Fall detection system

• US 10,198,928 B1
• Filed: 12/29/2017
• Issued: 02/05/2019
• Est. Priority Date: 12/29/2017
• Status: Active Grant

• Current Assignee: MedHab, LLC
• Original Assignee: MedHab, LLC
• Inventors: Ross, Johnny, Olsen, Simon Joakim, Samuelson, Donna Caroline, Gray, Mathew Aaron
• Primary Examiner(s): Alam, Mirza F
• Application Number:
  US15/858,930
• Time in Patent Office: 403 Days
• 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 5/7267 involving training the classification device
• CPC Class Codes: A61B 5/746 Alarms related to a physiological condition, e.g. details of setting alarm thresholds or avoiding false alarms alarms per se G08B
• 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: G16H 50/70 for mining of medical data, e.g. analysing previous cases of other patients

Abstract

A system used to detect a fall and to provide data has a wearable device that includes a pre-filter, an accelerometer and a transmitter. The pre-filter includes buffers, a low pass filter, a flag, and a measuring device. The pre-filter receives data points transmitted by the accelerometer. The pre-filter generates a magnitude value, and calculates a jerk value in relation to the magnitude value. The pre-filter appends the jerk value to one of the buffers and provides the jerk value to the measuring device. The pre-filter further transmits the magnitude value to a neural network.

Claims

1. A computer-implemented method for providing data to a neural network for detecting a fall of person having a portable electronic device, the method comprising the steps of:

phasing, via a processing device of a pre-filter, data points, wherein the phasing comprises dividing the data points into a predetermined number of data sets;

receiving a data point of the predetermined number of data sets at the processing device of the pre-filter, wherein the data point is selected for pre-filtering, and wherein the pre-filter comprises a first buffer, a second buffer, a low pass filter, and a measuring device;

calculating a first magnitude value in view of the data point;

passing the first magnitude value to the low pass filter to generate a second magnitude value;

calculating a jerk value in relation to the second magnitude value and a third magnitude value, wherein the third magnitude value is a magnitude value appended to the first buffer;

appending the jerk value to the second buffer;

appending the second magnitude value to the first buffer;

providing the jerk value to the measuring device; and

transmitting the second magnitude value to the neural network.

2. The computer-implemented method of claim 1, wherein the pre-filter comprises a flag and wherein the measuring device comprises a high threshold detector, the method further comprising a step of:

setting the flag in response to the high threshold detector detecting the jerk value meeting a high threshold, wherein the high threshold is greater than or equal to a predetermined value.

3. The computer-implemented method of claim 1, wherein the pre-filter comprises a flag and wherein the measuring device comprises a low threshold detector and when the flag is set, the method further comprising steps of:

passing the jerk value to the low threshold detector;

detecting the jerk value meeting a low threshold, wherein the low threshold is lower than a predetermined value; and

transmitting the second magnitude value to the neural network in response to the detecting of the jerk value meeting the low threshold.

4. The computer-implemented method of claim 1, wherein the pre-filter comprises a flag and in response the flag being set, meeting of a low threshold not being detected, and in response to greater than a predetermined number of data points being passed to the pre-filter since the flag was set, the method further comprising a step of:

unsetting the flag.

5. The computer-implemented method of claim 1, wherein the third magnitude value is a last magnitude value appended to the first buffer.

6. A system for providing data to a neural network, the system comprising:

a wearable device comprising a pre-filter, an accelerometer and a transmitter configured to communicate with a portable electronic device, wherein the pre-filter comprises a first buffer, a second buffer, a low pass filter, and a measuring device, wherein the pre-filter is configured to;

perform, via a processing device, phasing of data points, wherein the data points are transmitted by the accelerometer, and wherein the phasing comprises dividing data points into a predetermined number of data sets;

receive a data point of the predetermined number of data sets, wherein the data point is selected for pre-filtering;

calculate a first magnitude value in view of the data point;

pass the first magnitude value to the low pass filter to generate a second magnitude value;

calculate a jerk value in relation to the second magnitude value and a third magnitude value, wherein the third magnitude value is a magnitude value appended to the first buffer;

append the jerk value to the second buffer;

append the second magnitude value to the first buffer;

provide the jerk value to the measuring device; and

transmit the second magnitude value to the neural network.

7. The system of claim 6, wherein the pre-filter comprises a flag and wherein the measuring device comprises a high threshold detector, the pre-filter is configured to:

set the flag in response to the high threshold detector detecting the jerk value meeting a high threshold, wherein the high threshold is greater than or equal to a predetermined value.

8. The system of claim 6, wherein the pre-filter comprises a flag and wherein the measuring device comprises a low threshold detector and when the flag is set, the pre-filter is configured to:

pass the jerk value to the low threshold detector;

detect the jerk value meeting a low threshold, wherein the low threshold is lower than a predetermined value; and

transmit the second magnitude value to the neural network in response to detecting of the jerk value meeting the low threshold.

9. The system of claim 8, wherein the pre-filter sets the flag.

10. The system of claim 9, wherein the pre-filter is configured to unset the flag in response to:

the low threshold detector not detecting the jerk value meeting the low threshold; and

a predetermined number of data points being passed to the pre-filter since the flag was set.

11. A wearable device system comprising:

a pre-filter;

an accelerometer; and

a transmitter configured to communicate with a portable electronic device, wherein the pre-filter comprises a first buffer, a second buffer, a low pass filter, a flag, and a low threshold detector, wherein the pre-filter is configured to;

perform, via a processing device, phasing of data points, wherein the data points are transmitted by the accelerometer, and wherein the phasing comprises dividing data points into a predetermined number of data sets;

receive a data point of the predetermined number of data sets, wherein the data point is selected for pre-filtering;

calculate a first magnitude value in view of a data point;

pass the first magnitude value to the low pass filter to generate a second magnitude value;

calculate a jerk value in relation to the second magnitude value and a third magnitude value, wherein the third magnitude value is a magnitude value appended to the first buffer;

append the jerk value to the second buffer;

append the second magnitude value to the first buffer;

provide the jerk value to the low threshold detector in response to the flag being set; and

provide the second magnitude value to the transmitter in response to the low threshold detector detecting the jerk value meeting a low threshold, wherein the low threshold is lower than a predetermined value.

12. The wearable device system of claim 11, wherein the pre-filter further comprises a high threshold detector, the pre-filter is configured to:

set the flag in response to the high threshold detector detecting the jerk value meeting a high threshold, wherein the high threshold is greater than or equal to a predetermined value.

13. The wearable device system of claim 11, wherein the pre-filter sets the flag.

14. The wearable device system of claim 13, wherein the pre-filter is configured to unset the flag in response to:

the low threshold detector not detecting the jerk value meeting the low threshold; and

a predetermined number of data points being passed to the pre-filter since the flag was set.