Detection of periodic breathing during CPAP therapy

1. A breathing therapy device comprising:

a blower;

a processor;

a flow sensing element, for sensing total flow of air and

said processor configured to:

collect flow data from said flow sensing element;

analyze said flow data to detect patterns of periodic breathing in a user of said device;

rate the severity of said periodic breathing, wherein rating the severity of said periodic breathing comprises configuring said processor to:

calculate a power envelope of a normalized amplitude signal representing waxing and waning cycles of said patient flow data;

calculate a signal representing a zero crossing variance of said normalized amplitude signal; and

classify the severity of said periodic breathing using said power envelope of said normalized amplitude signal and said signal representing the zero crossing variance; and

alter therapy delivered to said user based on said rated severity of said periodic breathing.

2. The breathing therapy device of claim 1 wherein said normalized amplitude signal is

a ratio of a fast moving average and a slow moving average of a power envelope of a patient flow signal derived from said flow data.

3. The breathing therapy device of claim 2 wherein said power envelope of said patient flow signal and said power envelope of said normalized amplitude are calculated using Hilbert Transforms.

4. The breathing therapy device of claim 2 wherein in order to derive said patient flow signal from said flow data said processor is further configured to:

down sample said flow data received from said flow sensing element; and

apply a high pass filter to filter out a DC component of said flow data, representing bias flow.

5. The breathing therapy device of claim 4 wherein said high pass filter is a 2nd order Butterworth filter.

6. The breathing therapy device of claim 2 wherein, in order to rate the severity of said periodic breathing, said processor is further configured to:

define ranges for said power envelope of said normalized amplitude signal and for said signal representing the zero crossing variance of said normalized amplitude signal for each category of severity;

calculate said rating as a function of time based on a membership of said power envelope of said normalized amplitude signal and said signal representing the zero crossing variance of said normalized amplitude signal in each of said categories.

7. The breathing therapy device of claim 6 wherein said membership in said categories is calculated using a statistical analysis.

8. The breathing therapy device of claim 7 wherein said statistical analysis utilizes a minimum distance estimation algorithm.

9. The breathing therapy device of claim 2 wherein altering therapy delivered to said user includes ceasing an increase in pressure when apnea events are detected.

10. The breathing therapy device of claim 2 wherein altering therapy delivered to said user includes decreasing pressure when sustained periods of periodic breathing are detected.