Ensemble of partitioned sensor glucose models

Patent metadata

Field	Value
Publication number	US-2025134417-A1
Application number	US-202418908980-A
Country	US
Kind code	A1
Filing date	Oct 8, 2024
Priority date	Oct 27, 2023
Publication date	May 1, 2025
Grant date	—

Abstract

A processor-implemented method includes receiving sensor measurement data from a glucose sensor; selecting, based on the sensor measurement data, a first regional sensor glucose (SG) model from a first plurality of regional SG models for respective regions of a first plurality of regions of an input parameter space associated with the sensor measurement data, and a second regional SG model from a second plurality of regional SG models for respective regions of a second plurality of regions of the input parameter space; estimating a first SG value and a second SG value using the first regional SG model and the second regional SG model, respectively; and determining a predicted SG value based on a combination of the first SG value and the second SG value. The input parameter space is partitioned into the first plurality of regions and the second plurality of regions using different partition schemes.

First claim

What is claimed is:

1 . A processor-implemented method comprising:
receiving sensor measurement data measured by a glucose sensor, the sensor measurement data including at least one signal that is indicative of glucose level and at least one signal that is indicative of sensor health; selecting, based at least in part on the sensor measurement data, a first regional sensor glucose (SG) model from a first plurality of regional SG models for respective regions of a first plurality of regions of an input parameter space associated with the sensor measurement data, wherein the input parameter space is partitioned into the first plurality of regions based on a first partition scheme; estimating a first SG value using the first regional SG model and the sensor measurement data; selecting, based at least in part on the sensor measurement data, a second regional SG model from a second plurality of regional SG models for respective regions of a second plurality of regions of the input parameter space, wherein the input parameter space is partitioned into the second plurality of regions based on a second partition scheme that is different from the first partition scheme; estimating a second SG value using the second regional SG model and the sensor measurement data; and determining a predicted SG value based on a combination of the first SG value and the second SG value.

2 . The processor-implemented method of claim 1 , wherein at least one of the first partition scheme or the second partition scheme partitions the input parameter space based at least in part on sensor current (Isig), counter voltage (Vcntr), electrochemical impedance spectroscopy (EIS) data, age of the glucose sensor, temperature, age of a user of the glucose sensor, body mass index (BMI) of the user, or a combination thereof.

3 . The processor-implemented method of claim 1 , wherein the first partition scheme or the second partition scheme partitions the input parameter space based at least in part on a sensor current and an age of the glucose sensor.

4 . The processor-implemented method of claim 1 , wherein the first partition scheme and the second partition scheme use different combinations of input parameters to partition the input parameter space.

5 . The processor-implemented method of claim 1 , wherein the first partition scheme and the second partition scheme use different numbers of input parameters to partition the input parameter space.

6 . The processor-implemented method of claim 1 , wherein the first partition scheme and the second partition scheme partition the input parameter space using different ranges of a same set of input parameters.

7 . The processor-implemented method of claim 1 , wherein regions of the first plurality of regions of the input parameter space are characterized by different sizes, different shapes, different numbers of dimensions, or a combination thereof.

8 . The processor-implemented method of claim 1 , wherein a number of regions of the first plurality of regions of the input parameter space is the same as or different from a number of regions of the second plurality of regions of the input parameter space.

9 . The processor-implemented method of claim 1 , further comprising generating at least one of a derivative feature or a transformed feature based on the sensor measurement data, wherein input parameters of the first plurality of regional SG models and the second plurality of regional SG models include at least one of the derivative feature or the transformed feature.

10 . The processor-implemented method of claim 1 , wherein determining the predicted SG value based on the combination of the first SG value and the second SG value comprises determining an average or a weighted average of the first SG value and the second SG value.

11 . The processor-implemented method of claim 10 , further comprising:
determining a first probability or confidence level associated with the first SG value; and determining a second probability or confidence level associated with the first SG value, wherein weights associated with the first SG value and the second SG value for the weighted average are determined based on the first probability or confidence level, and the second probability or confidence level.

12 . The processor-implemented method of claim 1 , further comprising:
selecting, based at least in part on the sensor measurement data, a third regional SG model from a third plurality of regional SG models for respective regions of a third plurality of regions of the input parameter space, wherein the input parameter space is partitioned into the third plurality of regions based on a third partition scheme that is different from the first partition scheme and the second partition scheme; and estimating a third SG value using the third regional SG model and the sensor measurement data, wherein determining the predicted SG value includes determining the predicted SG value based on a combination of the first SG value, the second SG value, and the third SG value.

13 . The processor-implemented method of claim 1 , wherein:
each regional SG model of the first plurality of regional SG models and the second plurality of regional SG models includes one or more machine learning models, equations, functions, or a combination thereof; and a first regional SG model and a second regional SG model in the first plurality of regional SG models and the second plurality of regional SG models use different input parameters.

14 . A system comprising:
one or more processors; and one or more processor-readable media storing instructions which, when executed by the one or more processors, cause performance of operations including: receiving sensor measurement data measured by a glucose sensor, the sensor measurement data including at least one signal that is indicative of glucose level and at least one signal that is indicative of sensor health; selecting, based at least in part on the sensor measurement data, a first regional sensor glucose (SG) model from a first plurality of regional SG models for respective regions of an input parameter space associated with the sensor measurement data, wherein the input parameter space is partitioned into the first plurality of regions based on a first partition scheme; estimating a first SG value using the first regional SG model and the sensor measurement data; selecting, based at least in part on the sensor measurement data, a second regional SG model from a second plurality of regional SG models for respective regions of the input parameter space, wherein the input parameter space is partitioned into the second plurality of regions based on a second partition scheme that is different from the first partition scheme; estimating a second SG value using the second regional SG model and the sensor measurement data; and determining a predicted SG value based on a combination of the first SG value and the second SG value.

15 . The system of claim 14 , wherein at least one of the first partition scheme or the second partition scheme partitions the input parameter space based at least in part on sensor current (Isig), counter voltage (Vcntr), electrochemical impedance spectroscopy (EIS) data, age of the glucose sensor, temperature, age of a user of the glucose sensor, body mass index (BMI) of the user, or a combination thereof.

16 . The system of claim 14 , wherein the first partition scheme or the second partition scheme partitions the input parameter space based at least in part on a sensor current and an age of the glucose sensor.

Assignees

• Medtronic Minimed Inc

Inventors

• Basilico Jake G
• Nava-Guerra Leonardo
• Notghi Bahram B
• Emig Molly E
• Liu Jing
• Willats Adam A
• Mallas Georgios
• Zhang Yi

Classifications

• A61B5/7203

  for noise prevention, reduction or removal · CPC title

• A61B5/7264

  Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems · CPC title

• A61B5/14532 Primary

  for measuring glucose, e.g. by tissue impedance measurement · CPC title

• A61B5/7271

  Specific aspects of physiological measurement analysis (specific diagnostics methods using bioelectric or biomagnetic signals A61B5/316) · CPC title