Method and system for fall-onset detection

• US 8,260,570 B2
• Filed: 01/22/2008
• Issued: 09/04/2012
• Est. Priority Date: 01/22/2007
• Status: Expired due to Fees

• Current Assignee: National University of Singapore, Singapore General Hospital
• Original Assignee: National University of Singapore, Singapore General Hospital
• Inventors: Nyan, Myo Naing, Tay, Eng Hock Francis, Murugasu, Euan
• Primary Examiner(s): Bhat, Aditya
• Application Number:
  US12/523,808
• Publication Number:
  US 20100121603A1
• Time in Patent Office: 1,687 Days
• Field of Search: 702/160
• US Class Current: 702/160
• CPC Class Codes: A61B 2562/0219 Inertial sensors, e.g. accelerometers, gyroscopes, tilt switches
• CPC Class Codes: A61B 5/1117 Fall detection
• CPC Class Codes: A61B 5/6828 Leg
• CPC Class Codes: A61B 5/6831 Straps, bands or harnesses
• CPC Class Codes: G08B 21/0446 worn on the body to detect changes of posture, e.g. a fall, inclination, acceleration, gait

Abstract

A method and system for fall-onset detection is provided. The method includes the steps of monitoring acceleration at the thigh of a person; monitoring acceleration at the waist of the person; monitoring orientation of the thigh of the person; and detecting the fall-onset if the orientation of the thigh exceeds a first threshold value and a correlation between variation in acceleration of the thigh and variation in acceleration of the waist exceeds a second threshold value and a correlation between variation in the orientation of the thigh and a fall-template variation exceeds a third threshold value.

Claims

1. A method of fall-onset detection, the method comprising the steps of:

providing an accelerometer for monitoring acceleration;

monitoring acceleration at the thigh of a person;

monitoring acceleration at the waist of the person;

monitoring orientation of the thigh of the person; and

detecting the fall-onset if

the orientation of the thigh exceeds a first threshold value and

a correlation between variation in acceleration of the thigh and variation in acceleration of the waist exceeds a second threshold value and

a correlation between variation in the orientation of the thigh and a fall-template variation exceeds a third threshold value.

2. The method as claimed in claim 1, wherein the first threshold value is about ±

10°

.

3. The method as claimed in claim 1, wherein the second threshold value is about 0.88.

4. The method as claimed in claim 1, wherein the third threshold value is about 0.8.

5. The method as claimed in claim 4, wherein the acceleration of the waist is monitored three-dimensionally as:

{x_WS;

vertical (downward positive);

y_ws;

lateral (right positive);

z_ws;

sagittal (forward positive)}.

6. The method as claimed in claim 4, wherein the orientation of the thigh is measured two-dimensionally as:

lateral (left positive) and sagittal (back positive) directions.

7. The method as claimed in claim 1, wherein the acceleration of the thigh is monitored three-dimensionally as:

{x_TS;

vertical (downward positive);

y_TS;

lateral (right positive);

z_TS;

sagittal (forward positive)}.

8. The method as claimed in claim 1, wherein the monitored acceleration of the thigh and waist are transformed into two-dimensional body orientation and the correlation between the variation in acceleration of the thigh and variation in acceleration of the waist is determined with regard to the transformed two-dimensional body orientation.

9. The method as claimed in claim 8, wherein the monitored acceleration of the thigh and waist are transformed into two-dimensional body orientation deg_SAG, deg_LAT, as:

deg_SAG=−

atan (z_TSorWS/x_TSorWS)*(180/pi),

e=sqrt(1−

(y_TSorWS*y_TSorWS)), and

deg_LAT=atan (y_TSorWS/e)*(180/pi).

10. The method as claimed in claim 1, wherein the correlation between variation in acceleration of the thigh and variation in acceleration of the waist is determined over a first data segment of pre-determined sample length.

11. The method as claimed in claim 10, wherein the sample length is about 20.

12. The method as claimed in claim 1, wherein the correlation between variation in the orientation of the thigh and a fall-template variation is determined over a second data segment of pre-determined sample length.

13. The method as claimed in claim 12, wherein the sample length is about 80.

14. The method as claimed in claim 1, further comprising performing a feedback step in response to the fall-onset detection.

15. A system for fall-onset detection, the system comprising:

a first accelerometer for monitoring acceleration at the thigh of a person;

a second accelerometer for monitoring acceleration at the waist of the person;

a gyroscope for monitoring orientation of the thigh of the person; and

a processor coupled to the first and second accelerometers and to the gyroscope for detecting the fall-onset, the processor being operable to detect the fall-onset based if

the orientation of the thigh exceeds a first threshold value and

a correlation between variation in acceleration of the thigh and variation in acceleration of the waist exceeds a second threshold value and

a correlation between variation in the orientation of the thigh and a fall-template variation exceeds a third threshold value.