METHOD AND DEVICE AND PREDICTING PHYSIOLOGICAL VALUES
The invention relates generally to methods, systems, and devices for measuring the concentration of target analytes present in a biological system using a series of measurements obtained from a monitoring system and a Mixtures of Experts (MOE) algorithm. In one embodiment, the present invention describes a method for measuring blood glucose in a subject.
1 - 24 . (canceled)
25 . A method for measuring an amount or concentration of analyte present in a biological system, said method comprising:
determining a measurement value indicative of the amount or concentration of analyte present in the biological system;
providing two or more ranges of measurement values;
identifying the range in which said determined measurement value falls;
employing an algorithm for prediction of further measurement values wherein said algorithm is optimized for performance in the identified range; and
generating further measurement values indicative of amount or concentration of analyte present in the biological system using said algorithm
26 . The method of claim 25 , wherein said algorithm for prediction of further measurement values is a Mixtures of Experts algorithm and said Mixtures of Experts algorithm is trained using data from the identified range.
27 . The method of claim 28 , wherein for the Mixtures of Experts algorithm the individual experts have a linear form
An
=
∑
i
=
1
n
An
i
w
i
(
1
)
wherein (An) is an analyte of interest, n is the number of experts, An i is the analyte predicted by Expert i; and w i is a parameter, and the individual experts An i are further defined by the expression shown as Equation (2)
An
i
=
∑
j
=
1
m
a
ij
P
j
+
z
i
(
2
)
wherein, An i is the analyte predicted by Expert i; P j is one of m parameters, m is typically less than 100; a ij are coefficients; and z i is a constant; and further where the weighting value, w i , is defined by the formula shown as Equation (3)
w
i
=
d
i
[
∑
k
=
1
n
d
k
]
(
3
)
where e refers to the exponential function, d i is one of the d k , d i and d k are parameter sets analogous to Equation 2 used to determine the weight w i , the d k are given by Equation 4
d
k
=
∑
j
=
1
m
α
jk
P
j
+
ω
k
(
4
)
where a jk is coefficient, P j is one of m parameters, and where ω k is a constant.
28 . The method of claim 25 , wherein the analyte is glucose.
29 . The method of claim 25 , further comprising identifying in which range one or more of the further measurement values falls, and employing an algorithm for prediction of further measurement values wherein said algorithm is optimized for performance in the identified range.
30 . The method of claim 25 , wherein said generating further measurement values indicative of amount or concentration of analyte present in the biological system comprises obtaining a raw signal specifically related to analyte amount or concentration in the biological system and using said algorithm to correlate the raw signal with a measurement value.
31 . The method of claim 25 , wherein said determining a measurement value indicative of the amount or concentration of analyte present in the biological system compromises a calibration step.
32 . The method of claim 31 , wherein said calibrating step correlates a raw signal obtained from a sensing device with a concentration of analyte present in the biological system.
33 . The method of claim 44 , wherein said calibration step provides a calibrated signal by a method compromising
signal
=
BG
cp
active
cp
(
active
)
wherein, signal is the calibrated signal, BG cp is blood glucose value at a calibration point, active cp is an active signal that corresponds to an electrochemical sensor signal at the calibration point, and active is an active signal that corresponds to an electrochemical sensor signal.
34 . The method of claim 44 , wherein said calibration step provides a calibrated signal by a method comprising
signal
=
BG
cp
(
active
cp
+
offset
)
(
active
+
offset
)
wherein, signal is the calibrated, signal, BG cp is blood glucose value at a calibration point, active cp is an active signal that corresponds to an electrochemical sensor signal at the calibration point, active is an active signal that corresponds to an electrochemical sensor signal, and offset is a value that takes into account a non-zero y-intercept value.