Fall detection

1. A computer-based method for distinguishing between a fall event and a non-fall-event for a plurality of subjects, the method comprising:

obtaining first and second false fall detection rates for first and second sub-groups of the plurality of subjects, respectively, and wherein the first and second sub-groups comprise subjects with first and second differing values of a property of the plurality of subjects, respectively;

determining, based on the first false fall detection rate, a first decision value for distinguishing between a fall event and a non-fall-event of subjects of the first sub-group; and

determining, based on the second false fall detection rate, a second decision value for distinguishing between a fall event and a non-fall-event of subjects of the second sub-group,

wherein the fall detection rate for a sub-group of the plurality of subjects is based on a ratio between a number of false fall detections for the sub-group and the number of subjects in the sub-group.

2. The method of claim 1 , wherein determining the first decision value comprises determining a value at which a false fall detection rate for subjects of the first sub-group is equal to a predetermined rate.

3. The method of claim 2 , wherein the predetermined rate is defined so as to maintain a predetermined minimum detection probability value.

4. The method of claim 1 , wherein the first and second decision values are determined so as to maintain an overall false fall detection rate across the first and second groups remains constant whilst the an average true positive detection rate is increased.

5. The method of claim 1 , further comprising:

associating the first decision value with each subject of the first sub-group; and associating the second decision value with each subject of the second sub-group.

6. The method of claim 1 , wherein obtaining first and second false fall detection rates for first and second sub-groups of the plurality of subjects comprises:

identifying the first and second sub-groups of the plurality of subjects;

monitoring the plurality of subjects to detect potential fall events;

for each detected potential fall event, distinguishing between the detected potential fall event being a fall event and a non-fall-event based on a fall detection algorithm employing a decision value; and

determining first and second false fall detection rates based on the identified first and second sub-groups and the results of distinguishing the potential fall events between being fall events and non-fall-events.

7. The method of claim 1 , wherein the property of the plurality of subjects comprises: location; activity; fall risk; age; medical condition; weight; gender; diagnosis; disease; prescription; or user equipment/aids.

8. A computer program comprising code means for implementing the method of claim 1 when said program is run on a processing system.

9. A system for distinguishing between a fall event and a non-fall-event for a plurality of subjects, the system comprising:

an interface component adapted to obtain first and second false fall detection rates for first and second sub-groups of a plurality of subjects, respectively, and wherein the first and second sub-groups comprise subjects with first and second differing values of a property of the plurality of subjects, respectively; and

a decision value calculation unit adapted to:

determine, based on the first false fall detection rate, a first decision value for distinguishing between a fall event and a non-fall-event of subjects of the first sub-group;

determine, based on the second false fall detection rate, a second decision value for distinguishing between a fall event and a non-fall-event of subjects of the second sub-group; and

determine a value at which a false fall detection rate for subjects of the first sub-group is equal to a predetermined rate.

10. The system of claim 9 , wherein the predetermined rate is defined so as to maintain a predetermined minimum detection probability value.

11. The system of claim 8 , further comprising an assignment unit configured to:

associate the first decision value with each subject of the first sub-group; and

associate the second decision value with each subject of the second sub-group.

12. The system of claim 8 , wherein the interface component comprises:

a grouping component configured to identify the first and second sub-groups of the plurality of subjects;

a monitoring component configured to monitor the plurality of subjects to detect potential fall events;

a classification component configured, for each detected potential fall event, to distinguish between the detected potential fall event being a fall event and a non-fall-event based on a fall detection algorithm employing a decision value; and

a processing component configured to determine first and second false fall detection rates based on the identified first and second sub-groups and the results of distinguishing the potential fall events between being fall events and non-fall-events.

13. A system for detecting a fall of a subject, comprising:

one or more sensors for obtaining movement data responsive to a subject's movement;

the system for distinguishing between a fall event and a non-fall-event for a plurality of subjects of claim 9 , further configured to:

receive the movement data from the one or more sensors;

detect a potential fall event based on the movement data;

determine which sub-group the subject is a member of; and

classify the detected potential fall event as a fall event or a non-fall event by comparing information associated with the fall event with the decision value for the sub-group the subject is determined to be a member of; and

a user interface configured to provide a feedback function, which, when activated, instructs the system for distinguishing between the fall event and the non-fall-event to re-classify a detected fall event as a non-fall event, and to update a false detection rate based on the re-classification of the event.