Multiparameter method for quantifying balance

Patent metadata

Field	Value
Publication number	US-11607166-B2
Application number	US-201716346395-A
Country	US
Kind code	B2
Filing date	Nov 9, 2017
Priority date	Nov 9, 2016
Publication date	Mar 21, 2023
Grant date	Mar 21, 2023

Abstract

A method for quantifying the balance of an individual recording, on a memory, at least one statokinesigram of the individual obtained from a platform comprising pressure and/or force sensors; extracting, by a processor and from the at least one statokinesigram of the individual recorded on the memory, values of at least one position trajectory parameter of the pressure center and values of at least one stability trajectory parameter of the pressure center; determining, by the processor, the value of a plurality of quantifiers, from the values of the trajectory parameters extracted; comparing, by the processor, said values of the plurality of quantifiers with the values of the same quantifiers obtained from reference statokinesigrams; and determining, by the processor, said value representative of the balance of the individual at the end of the comparison.

First claim

The invention claimed is:

1. A method for quantifying the balance of an individual to obtain a value representative of the balance of said individual, said method being implemented by a device comprising at least one processor connected to a memory and said method comprising:
a) a step of recording, on the memory, at least one statokinesigram of the individual obtained from a platform comprising pressure and/or force sensors,
b) a step of extracting, by the processor and from the at least one statokinesigram of the individual recorded on the memory, values of at least one position trajectory parameter of a pressure center and values of at least one stability trajectory parameter of the pressure center, wherein the at least one position trajectory parameter of the pressure center is selected from: the position of the pressure center along an X axis, the position of the pressure center along a Y axis, and a radius in polar coordinates, wherein the at least one stability trajectory parameter of the pressure center is selected from: a radial balance, a time balance, and a ballistic interval,
c) a step of determining, by the processor, values of a plurality of quantifiers, from the values of the trajectory parameters extracted in step b), at least one of the plurality of quantifiers determined in step c) being selected from: a mean value, a median value, a variance value, a mean square value, and an extreme value, of at least one of said trajectory parameters,
d) a step of comparing, by the processor, said values of the plurality of quantifiers with values of the same quantifiers obtained from reference statokinesigrams, wherein the comparing is based on applying a learning model that is trained on a dataset from a set of individuals representative of a population and characterized by the reference statokinesigrams and the same quantifiers, the learning model configured to generate a prediction of an observation that is associated with the dataset, the learning model comprising a non-linear statistical model, and
e) a step of determining, by the processor, said value representative of the balance of the individual at the end of the comparison.

2. The method according to claim 1 , wherein:
the recording step further comprises recording body mass index (BMI) data of said individual, and is followed by a step of classifying, by the processor, the at least one statokinesigram of the individual in a BMI category depending on the recorded BMI data, said classification step taking place after the recording step a) and before the extraction step b),
during the step of comparing said values of said plurality of quantifiers, said values of said plurality of quantifiers are compared with the values of the same quantifiers obtained from reference statokinesigrams classified in the same BMI category as the BMI category determined during the classification step.

3. The method according to claim 2 , wherein the BMI category determined in the classification step is selected from at least three BMI categories.

4. The method according to claim 1 , wherein step b) further comprises extracting values of at least one dynamics trajectory parameter of the pressure center.

5. The method according to claim 4 , wherein the at least one dynamics trajectory parameter of the pressure center is selected from: a velocity of displacement of the pressure center, acceleration of the displacement of the pressure center, a power and the deviation.

6. The method according to claim 1 , wherein the at least one position trajectory parameter of the pressure center comprises its radius in polar coordinates.

7. The method according to claim 1 , wherein step b) is performed from two statokinesigrams generated during a Romberg test.

8. The method according to claim 1 , wherein at least one of the quantifiers determined in step c) is an extreme value of one of said trajectory parameters, said extreme value corresponding to a percentile greater than or equal to 5 and less than or equal to 15, or to a percentile greater than or equal to 85 and less than or equal to 95, of the trajectory parameter values.

9. The method according to claim 2 , wherein steps d) and e) are performed by implementing the values of the quantifiers determined in step c) in a scoring algorithm previously calibrated based on the values of the same quantifiers obtained from the reference statokinesigrams.

10. The method according to claim 9 , wherein the scoring algorithm is previously calibrated based on the values of the same quantifiers obtained from the reference statokinesigrams classified in the same BMI category as the BMI category determined in step a).

11. The method according to claim 10 , wherein the previously calibrated scoring algorithm is obtained by implementing a supervised learning statistical method comprising implementing a Bagging step.

12. The method according to claim 1 , further comprising, in step c), determining the values of at least five quantifiers.

13. The method according to claim 1 , wherein:
in step b), the extraction step is performed from a first said statokinesigram obtained while the individual's eyes are open and a second said statokinesigram obtained while the individual's eyes are closed; and
in step d), the comparing step comprises comparing values of the same quantifiers obtained from the first and second said statokinesigrams obtained with the eyes closed or with the eyes open.

14. The method according to claim 1 , further comprising determining, by the processor, a risk of the individual falling within a predetermined period of time based on the value representative of the balance of the individual.

15. The method according to claim 1 , further comprising ranking, by the processor, the individual within a binary label classification based on the value representative of the balance of the individual.

16. The method according to claim 1 , further comprising tracking, by the processor, progression of the value representative of the balance of the individual by one or more of: monitoring an evolution of the value over a predetermined period of time; monitoring a position of the value relative to other individuals; or identifying deviations of the value relative to a reference value.

17. The method according to claim 1 , further comprising evaluating, by the processor, the value representative of the balance of the individual to at least one or more of: target customized treatment for the individual; or determine an origin of a balance disorder associated with the individual.

18. The method according to claim 1 , wherein the learning model is configured to generate the prediction of the observation based on a binary label classification.

19. The method according to claim 1 , wherein steps (d) and (e) are performed by implementing the values of the plurality of quantifiers in a scoring algorithm previously calibrated based on the values of the same quantifiers obtained from the reference statokinesigrams, the previously calibrated scoring algorithm having been obtained by implementing a supervised statistical learning model that is built from a combination of a plurality of non-linear supervised statistical learning models that are configured to aggregate predictions of a same, randomly selected variable.

20. A device for quantifying the balance of an individual, said device comprising:
a receiver and/or transmitter configured to receive at least one statokinesigram of said individual,
a memory configured to record the at least one statokinesigram, and
at least

Assignees

• Centre Nat Rech Scient
• Ecole Normale Superieure Paris Saclay
• Etat Francais—Ministere De La Defense—Direction Centrale Du Service De Sante Des Armee
• Univ Paris Descartes

Inventors

• Vayatis Nicolas
• Vidal Pierre Paul
• Promponas Kefalas Nikolaos
• Audiffren Julien
• Yelnik Alain
• De Waele Vidal Catherine
• Ricard Damien

Classifications

• A61B5/4023 Primary

  Evaluating sense of balance · CPC title

• G16H50/30

  for calculating health indices; for individual health risk assessment · CPC title

• G16H40/63 Primary

  for local operation · CPC title