Fall detection and/or prevention systems

• US 9,011,352 B2
• Filed: 08/20/2009
• Issued: 04/21/2015
• Est. Priority Date: 08/28/2008
• Status: Active Grant

• Current Assignee: Koninklijke Philips N.V.
• Original Assignee: Koninklijke Philips N.V.
• Inventors: Ten Kate, Warner Rudolph Theophile, Baldus, Heribert, Jin, Sheng, Peng, Yang
• Primary Examiner(s): Dougherty, Sean
• Assistant Examiner(s): Stout, Michael C
• Application Number:
  US13/059,484
• Publication Number:
  US 20110230791A1
• Time in Patent Office: 2,070 Days
• Field of Search: 600/300, 600/301, 600/595
• US Class Current: 600/595
• CPC Class Codes: A61B 5/1117 Fall detection
• CPC Class Codes: G08B 21/043 detecting an emergency event, e.g. a fall
• CPC Class Codes: G08B 21/0446 worn on the body to detect changes of posture, e.g. a fall, inclination, acceleration, gait
• CPC Class Codes: G08B 29/22 Provisions facilitating manual calibration, e.g. input or output provisions for testing; Holding of intermittent values to permit measurement

Abstract

There is provided a fall detection and/or prevention system, comprising one or more sensors for detecting characteristics of movement of a user of the fall detection and/or prevention system and for generating corresponding signals; processing means for analyzing the signals from the one or more sensors using a fall detection algorithm to determine if a fall has taken place or is likely to take place; wherein the processing means is further adapted to update said fall detection algorithm based on the result of the analysis of the signals and an indication whether a fall has actually taken place from the user or a third party.

Claims

1. A fall detection and/or prevention system, comprising:

one or more sensors for detecting characteristics of movement of a user of the fall detection and/or prevention system and for generating corresponding signals;

one or more computers for analyzing the signals from the one or more sensors using a fall detection algorithm to determine if a fall has taken place or is likely to take place, the fall detection algorithm including one or more feature sets indicating falls or non-falls and representing signals from the one or more sensors, the one or more computers being configured to;

update said fall detection algorithm based on the result of the analysis of the signals using the fall detection algorithm to determine if a fall has taken place or is likely to take place and an indication as to whether a fall has actually taken place from the user or a third party, and

monitor the frequency with which the fall detection algorithm is updated, and if the frequency exceeds a threshold, remove one or more feature sets from the fall detection algorithm.

2. The fall detection and/or prevention system as claimed in claim 1, wherein the one or more computers are further configured to generate an alarm signal in the event that a fall has taken place or is likely to take place.

3. The fall detection and/or prevention system as claimed in claim 1, further comprising:

a memory configured to store the one or more feature sets, an indication from the fall detection algorithm of whether the fall detection algorithm has determined that a fall has taken place or is likely to take place, and the indication whether a fall has actually taken place.

4. The fall detection and/or prevention system as claimed in claim 1, wherein the indication whether a fall has actually taken place comprises a reset signal.

5. The fall detection and/or prevention system as claimed in claim 4, wherein one or more computers are further configured to:

determine whether the fall detection algorithm has provided a false positive in the event that the fall detection algorithm detects that a fall has taken place and the reset signal is present, and update the fall detection algorithm accordingly.

6. The fall detection and/or prevention system as claimed in claim 4, wherein the one or more computers are further configured to:

determine whether the fall detection algorithm has provided a true positive in the event that the fall detection algorithm detects that a fall has taken place and the reset signal is not present, and

update the fall detection algorithm accordingly.

7. The fall detection and/or prevention system as claimed in claim 4, further comprising:

an input which receives the reset signal from a third party.

8. The fall detection and/or prevention system as claimed in claim 4, further comprising a first user operable component for allowing a user to selectively generate the reset signal.

9. The fall detection and/or prevention system as claimed in claim 8, further comprising a second user-operable component for generating an alarm signal.

10. The fall detection and/or prevention system as claimed in claim 9, wherein the one or more processors are further configured to:

determine whether the fall detection algorithm has provided a false negative in the event that the fall detection algorithm does not detect that a fall has taken place and the alarm signal is present, and

update the fall detection algorithm accordingly.

11. The fall detection and/or prevention system as claimed in claim 9, wherein the one or more processors are further configured to:

determine whether the fall detection algorithm has provided a true positive in the event that the fall detection algorithm detects that a fall has taken place and the alarm signal is present, and update the fall detection algorithm accordingly.

12. The fall detection and/or prevention system as claimed in claim 1, wherein the fall detection algorithm comprises one or more feature sets representing signals from the one or more sensors.

14. The fall detection and/or prevention system as claimed in claim 12, wherein the one or more computers are configured to determine whether a fall has taken place or is likely to take place by comparing the one or more feature sets with the corresponding signals generated by the one or more sensors.

15. The fall detection and/or prevention system as claimed in claim 1, wherein the one or more computers are configured to update the fall detection algorithm to selectively optimize false positives, in response to:

the algorithm incorrectly detecting a fall or false negatives, or

incorrectly detecting that no fall has taken place to obtain a stable ratio between false positives and false negatives.

13. A fall detection and/or prevention system comprising:

one or more sensors configured to detect characteristics of movement of a user and generate signals corresponding to the detected characteristics of movement;

a computer which analyzes signals from the one or more sensors using a fall detection algorithm which includes one or more features sets representing signals from the one or more sensors to determine if a fall has taken place or is likely to take place, the computer being configured to;

update said fall detection algorithm based on the result of the analysis of the signals and an indication whether a fall has actually taken place from the user or a third party;

monitor the frequency with which the fall detection algorithm is updated, and

if the frequency exceeds a threshold, remove one or more feature sets from the fall detection algorithm.

16. A non-transitory computer-readable medium carrying computer executable code that, when executed on a computer or processor, controls the computer or processor to perform the steps of:

receiving signals indicating characteristics of movement of a user of a fall detection and/or prevention system;

analyzing the signals using a fall detection algorithm which includes a plurality of movement characteristics feature sets to determine if a fall has taken place or is likely to take place;

updating said fall detection algorithm based on the result of the analysis of the signals and an indication whether a fall has actually taken place; and

in response to a frequency with which the fall detection algorithm is updated exceeding a threshold, removing one or more of the feature sets.

17. A method of training a fall detection and/or prevention algorithm for use in a fall detection and/or prevention system, the method comprising:

with movement sensors, obtaining measurements of characteristics of movement of a user;

with one or more computers, analyzing the measurements using a fall detection algorithm which compares characteristics of movement feature sets to determine if a fall has taken place or is likely to take place and updating said fall detection algorithm based on an indication whether a fall has actually taken place and removing at least one of the feature sets in response to a frequency with which the fall detection algorithm is updated exceeding a threshold.

18. The method as claimed in claim 17, wherein the one or more computers update the fall detection algorithm in response to:

the fall detection algorithm determining that a fall has or is likely to take place and an input from the user or a third party indicating no fall has taken place;

or

the fall detecting algorithm not determining that a fall has or is likely to take place and an input from the user or a third party indicating a fall has taken place.