Monitoring sleep using microactivity states

1. A method for monitoring sleep of a user using a wearable device, comprising:

determining an activity amount of the user based on multi-axial acceleration data received by the wearable device in a time period, wherein the multi-axial acceleration data comprises acceleration data in multiple axes, wherein determining the activity amount comprises:

obtaining a first distance vector using the multi-axial acceleration data, wherein the first distance vector has a first size M1;

obtaining second distance vectors from the first distance vector by rotating elements of the first distance vector, wherein obtaining an n th second distance vector of the second distance vectors from the first distance vector comprises rotating n elements of the first distance vector from a head to a tail in the n th second distance vector, and wherein each second distance vector has the first size M1;

obtaining third distance vectors from the first distance vector and the second distance vectors, wherein obtaining each third distance vector comprises extracting first M2 elements from a respective first distance vector and a second distance vector, wherein M2 is a positive number that is less than the first size M1;

obtaining an activity vector (AV) using the third distance vectors, wherein the activity vector has a form AV=[val 1 , val 2 , . . . , val M ]; and

determining the activity amount from the activity vector as (val 1 +val 2+ . . . +val M ) 2 /2;

based on a determination that the activity amount is smaller than a first threshold, determining a microactivity feature value using the multi-axial acceleration data and predetermined weights; and

determining a microactivity state for the user in the time period based on the microactivity feature value, wherein the activity amount of the user is substantially zero in the microactivity state.

2. The method of claim 1 , further comprising:

based on a determination that the activity amount is greater than or equal to the first threshold, determining the user is in a non-sleep state.

3. The method of claim 1 , wherein determining the microactivity feature value comprises:

determining axial feature values for the multiple axes, wherein each axial feature value is determined using weighted acceleration data in a respective axis of the multiple axes; and

determining the microactivity feature value by summing the axial feature values.

4. The method of claim 3 , wherein determining the axial feature values for the multiple axes comprises:

determining temporal segments for acceleration data in an axis of the multiple axes;

determining average values for the axis, wherein each average value is determined using acceleration data in a respective temporal segment of the axis; and

determining an axial feature value for the axis using the average values and the predetermined weights.

5. The method of claim 4 , further comprising:

determining the predetermined weights using the multi-axial acceleration data received by the wearable device in a predetermined time period and a statistics technique.

6. The method of claim 1 , wherein determining the microactivity state for the user in the time period based on the microactivity feature value comprises:

determining whether the microactivity feature value is greater than a second threshold and smaller than a third threshold;

based on a determination that the microactivity feature value is greater than the second threshold and smaller than the third threshold, determining that the user is in the microactivity state and a sleep state in the time period;

based on a determination that the microactivity feature value is smaller than the second threshold, determining that the user is not wearing the wearable device in the time period; and

based on a determination that the microactivity feature value is greater than the third threshold, determining that the user is in the microactivity state and a waking state in the time period.

7. The method of claim 6 , further comprising:

determining the second threshold and the third threshold based on multiple microactivity feature values determined using a first statistics technique when the user is wearing the wearable device and in the sleep state.

8. The method of claim 7 , further comprising:

updating the second threshold and the third threshold based on multiple microactivity feature values determined using a second statistics technique when a different user is wearing the wearable device.

9. An apparatus for monitoring sleep of a user, comprising:

a processor; and

a memory coupled to the processor, the memory configured to store instructions which when executed by the processor become operational with the processor to:

determine an activity amount of the user based on multi-axial acceleration data received by a wearable device in a time period, wherein the multi-axial acceleration data comprises acceleration data in multiple axes and the apparatus comprises the wearable device, wherein to determine the activity amount comprises to:

obtain a first distance vector using the multi-axial acceleration data;

obtain second distance vectors from the first distance vector by rotating elements of the first distance vector, wherein obtaining an n th second distance vector of the second distance vectors from the first distance vector comprises rotating n elements of the first distance vector from a head to a tail in the n th second distance vector;

obtain an activity vector (AV) using the second distance vectors, wherein the activity vector has a form AV=[val 1 , val 2 , . . . , val M ]; and

determine the activity amount from the activity vector as (val 1 +val 2+ . . . +val M ) 2 /2;

based on a determination that the activity amount is smaller than a first threshold, determine a microactivity feature value using the multi-axial acceleration data and predetermined weights; and

determine a microactivity state for the user in the time period based on the microactivity feature value, wherein the activity amount of the user is substantially zero in the microactivity state.

10. The apparatus of claim 9 , wherein the memory further comprises instructions when executed by the processor become operational with the processor to:

based on a determination that the activity amount is greater than or equal to the first threshold, determine the user is in a non-sleep state.

11. The apparatus of claim 9 , wherein the instructions operational with the processor to determine the microactivity feature value further comprise instructions to:

determine temporal segments for acceleration data in each axis of the multiple axes;

determine average values for each axis, wherein each average value is determined using acceleration data in a respective temporal segment of the axis;

determine an axial feature value for each axis using the average values and the predetermined weights; and

determine the microactivity feature value by summing axial feature values of the multiple axes.

12. The apparatus of claim 11 , wherein the memory further comprises instructions when executed by the processor become operational with the processor to:

determine the predetermined weights using the multi-axial acceleration data received by the wearable device in a predetermined time period and a statistics technique.

13. The apparatus of claim 9 , wherein the instructions operational with the processor to determine the microactivity state for the user in the time period based on the microactivity feature value further comprise instructions to:

determine whether the microactivity feature value is greater than a second threshold and smaller than a third threshold;

based on a determination that the microactivity feature value is greater than the second threshold and smaller than the third threshold, determine that the user is in the microactivity state and a sleep state in the time period;

based on a determination that the microactivity feature value is smaller than the second threshold, determine that the user is not wearing the wearable device in the time period; and

based on a determination that the microactivity feature value is greater than the third threshold, determine that the user is in the microactivity state and a waking state in the time period.

14. The apparatus of claim 13 , wherein the memory further comprises instructions when executed by the processor become operational with the processor to:

determine the second threshold and the third threshold based on multiple microactivity feature values determined using a first statistics technique when the user is wearing the wearable device and in the sleep state.

15. The apparatus of claim 14 , wherein the memory further comprises instructions when executed by the processor become operational with the processor to:

update the second threshold and the third threshold based on multiple microactivity feature values determined using a second statistics technique when a different user is wearing the wearable device.

16. A wearable device for monitoring sleep of a user, comprising:

an accelerometer having multiple axes;

a processor; and

a memory coupled to the processor, the memory configured to store instructions which when executed by the processor become operational with the processor to:

determine an activity amount of the user based on multi-axial acceleration data in the multiple axes received by the accelerometer in a time period, wherein to determine the activity amount comprises to:

obtain a first distance vector using the multi-axial acceleration data, wherein the first distance vector has a first size M1;

obtain second distance vectors from the first distance vector by rotating elements of the first distance vector in a tail-to-head manner, wherein each second distance vector has the first size M1;

obtain third distance vectors from the first distance vector and the second distance vectors, wherein obtaining each third distance vector comprises extracting first M2 elements from a respective first distance vector or a second distance vector, wherein M2 is a positive number that is less than the first size M1;

obtain an activity vector using the third distance vectors; and

determine the activity amount by operations including summing all elements of the activity vector and squaring the sum;

based on a determination that the activity amount is smaller than a first threshold, determine a microactivity feature value using the multi-axial acceleration data and predetermined weights; and

determine a microactivity state for the user in the time period based on the microactivity feature value, wherein the activity amount of the user is substantially zero in the microactivity state, and wherein

the user is in the microactivity state and a sleep state when the microactivity feature value is greater than a second threshold and smaller than a third threshold,

the user is not wearing the wearable device when the microactivity feature value is smaller than the second threshold, and

the user is in the microactivity state and a waking state when the microactivity feature value is greater than the third threshold.

17. The wearable device of claim 16 , wherein the memory further comprises instructions when executed by the processor become operational with the processor to:

determine the predetermined weights using the multi-axial acceleration data received by the accelerometer in a predetermined time period and a statistics technique.

18. The wearable device of claim 16 , wherein the memory further comprises instructions when executed by the processor become operational with the processor to:

determine the second threshold and the third threshold based on multiple microactivity feature values determined using a first statistics technique when the user is wearing the wearable device and in the sleep state; or

update the second threshold and the third threshold based on multiple microactivity feature values determined using a second statistics technique when a different user is wearing the wearable device.