System and method for predicting blood glucose level

• US 5,971,922 A
• Filed: 10/16/1998
• Issued: 10/26/1999
• Est. Priority Date: 04/07/1998
• Status: Expired due to Fees

• Current Assignee: Research Institute of Application Technologies For Chaos & Complex Systems Co., Ltd.
• Original Assignee: Kabushiki Kaisha Meidensha
• Inventors: Iokibe, Tadashi, Yoneda, Masaya, Arita, Seizaburou
• Primary Examiner(s): O'Connor, Cary
• Assistant Examiner(s): Natnithithadha, Navin
• Application Number:
  US09/174,258
• Time in Patent Office: 375 Days
• Field of Search: 600/365, 600/319, 600/300, 600/301, 706/58, 706/45
• US Class Current: 600/365
• CPC Class Codes: A61B 5/14532 for measuring glucose, e.g. by tissue impedance measurement
• CPC Class Codes: A61B 5/7264 Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems neural networks per se G06N3/00; expert systems per se G06N5/00
• CPC Class Codes: A61B 5/7275 Determining trends in physiological measurement data; Predicting development of a medical condition based on physiological measurements, e.g. determining a risk factor
• CPC Class Codes: G16H 20/10 relating to drugs or medications, e.g. for ensuring correct administration to patients
• CPC Class Codes: G16H 40/63 for local operation
• CPC Class Codes: G16H 50/30 for calculating health indices; for individual health risk assessment
• CPC Class Codes: G16H 50/50 for simulation or modelling of medical disorders

Abstract

A system for predicting a blood glucose level is comprised of a time series measurement data storing section for storing blood glucose level measured data in a blood glucose time series file to treat the data as time series data. A dynamics estimating section estimates a dynamics which most preferably represents a phase characteristic of the time series data stored in the time series measurement data storing section. A parameter storing section stores an embedding dimension n and a time delay τ of the dynamics estimated in the dynamics estimating section as parameters for embedding the estimated dynamics in multidimensional state space. A blood glucose predicting section predicts a near future value of blood glucose level by means of the Local Fuzzy Reconstruction Method on the basis of the stored data of the blood glucose level data stored in the blood glucose time series file and the parameters corresponding to the data and for storing the predicted future value in a predicted blood glucose level file. A display section displays the data of the blood glucose time series file and the predicted blood glucose level file.

Claims

1. A system for predicting a blood glucose level comprising:

a time series measurement data storing means for storing blood glucose level measured data in a blood glucose time series file to treat the data as time series data;

a dynamics estimating means for estimating a dynamics which most preferably represents a phase characteristic of the time series data stored in said time series measurement data storing means;

a parameter storing means for storing an embedding dimension n and a time delay τ

of the dynamics estimated in said dynamics estimating means as parameters for embedding the estimated dynamics in a multidimensional state space;

a blood glucose predicting means for predicting a near future value of the blood glucose level by means of the Local Fuzzy Reconstruction Method on the basis of the data of the blood glucose level stored in the blood glucose time series file and the parameters corresponding to the data, said blood glucose predicting means storing the predicted future value in a predicted blood glucose level file; and

a display means for displaying the data of the blood glucose time series file and the predicted blood glucose file.

2. A method for predicting blood glucose level comprising the steps of:

preparing blood glucose level data measured at latest time and past time for use as time series data;

constructing an attractor by embedding the time series data in a multidimensional state space according to the Takens'"'"' embedding theorem;

selecting a data vector z(T) on the attractor which includes the latest blood glucose level data;

selecting a plurality of neighboring vectors x(i) on the other trajectory passing through a neighbor space of the data vector z(T) on the basis of a selecting reference that the Euclidean distance thereof is smaller that a predetermined value;

selecting a data vector x(i+s) at s steps future with respect to the data vector x(i) from the attractor;

estimating (inferring) a predicted value z(T+s) at s steps future with respect to the data vector z(T) by using the data vectors z(T), x(i) and x(i+s) by means of the Local Fuzzy Reconstruction Method; and

obtaining a predicted blood glucose level y(T+s) at s step future on the basis of the predicted value z(T+s).

3. A method as claimed in claim 2 further comprising a step for displaying the predicted value z(T+s).

4. A method as claimed in claim 2 stored in a medium usable in a microcomputer.

5. A system for predicting a blood glucose level comprising:

a self-checked blood glucose input section sending a self checked value of blood glucose level measured by a diabetic at predetermined intervals;

a blood glucose time series file storing the self-checked blood glucose data sent from said blood glucose input section as time series data by each patient;

a dynamics estimating section estimating,a dynamics which most preferably represents a phase characteristic of each time series data stored in said blood glucose time series file;

an optimum embedding parameter file storing an embedding dimension and a time delay obtained at said dynamics estimating section as parameters for each patient;

a blood glucose predicting section executing a prediction of the blood glucose level at a near future on the basis of the selected data and the parameters by means of the Local Fuzzy Reconstruction Method;

a predicted blood glucose file storing the blood glucose level dada predicted at said blood glucose predicting section for each patient;

an insulin administration amount input section sending the amount of the insulin practically administrated to the patient;

an insulin administration amount time series file storing the insulin administration amount data sent from said insulin administration amount input section as time series data by each patient; and

a display section displaying data for each patient upon searching the data from said blood glucose time series file, said predicted blood glucose file and said insulin administration amount time series file to provide necessary information for the treatment for diabetic to a doctor.

6. A system as claimed in claim 5 wherein the self-checked blood glucose data is sent from said blood glucose input section by means of a communication means which includes internet communication, personal handy-phone system, on-line personal computer communication, pocket pager, and facsimile.