APPARATUS AND METHOD FOR THE DETECTION OF THE BODY POSITION WHILE SLEEPING

Abstract

The invention relates to a method and an apparatus for the detection of the body position, especially while sleeping. More particularly, the invention relates to how the main body positions during sleep can be derived from the distribution of the reflection of a projected IR light from a subject's body under a blanket. Additionally, the breathing signals can be analyzed to determine the body posture.

Claims

1. A method for detection of a posture of a body, the method comprising: providing a bedding; wherein the bedding lies upon a body; identifying a plurality of segments of the body; projecting a pattern of electromagnetic radiation at least on a part of the bedding; detecting a reflection pattern of the projected pattern caused by reflections from the bedding; comparing the reflection pattern with a plurality of predefined reflection patterns representing a corresponding plurality of body postures; and determining the posture of the body based on the comparison of the reflection pattern with the plurality of predefined reflection patterns; and recording the posture of the body over time for subsequent diagnoses; wherein the reflection pattern includes a plurality of reflection magnitudes; wherein each of the plurality of reflection magnitudes correspond to reflections from corresponding segments of the body; and wherein each of the predefined reflection patterns include a plurality of predefined reference magnitudes corresponding to each segment of the body at each corresponding body posture.

2. The method of claim 1, wherein the segments of the body include at least two of: upper body part, middle body part, and bottom body part.

3. The method of claim 1, wherein a light source for the projection of the pattern of electromagnetic radiation comprises a laser device; wherein the laser device is a solid state device; and wherein the laser device emits light in the infrared spectrum.

4. The method of claim 1, wherein the projecting of the pattern of electromagnetic radiation comprises projecting the pattern of electromagnetic radiation intermittently.

5. The method of claim 1, wherein the detecting of the reflection of the pattern of electromagnetic radiation is via a low-resolution sensor array.

6. The method of claim 1, wherein, in addition to the reflections from the bedding, one or more additional information items are considered for the determining of the body posture.

7. The method of claim 6, wherein the one or more additional information items include breathing amplitude information.

8. The method of claim 7, wherein the breathing amplitude information is based on images of the body from a video signal.

9. The method of claim 6, wherein the one or more additional information items include acoustical information that is retrieved via at least two microphones positioned on both sides of the bedding.

10. The method of claim 9, wherein the determining of the posture of the body is based on heuristics; wherein the heuristics include an orientation of one or more light sources that project the pattern of electromagnetic radiation; and wherein the heuristics include an orientation of one or more sensors that detect the reflection of the projected pattern.

11. The method of claim 6, wherein the one or more information items include movement information.

12. An apparatus comprising: a projector that projects a pattern of electromagnetic waves on a bedding; wherein the bedding lies upon a body; a detector that detects a reflection pattern of the pattern of electromagnetic waves; a data processor that: compares the reflection pattern to a plurality of predefined reflection patterns; determines the posture of the body based on the comparison of the reflection pattern with the plurality of predefined reflection patterns; and records the posture of the body over time for subsequent diagnoses; wherein the reflection pattern includes a plurality of reflection magnitudes; wherein each of the plurality of reflection magnitudes correspond to reflections from corresponding segments of the body; and wherein each of the predefined reflection patterns include a plurality of predefined reference magnitudes corresponding to each segment of the body at each corresponding body posture.

13. The apparatus of claim 12, further comprising at least one microphone connected to the data processor.

14. The apparatus of claim 12, wherein the projector comprises a laser device; wherein the laser device is a solid state device; and wherein the laser device emits light in the infrared spectrum.

15. The method of claim 2, wherein the upper body part comprises an upper left body part and an upper right body part; wherein the middle body part comprises a middle left body part and a middle right body part; and wherein the lower body part comprises a lower left body part and a lower right body part.

16. The method of claim 1, wherein the projecting of the pattern of electromagnetic radiation includes modulating the electromagnetic radiation.

17. The method of claim 6, wherein the one or more additional information items include acoustical information.

18. The method of claim 1, wherein each reflection magnitude and each reference reflection magnitude is a binary value.

19. The method of claim 1, comprising diagnosing a physical health condition based on the determined posture of the body.

20. The method of claim 1, comprising controlling an environmental situation of the body based on the determined posture of the body.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0045] These and other aspects of the invention will be apparent from and elucidated with reference to the embodiments described hereinafter.

[0046] In the drawings:

[0047] FIG. 1 shows the most common body postures for sleep (Dunkell, Samuel, Sleep Positions, 1977);

[0048] FIG. 2 shows the pattern reflection caused by different body postures at different resolution;

[0049] FIG. 3 shows a schematic illustration of the body posture detection according to an embodiment of the invention;

[0050] FIG. 4 shows the intensity distribution comparison of the reflection between different body postures with a light projector mounted on a wall at the lower end of the bedding;

[0051] FIG. 5 shows the intensity distribution comparison of the reflection between different body postures with a light projector mounted on the ceiling above the head of a subject on the bedding;

[0052] FIG. 6 shows the intensity distribution comparison of the reflection between different body postures with a light projector mounted on the left lower side of the bedding;

[0053] FIG. 7 shows the breathing amplitude comparison when lying on the back versus lying on the belly;

[0054] FIG. 8 shows cumulative audio events over a full-night recording of a subject;

[0055] FIG. 9 shows a left/right posture estimation over a full night;

[0056] FIG. 10 shows integrated actuators usable in combination with the inventive method;

[0057] FIG. 11 shows a bed lifting device usable in combination with the inventive method.

DETAILED DESCRIPTION OF EMBODIMENTS

[0058] While the invention has been illustrated and described in detail in the drawings and foregoing description, such illustration and description are to be considered illustrative or exemplary and not restrictive; the invention is not limited to the disclosed embodiments. Other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims. In the claims, the word comprising does not exclude other elements or steps, and the indefinite article a or an does not exclude a plurality. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage. Any reference signs in the claims should not be construed as limiting the scope.

[0059] FIG. 1 shows the most common body postures for sleep. Shown are the Prone, the Semi-Fetal, the Full Fetal, the Flamingo, the Sandwich, the Royal, the Cyclops, and the Water-Wings body posture. The Prone posture is lying face down with the arms extended over the head and the legs stretched out with the feet somewhat apart. The Semi-Fetal position is lying on the side with the knees drawn partway up. The Full Fetal position is lying in a folded position that obscures the face. The legs are flexed at the knees and the knees are drawn up. The Flamingo position is lying on the side with one leg straight out while the other leg is bent at the knee and flexed at a sharp angle. The Sandwich position is lying on the side with the legs placed precisely on top on one another, the thigh, knee and angle of the leg parallel to that of the other. The Royal posture is lying flat on the back. The Cyclops posture is lying flat on the back with one hand covering the eyes. In the Water-Wings posture the head rests in the palms of the hands with the elbows extended on either side.

[0060] FIG. 2 shows a pattern reflection caused by different body postures at different resolution. In row 1 and 2 body postures are shown where the higher intensities of reflection in the middle segments are visible. These are the postures when the subject lies on the side. In row 3 images of a high and low resolution are compared. The low resolution image is sufficient for discriminating side from flat lying postures. Another possibility is to analyze the reflected light segments indicating the orientation of the legs as shown in FIG. 2. From the leg orientation, the head orientation can be directly obtained.

[0061] FIG. 3 shows an illustration of the body posture detection according to an embodiment of the invention. In a video signal captured from a camera the distribution of the reflected light is detected. For discriminating similar reflection pattern of flat and side postures, the audio signal coming from two microphones on either side of the bedding are taken in consideration. Further, for discriminating a back from a belly posture the breathing amplitude is taken into consideration. By analyzing the distribution of the light reflectance, the flat or side body position can be identified. When a side body position is detected, the orientation of the face can be either determined by an additional audio signal and/or by the reflected intensity distribution in the lower leg area. Adding another modality renders the system more robust. When only information on the flat or side position is needed, the images can be captured with a low resolution optical sensor, e.g. an optical mouse sensor, e.g. with a resolution of 1919. Due to the reduced information needed by the algorithm, the higher resolution image from a camera is not needed. This is especially relevant to preserve the privacy of the sleeping person (see third row in FIG. 2). When the flat body position is detected, the breathing amplitude provides an indication on whether the subject is on the belly or on the back since the chest movement is more prominent when the sleeping subject is on his/her back. Heuristics can be derived for the main orientations of the camera/sensor and the light source(s) with regard to the bed (e.g., on the top, from the bottom side of the bed, from the left side of the bed, and/or from the right side of the bed). Automatically, the corresponding heuristics can be applied when the user inputs the estimated location of the camera/sensor and light source(s) with regard to the bed in a one-time installation.

[0062] FIGS. 4 to 6 show the intensity distribution comparison of the reflection between different body postures for light projector positions. In the Figs. the reflection is segmented into six areas, top left/right, middle left/right, and bottom left/right, however, the reflection can be segmented into as low as two segments. For some postures, a smaller number of segments is sufficient to determine body posture. It can also be envisioned to segment the reflection into a larger number than six segments. Additionally, the grids do not need to be rectangular in order to determine a subject's body posture while sleeping. In FIGS. 4 to 6 each posture gives a specific distribution of reflection between these six segments. The distribution varies with the position of the light projector and/or the sensor/camera detecting the reflection. The following heuristics are derived for three light source locations in the bedroom (the light source is always positioned higher than the bed). As a dividing line, the threshold is chosen as the mean of the whole intensity curve within one segment. This could also have been done differently; it serves only as an approximate indication so that high-low intensities can be distinguished. The x-axis coding reflects flat royal (on back, FR), flat prone (on belly, FP), side right (SR), side left (SL). In total, 71 body positions with 4 different test subjects were measured. A classification accuracy of 96% correct detections and 4% false detections is achieved.

[0063] Bottom (foot part, see FIG. 4): [0064] At least 1 middle segment high: Side [0065] Both middle segments low: Flat [0066] 1 middle segment low, 1 middle segment close to thresh: If 1 bottom segment high: Side, else Flat

[0067] Top (on ceiling above upper body, see FIG. 5: [0068] At least 1 middle segment low: Side [0069] Both middle segments high: Flat [0070] 1 middle segment close to thresh: If at least 1 top segment high: Side, else Flat Bottom left, see FIG. 6: [0071] At least 1 top segment high: Side [0072] Both top segments low: Flat [0073] 1 top segment low, 1 top segment close to thresh: If 1 middle segment high: Side, else Flat

[0074] FIG. 7 shows the breathing amplitude comparison when lying on the back versus lying on the belly. When a flat body position is detected on the basis of the reflected pattern, to distinguish on the back from on the belly, the respiration analysis output can be included. The breathing characteristics extracted from a video signal are different when the person lies on the belly compared to when the person lies on the back. When the person lies on the back the chest is free to move into open space without any large barrier blocking its movement. However, when the person is on the belly, the chest movement goes into the mattress and the amplitude perceived by the video is reduced. Empirically, a 25% higher breathing amplitude is measured when a subject is on his back. The decline in the breathing amplitude towards the end of the back sequence is assumed due to the more relaxed state of the subject with more shallow breathing (reduced air flow and chest expansion).

[0075] FIG. 8 shows cumulative audio events over a full-night recording of a subject. In this plot, one can see that for each microphone there are around 5000 event detections. Clearly, it can be seen that between 400 and 500 minutes, the posture is mainly toward the left microphone. In order to more clearly detect if the breathing person lies toward the left or the right microphone, one will look at the number of detected events in an epoch of 1 minute. First, a quality measure can be computed as follows:

[00001]$\frac{\#.Math.\mathrm{events}\left({\mathrm{mic}}_{\mathrm{left}}\right)-\#.Math.\mathrm{event}\left(\mathrm{mic\_right}\right)}{25}$

[0076] The division by 25 is chosen as one may assume that roughly 25 breaths are maximally possible during 1 minute. This measure is depicted for a full-night recording in FIG. 9. A low-pass filtering of this measure was performed to smooth the data. Finally, the low-pass filtered signal is compared with the mean level in order to detect the posture. As one can see, the posture between 400 and 500 minutes is mainly pointed toward the left microphone, 1 represents a left side posture, 1 represents a right side posture.

Example 1: Alleviation of Obtrusive Sleep Apnea (OSA)

[0077] In this embodiment, it is proposed a positional sleep apnea apparatus for monitoring the sleep position of a person, in an unobtrusive manner, comprising: [0078] hardware: a camera that makes use of reflected light, and a microphone; [0079] software/algorithms: for detecting the sleep position (lateral or supine), based on the images of the camera and the microphone output. Said algorithms can also include the amount of time the person is on the back or side, and the changes over the night, etc.

[0080] If OSA events are detected, the output unit can also relate these events (and the number of occurrences of the event) to the sleep positions during the night. For example:

[0081] Number of OSA events in supine position: 20

[0082] Number of OSA events in lateral position: 1.

[0083] This can be depicted visually in a graph or using text or other modality. If sleep quality or sleep depth is determined, the output unit can also relate the sleep quality or depth to the OSA events and to the body position during sleep. For example, in the lateral position, the subject's sleep quality was 30% higher than in supine position due to less OSA events. This can also be depicted visually in a graph or using text or other modality.

[0084] The device can also comprise one or both of the following: an actuator to help subjects who sleep on their backs to move to sleep on their side. This can be done using tactile stimulation by, e.g., a bed that automatically lifts up or down, vibrations in the bed, a t-shirt, mattress, or pillow. The system can also behave smarter by detecting the location of a subject in bed in order to trigger certain actuators for optimum and effective turning stimulation as shown in FIG. 10 and FIG. 11. For example, the subject can be stimulated to change the body posture to one in which less OSA events occur, like e.g. the lateral body posture.