒的是，除非另有具体说明，否则这些实施例中阐述的方法不限制本申请的范围。In order to facilitate those skilled in the art to better understand the technical solution of the present application, the solution of the present application will be further described below in conjunction with the accompanying drawings and various exemplary embodiments. It should be reminded that unless otherwise specifically stated, the methods described in these examples do not limit the scope of the present application.
本申请的实施例中，除特殊说明外，所引用的附图均是基于正常健康人体的血糖浓度数据(某一时间段)、模拟本申请公开的方案并输出浓度变化实时趋势的实时趋势汇总图。横坐标为采集周期，可以理解为在当时时刻进行信号采集并输出了浓度数据，纵坐标为分析度(血糖)的浓度，其中信号采集周期/输出浓度数据的间隔为3min。In the embodiments of the present application, unless otherwise specified, the cited drawings are based on the blood sugar concentration data (a certain period of time) of normal healthy people, simulating the scheme disclosed in the application and outputting a real-time trend summary of the real-time trend of concentration changes picture. The abscissa is the collection period, which can be understood as the signal collection at that time and output the concentration data, and the ordinate is the concentration of the degree of analysis (blood sugar), where the interval between the signal collection period and the output concentration data is 3 minutes.
需要提醒的是，尽管我们现在可能可以从汇总图中判断出血糖浓度的变化在当时是一个怎样的趋势，但是在当时输出浓度变化实时趋势时，并不知道下 一个采集时刻血糖浓度是多少。What needs to be reminded is that although we may be able to judge the trend of the blood sugar concentration change at that time from the summary graph, we do not know the blood sugar concentration at the next collection time when the real-time trend of the concentration change is output at that time.
实施例1。Example 1.
从图6可知(血糖浓度数据为餐后三小时内某一段时间内数据)，假定当前采集时刻为第1796个采集周期时刻，与第1795个采集周期时刻相比此时血糖浓度有稍微地上升。根据第1795～1796采集周期时刻(m 1＝1)的血糖浓度拟合得到的P 1符号为正，根据第1791～1796共5个采集周期(m 2Δt＝15min)的血糖浓度拟合得到的P 2符号为负。P 1与P 2符号不同，则判定在采集时刻(第1796个采集周期时刻)的血糖浓度存疑，仅输出浓度数据，不输出分析物浓度变化实时趋势。 It can be seen from Figure 6 (the blood glucose concentration data is the data within a certain period of time within three hours after the meal), assuming that the current collection time is the 1796th collection cycle time, compared with the 1795th collection cycle time, the blood sugar concentration at this time has slightly increased . The sign of P 1 obtained by fitting the blood glucose concentration at the 1795th to 1796th collection period (m 1 = 1) is positive, and the P 1 is obtained by fitting the blood glucose concentration at the 1791st to 1796th five collection periods (m 2 Δt = 15min) The sign of P2 is negative. If the signs of P 1 and P 2 are different, it is determined that the blood glucose concentration at the collection time (the 1796th collection cycle time) is suspicious, and only the concentration data is output, and the real-time trend of the analyte concentration change is not output.
类似地，假定当前采集时刻为第1797个采集周期时刻，此时与第1796、1795个采集周期时刻相比血糖浓度均有稍微地上升。根据第1795～1797采集周期时刻(m 1＝2)的血糖浓度拟合得到的P 1符号为正，根据第1792～1797共5个采集周期(m 2Δt＝15min)的血糖浓度拟合得到的P 2符号为负。P 1与P 2符号不同，则判定在采集时刻(第1797个采集周期时刻)的血糖浓度存疑，仅输出浓度数据，不输出分析物浓度变化实时趋势。 Similarly, assuming that the current collection time is the 1797th collection cycle time, the blood glucose concentration at this time has slightly increased compared with the 1796th and 1795th collection cycle time points. The sign of P 1 obtained by fitting the blood glucose concentration at the 1795th to 1797th collection period (m 1 = 2) is positive, and it is obtained by fitting the blood glucose concentration of 5 collection periods (m 2 Δt = 15min) from the 1792nd to 1797th The sign of P2 is negative. If the signs of P 1 and P 2 are different, it is determined that the blood glucose concentration at the collection time (the 1797th collection cycle time) is suspicious, and only the concentration data is output, and the real-time trend of the analyte concentration change is not output.
类似地，假定当前采集时刻为第1797个采集周期时刻，此时与第1796个采集周期时刻相比血糖浓度仍有稍微地上升。根据第1796～1797采集周期时刻(m 1＝1)的血糖浓度拟合得到的P 1符号为正，根据第1792～1797共5个采集周期(m 2Δt＝15min)的血糖浓度拟合得到的P 2符号为负。P 1与P 2符号不同，则判定在采集时刻(第1797个采集周期时刻)的血糖浓度存疑，仅输出浓度数据，不输出分析物浓度变化实时趋势。 Similarly, assuming that the current collection time is the 1797th collection period, the blood glucose concentration still rises slightly compared with the 1796th collection period at this time. The sign of P 1 obtained by fitting the blood glucose concentration at the time of the 1796th to 1797th collection period (m 1 = 1) is positive, and it is obtained from the fitting of the blood glucose concentration at the 1792th to 1797th collection period of 5 collection periods (m 2 Δt = 15min) The sign of P2 is negative. If the signs of P 1 and P 2 are different, it is determined that the blood glucose concentration at the collection time (the 1797th collection cycle time) is suspicious, and only the concentration data is output, and the real-time trend of the analyte concentration change is not output.
类似地，P 1符号为负、P 2符号为正，或者两者中任意一个为0，判断方式是一样的，输出逻辑也是一样的。 Similarly, if the sign of P 1 is negative, the sign of P 2 is positive, or either of them is 0, the judgment method is the same, and the output logic is also the same.
实施例2。Example 2.
从图6可知(血糖浓度数据为餐后三小时内某一段时间内数据)，P 2可以根据第1784～1796共12个采集周期(m 2Δt＝36min)，或更长的采集周期(图中未示出)的血糖浓度拟合得到的，P 2符号为负。 It can be seen from Figure 6 (the blood glucose concentration data is the data within a certain period of time within three hours after a meal), P 2 can be based on 12 collection cycles (m 2 Δt=36min) from the 1784th to 1796th, or a longer collection cycle (Fig. not shown in ) , obtained by fitting the blood glucose concentration, and the sign of P2 is negative.
其方法和实施例1类似。Its method is similar to Example 1.
实施例3。Example 3.
图7中，人为地将第273和274个采集周期时刻的血糖浓度改为异常下降 的数据(数据很小)。In Fig. 7, the blood glucose concentration at the 273rd and 274th collection period is artificially changed to abnormally lower data (the data is very small).
假定当前采集时刻为第273个采集周期时刻，根据实施例1的方法(m 1＝1，m 2Δt＝15min)，获得的P 1和P 2符号相同，均为负。 Assuming that the current acquisition time is the 273rd acquisition cycle time, according to the method in Embodiment 1 (m 1 =1, m 2 Δt = 15min), the obtained P 1 and P 2 have the same sign and are both negative.
此时，则舍弃第273个采集周期时刻(m x＝1)的血糖浓度数据，采用第272个采集周期数的时刻的血糖浓度数据及前5个采集周期(即第267～272周期数，m 3Δt＝15min)的血糖浓度拟合得到的P 3符号为负。 At this time, the blood glucose concentration data at the 273rd acquisition cycle time (m x = 1) is discarded, and the blood glucose concentration data at the 272nd acquisition cycle time and the first 5 acquisition cycles (that is, the 267th to 272th cycle number, The sign of P 3 obtained by fitting the blood glucose concentration of m 3 Δt=15 min) is negative.
比较P 1和P 3符号，相同。则进一步比较P 1与P 3的大小关系或比例关系。 Comparing P 1 and P 3 symbols, same. Then further compare the size relationship or proportional relationship between P 1 and P 3 .
在本实施例中，针对的是血糖浓度，则优选比较P 1与P 3的比例关系，P 1/P 3>10，则判定在采集时刻(第273个采集周期数时)的血糖浓度存疑，仅输出浓度数据，不输出分析物浓度变化实时趋势。 In this embodiment, for the blood glucose concentration, it is preferable to compare the proportional relationship between P 1 and P 3 , and if P 1 /P 3 >10, it is determined that the blood glucose concentration at the collection time (273th collection cycle number) is suspicious , only the concentration data is output, and the real-time trend of the analyte concentration change is not output.
通常情况下，P 1与P 3差值的阈值或P 1与P 3比值的阈值可以由医学常识得到，即通过医学研究与经验确定分析物浓度变化的单位时间的理论最大值。具体地，针对血糖浓度，P 1和P 3两者的比值超出10，则认为该采集时刻的血糖浓度存疑。 Usually, the threshold value of the difference between P 1 and P 3 or the threshold value of the ratio of P 1 to P 3 can be obtained from medical common sense, that is, the theoretical maximum value per unit time of analyte concentration change determined through medical research and experience. Specifically, for the blood glucose concentration, if the ratio between P 1 and P 3 exceeds 10, the blood glucose concentration at the collection moment is considered to be suspicious.
类似地，舍弃第273和272个采集周期时刻(m x＝2)的血糖浓度数据，采用第271个采集周期数的时刻的血糖浓度数据及前5个采集周期(即第266～271周期数，m 3Δt＝15min)的血糖浓度拟合得到的P 3符号为负。 Similarly, discard the blood glucose concentration data at the 273rd and 272nd acquisition period (m x = 2), and use the blood glucose concentration data at the 271st acquisition period and the first 5 acquisition periods (that is, the 266th to 271st period , m 3 Δt=15min), the sign of P 3 obtained by fitting the blood glucose concentration is negative.
类似地，人为地将第273和274个采集周期数时刻的血糖浓度改为异常上升的数据(数据很大)，根据实施例1的方法，倘若获得的P 1符号为正、P 2符号为负，或P 1符号为负、P 2符号为正，则在采集时刻(第273个采集周期数时)不输出分析物浓度变化实时趋势(未示出图例)。 Similarly, artificially change the blood glucose concentration at the 273rd and 274th acquisition cycle times to abnormally rising data (the data is very large), according to the method in Example 1 , if the P1 sign obtained is positive, and the P2 sign is Negative, or the sign of P 1 is negative and the sign of P 2 is positive, then the real-time trend of the change of the analyte concentration will not be output at the acquisition time (at the 273th acquisition cycle number) (the legend is not shown).
此种判断方法主要针对于信号传感器设备短路(数据很小)或断路(数据很大)的情况下的数据异常。This kind of judging method is mainly aimed at data abnormality in the case of short circuit (small data) or open circuit (large data) of signal sensor equipment.
实施例4Example 4
如图8所示，当计算P 2采用5个采集周期(m 2＝5)的数据时，出现连续3个采集时刻(第1842～1844采集周期时刻)不输出实时趋势；如图9所示，当计算P 2采用10个采集周期(m 2＝10)的数据时，出现连续4个采集时刻(第1842～1845采集周期时刻)不输出实时趋势；可见，这两个m 2的取值均不太满足患者的使用要求。 As shown in Figure 8, when the data of 5 acquisition periods (m 2 = 5) are used to calculate P2, there are 3 consecutive acquisition moments (the 1842nd to 1844th acquisition period moments) and no real-time trend is output; as shown in Figure 9 , when the data of 10 acquisition periods (m 2 =10) are used to calculate P 2 , there are 4 consecutive acquisition moments (the 1842nd to 1845th acquisition period moments) and no real-time trend is output; it can be seen that the values of these two m 2 All do not quite satisfy the use requirement of the patient.
进一步地，通过回顾式的方法，可以遍历人为设定的阈值下的一段数据内的所有确定不输出分析物浓度变化实时趋势的集合A。通过有监督的机器学习 或K均值聚类算法等方法将不输出分析物浓度变化实时趋势的结果归类为集合B。通过A与B的比较可以得到m 2的最优解。 Further, through a retrospective method, it is possible to traverse all the sets A that do not output the real-time trend of the analyte concentration change within a period of data under the artificially set threshold. The results that do not output the real-time trend of analyte concentration changes are classified as set B by methods such as supervised machine learning or K-means clustering algorithm. The optimal solution of m2 can be obtained by comparing A and B.
尽管上述各实施例是针对血糖浓度的，应当理解，对于其他分析物，如动物体分泌物、代谢产物，或摄入药物等均可采用类似上述实施例的判断方法，所不同的是依据医学研究与经验判断的分析物浓度事件重新确定不同事件对应的m 2Δt最长时间。 Although the above-mentioned embodiments are aimed at blood sugar concentration, it should be understood that for other analytes, such as animal body secretions, metabolites, or ingested drugs, the judgment methods similar to the above-mentioned embodiments can be used, the difference is based on medical Analyte concentration events that were studied and empirically judged re-determined the m 2 Δt maximum time corresponding to different events.
上述实施例是示例性的，并非穷尽性的。在不偏离所说明的各实施例的范围和精神的情况下，对于本技术领域的普通技术人员来说许多修改和变更都是显而易见的。本文中所用术语的选择，旨在最好地解释各实施例的原理、实际应用或对市场中的技术改进，或者使本技术领域的其它普通技术人员能理解本文披露的各实施例。The above-described embodiments are exemplary and not exhaustive. Many modifications and alterations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principle of each embodiment, practical application or technical improvement in the market, or to enable other ordinary skilled in the art to understand each embodiment disclosed herein.

Claims (13)

  1. 一种提高动物体分析物浓度连续监测过程中浓度变化实时趋势准确率的方法，其为：在具有一定采集周期Δt的采集时刻t输出与采集时刻t对应的分析物浓度数据C(t)，在当前采集时刻T仅输出的分析物浓度C(T)，不输出分析物浓度变化实时趋势，包括如下步骤：A method for improving the real-time trend accuracy of concentration changes in the continuous monitoring process of animal body analyte concentration, which is: outputting the analyte concentration data C(t) corresponding to the collection time t at a collection time t with a certain collection period Δt, At the current acquisition time T, only the analyte concentration C(T) is output, and the real-time trend of the analyte concentration change is not output, including the following steps:
    采用多项式拟合：对以当前采集时刻T为终点的m 1个采集周期内的分析物浓度数据C进行直线拟合，得到一阶线性拟合方程C(t 1)＝P 1·t 1+X 1(t 1∈(T-m 1Δt,T)，1≤m 1，P 1为拟合直线的斜率，X 1为常量)； Using polynomial fitting: performing straight line fitting on the analyte concentration data C within m 1 collection periods with the current collection time T as the end point, and obtaining the first-order linear fitting equation C(t 1 )=P 1 ·t 1 + X 1 (t 1 ∈ (Tm 1 Δt,T), 1≤m 1 , P 1 is the slope of the fitted line, X 1 is a constant);
    采用多项式拟合：对以当前采集时刻T为终点的m 2个采集周期内的分析物浓度数据C进行直线拟合，得到一阶线性拟合方程C(t 2)＝P 2·t 2+X 2(t 2∈(T-m 2Δt,T)，m 1<m 2，P 2为拟合直线的斜率，X 2为常量)； Using polynomial fitting: performing linear fitting on the analyte concentration data C within m 2 collection periods with the current collection time T as the end point, and obtaining the first-order linear fitting equation C(t 2 )=P 2 ·t 2 + X 2 (t 2 ∈ (Tm 2 Δt,T), m 1 <m 2 , P 2 is the slope of the fitted line, X 2 is a constant);
    比较P 1和P 2的符号， Comparing the signs of P1 and P2,
    (1)符号不一致，则不输出分析物浓度变化实时趋势；或(1) If the signs are inconsistent, the real-time trend of the analyte concentration change will not be output; or
    (2)符号一致，则(2) The signs are consistent, then
    采用多项式拟合：对以采集时刻T-m xΔt为终点的m 3个采集周期内的分析物浓度数据C进行直线拟合，得到一阶线性拟合方程C(t 3)＝P 3·t 3+X 3(t 3∈(T-(m 3+m x)Δt，T-m xΔt)，1≤m x，m 1<m 3，P 3为拟合直线的斜率，X 3为常量)， Using polynomial fitting: linear fitting is performed on the analyte concentration data C within m 3 acquisition periods with the acquisition time Tm x Δt as the end point, and the first-order linear fitting equation C(t 3 )=P 3 ·t 3 is obtained +X 3 (t 3 ∈(T-(m 3 +m x )Δt, Tm x Δt), 1≤m x , m 1 <m 3 , P 3 is the slope of the fitted line, X 3 is a constant),
    比较P 1与P 3的大小关系或比例关系，若P 1-P 3或P 1/P 3处于设定阈值区间之外，则不输出分析物浓度变化实时趋势。 Comparing the size relationship or proportional relationship between P 1 and P 3 , if P 1 -P 3 or P 1 /P 3 is outside the set threshold range, the real-time trend of analyte concentration change will not be output.
  2. 根据权利要求1所述的方法，其中，所述分析物为动物体分泌物、代谢产物，或摄入药物中的任意一种。The method according to claim 1, wherein the analyte is any one of animal body secretions, metabolites, or ingested drugs.
  3. 根据权利要求1或2所述的方法，其中，所述分析物为血液中葡萄糖。The method according to claim 1 or 2, wherein the analyte is glucose in blood.
  4. 根据权利要求1所述的方法，其中，所述多项式为最小二乘法多项式。The method of claim 1, wherein the polynomial is a least squares polynomial.
  5. 根据权利要求1所述的方法，其中，所述m 1≤3。 The method according to claim 1, wherein said m 1 ≦3.
  6. 根据权利要求1所述的方法，其中，所述m 1＝1。 The method according to claim 1, wherein said m 1 =1.
  7. 根据权利要求1所述的方法，其中，所述m 2依据医学研究与经验判断的分析物浓度事件所持续的最长时间确定，使得m 2Δt≤分析物浓度事件所持续的最长时间。 The method according to claim 1, wherein the m 2 is determined according to the longest duration of the analyte concentration event judged by medical research and experience, so that m 2 Δt≤the longest duration of the analyte concentration event.
  8. 根据权利要求1所述的方法，其中，所述分析物为血液中葡萄糖时，m 2Δt≤15min。 The method according to claim 1, wherein when the analyte is glucose in blood, m 2 Δt≤15min.
  9. 根据权利要求1所述的方法，其中，所述m 2为通过机器学习方法来获得。 The method according to claim 1, wherein the m2 is obtained by a machine learning method.
  10. 根据权利要求9所述的方法，其中，所述机器学习方法为K均值聚类算法。The method according to claim 9, wherein the machine learning method is a K-means clustering algorithm.
  11. 根据权利要求1所述的方法，其中，所述m x≤3。 The method according to claim 1, wherein said m x ≤ 3.
  12. 根据权利要求1所述的方法，其中，所述m x＝1。 The method according to claim 1, wherein said m x =1.
  13. 根据权利要求1所述的方法，其中，所述设定阈值区间可通过医学研究与经验确定。The method according to claim 1, wherein the set threshold interval can be determined through medical research and experience.
PCT/CN2021/126936 2021-07-19 2021-10-28 Method for improving accuracy of real-time concentration change trend during continuous monitoring of analyte concentration in animal body WO2023000530A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110811537.6 2021-07-19
CN202110811537.6A CN113380411B (en) 2021-07-19 2021-07-19 Method for improving accuracy of concentration change real-time trend in continuous monitoring process of animal analyte concentration

Publications (1)

Publication Number Publication Date
WO2023000530A1 true WO2023000530A1 (en) 2023-01-26

Family

ID=77582349

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/126936 WO2023000530A1 (en) 2021-07-19 2021-10-28 Method for improving accuracy of real-time concentration change trend during continuous monitoring of analyte concentration in animal body

Country Status (2)

Country Link
CN (1) CN113380411B (en)
WO (1) WO2023000530A1 (en)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113380411B (en) * 2021-07-19 2024-03-01 苏州百孝医疗科技有限公司 Method for improving accuracy of concentration change real-time trend in continuous monitoring process of animal analyte concentration
CN113951879B (en) * 2021-12-21 2022-04-05 苏州百孝医疗科技有限公司 Blood glucose prediction method and device and system for monitoring blood glucose level
CN114166913B (en) * 2022-02-10 2022-05-27 苏州百孝医疗科技有限公司 Automatic calibration method and device, system for monitoring analyte concentration level
CN114145738B (en) * 2022-02-10 2022-06-24 苏州百孝医疗科技有限公司 Analyte concentration data generation method and device, and system for monitoring analyte level

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110077494A1 (en) * 2009-09-29 2011-03-31 Abbott Diabetes Care Inc. Method and Apparatus for Providing Notification Function in Analyte Monitoring Systems
US20140095081A1 (en) * 2012-09-28 2014-04-03 Abbott Diabetes Care Inc. Methods of Analyte Monitoring, and Devices and Systems Related Thereto
US20150216456A1 (en) * 2012-09-26 2015-08-06 Abbott Diabetes Care Inc. Method and Apparatus for Improving Lag Correction During In Vivo Measurement of Analyte Concentration with Analyte Concentration Variability and Range Data
CN110175353A (en) * 2019-04-16 2019-08-27 浙江新再灵科技股份有限公司 Elevator operating index trend analysis
CN112417791A (en) * 2020-11-19 2021-02-26 山东山大电力技术股份有限公司 Power equipment monitoring and early warning method and system based on linear regression and AHP algorithm
CN112990552A (en) * 2021-02-20 2021-06-18 节点互联(北京)科技有限公司 Equipment operation parameter short-time prediction method and system based on change rate
CN113380411A (en) * 2021-07-19 2021-09-10 苏州百孝医疗科技有限公司 Method for improving concentration change real-time trend accuracy in continuous monitoring process of animal analyte concentration

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN100563744C (en) * 2002-10-03 2009-12-02 斯科特实验室公司 Be used for providing the system of trend analysis at calm and analgesic systems
CN107390661B (en) * 2017-08-28 2019-05-24 南京富岛信息工程有限公司 A kind of method for early warning of process flow industry process abnormal state
US11529104B2 (en) * 2017-11-13 2022-12-20 Ai Care Corporation Method of blood pressure estimation using trend analysis

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110077494A1 (en) * 2009-09-29 2011-03-31 Abbott Diabetes Care Inc. Method and Apparatus for Providing Notification Function in Analyte Monitoring Systems
US20150216456A1 (en) * 2012-09-26 2015-08-06 Abbott Diabetes Care Inc. Method and Apparatus for Improving Lag Correction During In Vivo Measurement of Analyte Concentration with Analyte Concentration Variability and Range Data
US20140095081A1 (en) * 2012-09-28 2014-04-03 Abbott Diabetes Care Inc. Methods of Analyte Monitoring, and Devices and Systems Related Thereto
CN110175353A (en) * 2019-04-16 2019-08-27 浙江新再灵科技股份有限公司 Elevator operating index trend analysis
CN112417791A (en) * 2020-11-19 2021-02-26 山东山大电力技术股份有限公司 Power equipment monitoring and early warning method and system based on linear regression and AHP algorithm
CN112990552A (en) * 2021-02-20 2021-06-18 节点互联(北京)科技有限公司 Equipment operation parameter short-time prediction method and system based on change rate
CN113380411A (en) * 2021-07-19 2021-09-10 苏州百孝医疗科技有限公司 Method for improving concentration change real-time trend accuracy in continuous monitoring process of animal analyte concentration

Also Published As

Publication number Publication date
CN113380411B (en) 2024-03-01
CN113380411A (en) 2021-09-10

Similar Documents

Publication Publication Date Title
WO2023000530A1 (en) Method for improving accuracy of real-time concentration change trend during continuous monitoring of analyte concentration in animal body
US20200105397A1 (en) Method and apparatus for determining meal start and peak events in analyte monitoring systems
Maran et al. Lack of preservation of higher brain function during hypoglycaemia in patients with intensively-treated IDDM
CN103959291B (en) Glucose predictions device based on the regularization network with adaptively selected core and regularization parameter
Signal et al. Impact of retrospective calibration algorithms on hypoglycemia detection in newborn infants using continuous glucose monitoring
CN106413541A (en) Systems and methods for diagnosing sleep
WO2023115751A1 (en) Blood glucose prediction method and apparatus, and blood glucose level measurement system
JP4626700B2 (en) Biological rhythm information acquisition method
Pasquet et al. Relationship between taste thresholds and hunger under debate
Chen et al. Committed moving horizon estimation for meal detection and estimation in type 1 diabetes
Gosa et al. Reliability for identification of a select set of temporal and physiologic features of infant swallows
Joshi et al. Effect of examination stress on the plasma cortisol level
Kongsbak et al. Acute phase protein levels, T. trichiura, and maternal education are predictors of serum zinc in a cross-sectional study in Bangladeshi children
CN110911008A (en) Method for establishing biliary tract occluded alignment chart prediction model and application thereof
CN115662617B (en) Result illness state prediction system based on CGM and prediction method thereof
Marion et al. Congenital central hypoventilation syndrome and the PHOX2B gene mutation
CN113447655A (en) Blood glucose index combination for judging CPAP treatment effect
CN112768031A (en) Health management method for conditioning physical examination of comprehensive diet sports body and preventing chronic diseases
Syafaah et al. Self-Monitoring and Detection of Diabetes with Smart Toilet based on Image Processing and K-Means Technique
Owens History and vision: what is important for patients with diabetes?
CN106885910B (en) Tumor susceptibility microenvironment detection device and kit
EP3373173A1 (en) Method and apparatus for monitoring a subject
CN114431898A (en) Ketone body detection device based on respiratory gas
US20230019703A1 (en) Personal health tracking urinalysis device
Brummer et al. Continuous glucose monitoring for automatic real-time assessment of eating events and nutrition: a scoping review

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21950768

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE