US 12,094,608 B2
Pattern recognition engine for blood glucose measurements
Milena Saleh, Frankfurt am Main (DE); Anton Petkov, Frankfurt am Main (DE); Jochen Sieber, Frankfurt am Main (DE); Giacomo Vespasiani, San Benedetto del Tronto (IT); and Sandro Girolami, San Benedetto del Tronto (IT)
Assigned to Sanofi, Paris (FR)
Appl. No. 17/312,704
Filed by Sanofi, Paris (FR)
PCT Filed Dec. 17, 2019, PCT No. PCT/EP2019/085476
§ 371(c)(1), (2) Date Jun. 10, 2021,
PCT Pub. No. WO2020/127137, PCT Pub. Date Jun. 25, 2020.
Claims priority of application No. 18306744 (EP), filed on Dec. 19, 2018.
Prior Publication US 2022/0051797 A1, Feb. 17, 2022
Int. Cl. G16H 50/20 (2018.01); A61B 5/145 (2006.01); G16H 10/40 (2018.01)
CPC G16H 50/20 (2018.01) [G16H 10/40 (2018.01); A61B 5/14532 (2013.01)] 17 Claims
OG exemplary drawing
 
1. A computer implemented blood glucose analysis method comprising:
receiving a plurality of blood glucose measurements relating to an individual taken over a plurality of temporal frames;
identifying hypoglycaemic and/or hyperglycaemic blood glucose measurements in the plurality of blood glucose measurements;
identifying a plurality of sub-intervals of the temporal frames in which there are a number of hypoglycaemic and/or hyperglycaemic blood glucose measurements that satisfies a threshold condition;
inferring a temporal relationship between blood glucose measurements in a first sub-interval of the identified sub-intervals and blood glucose measurements in a second sub-interval of the identified sub-intervals, wherein inferring the temporal relationships comprises:
determining a plurality of matching pairs of blood glucose measurements, wherein a matching pair comprises a blood glucose measurement in the first sub-interval and a blood glucose measurement in the second sub-interval and wherein determining matching pairs of blood glucose measurements comprises:
comparing time stamps associated with the blood glucose measurements in the first sub-interval to time stamps associated with the blood glucose measurements in the second sub-interval; and
determining a matching pair if a time stamp associated with a blood glucose measurement in the first sub-interval correspond to a time stamp associated with a blood glucose measurement in the second sub-interval,
determining if the total number of matching pairs exceeds a threshold number of pairs, wherein the threshold number of pairs is equal to or more than two; and
in the event of a positive determination, inferring a temporal relationship between the first sub-interval and the second sub-interval;
outputting the identified one or more sub-intervals of the temporal frames; and
outputting a suggested causal driver for the inferred temporal relationship, wherein the causal driver is selected from a database based on the inferred temporal relationship.