Algorithm for automatic positive air pressure titration

• US 20050241639A1
• Filed: 07/12/2005
• Published: 11/03/2005
• Est. Priority Date: 01/30/2003
• Status: Active Grant

• Current Assignee: Compumedics Medical Innovation Pty Limited
• Original Assignee: Compumedics, Inc.
• Inventors: Zilberg, Eugene
• Granted Patent: US 8,021,309 B2
• Time in Patent Office: Days
• US Class Current: 128/204.210
• CPC Class Codes: A61B 5/0803 Recording apparatus specially adapted therefor
• CPC Class Codes: A61B 5/0816 Measuring devices for examining respiratory frequency measuring frequency of electric signals G01R23/00
• CPC Class Codes: A61B 5/087 Measuring breath flow
• CPC Class Codes: A61B 5/4818 Sleep apnoea
• CPC Class Codes: A61M 16/0006 with means for creating vibrations in patients' airways
• CPC Class Codes: A61M 16/024 including calculation means, e.g. using a processor
• CPC Class Codes: A61M 2016/0036 in the breathing tube and used in both inspiratory and expiratory phase
• CPC Class Codes: A61M 2230/18 Rapid eye-movements [REM]
• CPC Class Codes: A61M 2230/40 Respiratory characteristics
• CPC Class Codes: A61M 2230/42 Rate
• CPC Class Codes: A61M 2230/62 Posture

Abstract

An algorithm for measuring the breathing of a patient for diagnosing breathing disorders and for controlling a breathing aid device for treating a patient. Sensors record pressure and flow rates of air delivered to a patient. The data is statistically manipulated to find the start of a breath, the end of a breath, the duration of a breath, the shape of a breath, the volume of a breath, the breathing rates, flow rates, snoring index, flattening index, and other useful data to diagnose and treat a patient. The data is used to diagnose the condition of a patient and to control a breathing aide device for treating a patient at the most optimal pressures and flow rates for the patient'"'"'s condition. Inspiration and expiration data are tracked separately and compared throughout the algorithm as a check the accuracy of the data manipulation by the algorithm.

Claims

1. A breath monitoring and treating method comprising:

collecting flow rate data of air applied to a patient,

compiling the flow rate data into inspiration arrays and expirations arrays by approximating the start and end times of the inspirations and expirations, and determining the maximum peak flow rates during inspiration and minimum peak flow rates during expiration from the flow rate data,

adjusting the start and end times of the inspirations and expirations in the arrays of flow rate data to best fit the flow rate data collected.

2. A breath monitoring and treating method as in claim 1 including the further step of:

removing noise from the array of flow rate data.

3. A breath monitoring and treating method as in claim 1 including the further step of:

collecting pressure data of the air applied to the patient.

4. A breath monitoring and treating method as in claim 1 including the further step of:

calculating leak compensation rates to subtract the leak rate from the measured rate to determine the flow rate of air applied to the patient.

5. A breath monitoring and treating method as in claim 1 including the further step of:

comparing inspiration flow rate data to expiration flow rate data as an accuracy check.

6. A breath monitoring and treating method as in claim 5 including the further step of:

comparing inspiration volume data to expiration volume data as an accuracy check.

7. A breath monitoring and treating method as in claim 5 including the further step of:

comparing inspiration end time data to expiration start time data and comparing inspiration start time data to expiration end time data as an accuracy check.

8. A breath monitoring and treating method as in claim 1 including the further step of:

compiling the flow rate data into breath shapes,

adjusting the start and end times for each inspiration and expiration from the breath shapes obtained from the flow rate data.

9. A breath monitoring and treating method as in claim 2 including the further step of:

using a linear smoothing technique for removing the noise.

10. A breath monitoring and treating method as in claim 1 wherein:

finding break points for approximating the start and end times of inspirations and expirations.

11. A breath monitoring and treating method as in claim 10 including the step of:

collecting raw data for inspirations, smoothing the raw data for the first inspiration in the array with a linear smoothing process and then attaching raw data collected for the next inspiration to the smoothed data and smoothing the raw data of the next inspiration with the linear smoothing process to compile an inspiration array of flow rate data and eliminate noise from the data and repeating the attaching and smoothing of new inspiration data collected until the array is completed to compile the flow data into inspiration arrays.

12. A breath monitoring and treating method as in claim 11 including the step of:

attaching the inspiration raw data to the array of smoothed data by matching the break point end of the inspiration of the previous breath to the break point of the raw data of the succeeding breath by a best fit method, adjusting the array accordingly and then linearly smoothing the raw data to form the inspiration array.

13. A breath monitoring and treating method as in claim 12 wherein the best fit method includes:

attaching the end of the inspiration break point of a first breath to a local minimum data point in the raw data of the next breath when the next breath raw data is available before the end of inspiration break point data is available and the next breath raw data starts at a lower value than the end of inspiration break point of the previous breath.

14. A breath monitoring and treating method as in claim 12 wherein the best fit method includes:

attaching the end of inspiration break point of a first breath to the start of the next breath inspiration break point when they have the same or approximately the same value and the calculated start of inspiration break point is available at approximately the same time as the calculated value of the end of inspiration break point of the previous inspiration.

15. A breath monitoring and treating method as in claim 12 wherein the best fit method includes:

attaching a point on the smoothed inspiration data as the inspiration break point to a local minimum data point in the raw data of the next breath at a point where the smoothed data and the raw data overlap when the next breath raw data is available before the end of inspiration break point data is available and the next breath raw data start of inspiration break point starts at a higher value than the end of inspiration break point.

16. A breath monitoring and treating method as in claim 12 wherein the best fit method includes:

attaching the end of inspiration break point of a first breath to the raw data of the next breath at a point where the raw data equals the end of the inspiration break point when the next breath raw data is available approximately simultaneously with the smoothed data such that they nearly overlap in time and the next breath raw data extends to a lower value than the end of the inspiration break point of the first breath.

17. A breath monitoring and treating method as in claim 12 wherein the best fit method includes:

attaching the end of inspiration break point of a first breath to the raw data of the next breath starting break point where there is a time gap between the first breath smoothed and the next breath starting break point raw data being available by filling in the gap using the slope of the linearly smoothed inspiration between the maximum and the end of breath break point as the slope of the line to fill in the gap, and moving the time of the next breath starting break point to intercept the line filling the gap.

18. A breath monitoring and treating method as in claim 12 including the step of:

checking the inspiration array for two maximum peaks within a specified time between the starting break point and the ending break point of the inspiration which indicates a double peak,

linearly smoothing the inspiration by selecting the maximum peak nearest the half duration of the inspiration as the true maximum peak and linearly smoothing the maximum peak to the starting break point and the ending break point.

19. A breath monitoring and treating method as in claim 18 including the step of:

identifying noisy maximum peaks in the array of inspiration peaks.

20. A breath monitoring and treating method as in claim 19 including the step of:

using a series of duration between local peaks tests and variations of flow rates between the peaks tests for determining which breaths fit within a predetermined threshold of parameters defining a noisy breath.

21. A breath monitoring and treating method as in claim 19 including the step of:

examining the upward slope of each inspiration in the array first to see if there is a noisy slope and if no noisy slope is found then checking the downward slope of each inspiration in the array to see if there is a noisy slope.

22. A breath monitoring and treating method as in claim 19 including the step of:

smoothing noisy maximum peaks found in the array by a least error selection made from a first order linear approximation, a second order liner approximation and a third order linear approximation.

23. A breath monitoring and treating method as in claim 22 including the step of:

comparing inspiration volumes to adjacent breath expiration volumes in an array of breaths for approximately the same volume of inspiration and expiration as a check on the accuracy of the data collected.

24. A breath monitoring and treating method as in claim 23 including the step of:

accepting the data as accurate if the inspiration volume approximately equals the adjacent breath expiration volume.

25. A breath monitoring and treating method as in claim 23including the step of:

rejecting the data as inaccurate if the inspiration volume does not approximately equal the adjacent breath expiration volume.

26. A breath monitoring and treating method as in claim 24 including the step of:

testing the raw data between two peaks in a maximum pair to see if any peaks are above a specified threshold which disqualifies the two peaks as a maximum pair.

27. A breath monitoring and treating method as in claim 26 including the step of:

testing the duration between maximum peaks for a predetermined range of values, if the duration is not within the predetermined range the maximum pair is disqualified.

28. A breath monitoring and treating method as in claim 27 including the step of:

graphing the variation of the flow rate in a maximum pair raw data set having a first maximum peak and a second maximum peak, and linearly smoothing the raw data showing the decreasing values as they decrease between the first maximum peak in the maximum pair and a minimum point, the graph showing no change during increasing variations, and showing the increasing values between the minimum point and the second maximum peak of the maximum pair, while showing no change during decreasing variations and noting the points at which the variations decrease to their 95 percentage points, their 75 percentage points, their 50 percentage points, their 25 percentage points, and their 5 percentage points, of the variation from the first maximum peak and noting the points at which the variations increase from their 5 percentage points, their 25 percentage points, their 50 percentage points, their 75 percentage points, their 95 percentage points, from the second maximum peak in the maximum pair.

29. A breath monitoring and treating method as in claim 28 including the step of:

compiling the differences in variations between the raw data and the linearly smoothed data at the 5%, 25%, 50%, 75% and 95% positions.

30. A breath monitoring and treating method as in claim 29 including the step of:

forming maximum trains out of maximum pairs using a similarity test on the maximum pairs to select which maximum pairs to add to the maximum train.

31. A breath monitoring and treating method as in claim 30 including the step of:

calculating the mean, standard deviation, average and maximum error for each of the following maximum pair parameters;

the duration, the variation in signal level in external duration, the average signal level in the internal duration signal, asymmetry measures of shape, signal ‘

shape’

in external duration, signal ‘

shape’

in internal duration, average noise level, and maximum noise level.

32. A breath monitoring and treating method as in claim 31 including the step of:

establishing mean index and error index values based on ranges of values on a predetermined table for adding maximum pairs to a maximum train.

33. A breath monitoring and treating method as in claim 30 including the step of:

verifying the maximum trains by comparing the inspiration maximum train starting break points to the expiration maximum train expiration ending break points and the inspiration maximum train ending break points to the expiration maximum train expiration beginning break points.

34. A breath monitoring and treating method as in claim 33 including the step of:

adding a maximum peak from the raw data if the inspiration and expiration break points do not match and the duration between maximum peaks is longer than a predefined threshold by left and right side maximum peak verification.

35. A breath monitoring and treating method as in claim 34 including the step of:

using flow rate signal level verification, to see if the signal level remains constant indicating the mask may have come off and the maximum pair must then be eliminated from the maximum train.

36. A breath monitoring and treating method as in claim 34 including the step of:

testing the duration between two maximum peaks and the volume of the inspirations as measured by the area under the maximum peaks, if the duration and volume are not within predetermined limits the maximum pairs are too close together and the smaller amplitude maximum peak is eliminated from the array.

37. A breath monitoring and treating method as in claim 34 including the step of:

testing the duration between two maximum peaks for durations that exceed predetermined limits and adding back maximum peaks to the array if the two tests are passed,

a 4 second duration test and,

a 10 second verification test.

38. A breath monitoring and treating method as in claim 1 including the step of:

shaping the inspiration and expiration data by a best fit method employing a piece wise linear approximation of the breath shape.

39. A breath monitoring and treating method as in claim 34 including the step of:

shaping the inspiration and expiration data by a best fit method employing a piece wise linear approximation of the breath shape.

40. A breath monitoring and treating method as in claim 38 including the step of:

using a third order piecewise linear approximation to determine the breath shape.

41. A breath monitoring and treating method as in claim 39 including the step of:

using a third order piecewise linear approximation to determine the breath shape.

42. A breath monitoring and treating method as in claim 40 including the step of:

fine tuning the start and end break points of the inspirations and expirations by a comparison of the relative changes of flow signal level at the start and end times of the inspirations to find the beginning of inspiration and the ending inspiration times.

43. A breath monitoring and treating method as in claim 41 including the step of:

comparing the relative changes of the flow signal level at the break points to find the beginning of inspiration and the ending inspiration times to fine tune the break points.

44. A breath monitoring and treating method as in claim 1 including the step of:

detecting reductions of inspiration flow rates in the inspiration arrays to identify apnea events.

45. A breath monitoring and treating method as in claim 1 including the step of:

detecting flattening of the inspiration flow rates in the inspiration arrays.

46. A breath monitoring and treating method as in claim 45 including the step of:

using a least square error method to find the smallest flattening error in the inspiration flow rate data to yield a flattening index.

47. A breath monitoring and treating method as in claim 46 including the step of:

using a best fit of a two dimensional searching method for detecting the flattening index.

48. A breath monitoring and treating method as in claim 47 including the step of:

using a straight line length of 60% of the inspiration interval for detecting the flattening index.

49. A breath monitoring and treating method as in claim 46 including the step of:

using the flattening index to determine the degree of blockage of the patient'"'"'s airway.

50. A breath monitoring and treating method as in claim 1 including the step of:

determining the inspiration flow amplitude from the flow rate data and using the inspiration flow amplitude to identify if the patient is experiencing apnea.

51. A breath monitoring and treating method as in claim 1 including the step of:

determining the inspiration flow amplitude from the flow rate data and using the inspiration flow amplitude to identify if the patient is experiencing hypopnea.

52. A breath monitoring and treating method as in claim 1 including the step of:

detecting reductions of inspiration flow rates to identify hypopnea events.

53. A breath monitoring and treating method as in claim 1 including the step of:

detecting snoring by measuring the amount of noise in terms of amplitude and frequency.

54. A breath monitoring and treating method as in claim 1 including the step of:

detecting snoring by use of a snoring index.

55. A breath monitoring and treating method as in claim 1 including the step of:

detecting a high frequency flow rate on top of the flow rate signal to identify snoring.

56. A breath monitoring and treating method as in claim 1 including the step of:

detecting flattening of the inspiration flow rates in the inspiration arrays and establishing a flattening index based on the degree of flattening to assign a value to the inspiration flow.

57. A breath monitoring and treating method as in claim 3 including the step of:

adjusting the pressure applied to the patient based on a flow rate data analysis.

58. A breath monitoring and treating method as in claim 57 including the step of:

using a snoring index to adjust the pressure applied to a patient.

59. A breath monitoring and treating method as in claim 57 including the step of:

using a flattening index to adjust the pressure applied to a patient.

60. A breath monitoring and treating method as in claim 57 including the step of:

changing the pressure applied to a patient within a predetermined range based on the current pressure applied to the patent, a flattening index of a reduction in the flow of air applied to a patient, and a snoring index of the degree of snoring.

61. A breath monitoring and treating method as in claim 60 including the step of:

changing the pressure applied to a patient based on the number of breaths measured in the flow rate data analysis.

62. A breath monitoring and treating method as in claim 57 including the step of:

determining an inspiration flow amplitude from the flow rate data of air applied to a patient,

changing the pressure applied to a patient within a predetermined range based on the current pressure applied to the patent and the inspiration flow amplitude.

63. A breath monitoring and treating method as in claim 62 including the step of:

changing the pressure applied to a patient based on the number of breaths measured in the flow rate data analysis.

64. A breath monitoring and treating method as in claim 62 including the step of:

using a leakage rate combined with the flattening index to determine the change in pressure applied to the patient.

65. A breath monitoring and treating method as in claim 3 including the step of:

changing the pressure applied to the patient based on the current pressure applied and the snoring index.

66. A breath monitoring and treating method as in claim 65 including the step of:

changing the pressure applied to the patient based on the flow level of air applied to the patient.

67. A breath monitoring and treating method as in claim 66 including the step of:

changing the pressure applied to the patient based on the flattening index.

68. A breath monitoring and treating method as in claim 50 including the step of:

measuring the air pressure applied to the patient, measuring the flow rate applied to the patient and changing the pressure applied to the patient according to a predetermined formula.

69. A breath monitoring and treating method as in claim 51 including the step of:

measuring the air pressure applied to the patient, measuring the flow rate applied to the patient and changing the pressure applied to the patient according to a predetermined formula.

70. A breath monitoring and treating method as in claim 3 including the step of:

reducing the pressure of air applied to a patient when normal breathing is established.

71. A breath monitoring and treating method as in claim 3 including the step of:

reducing the pressure of air applied to a patient when leak rates reach a predetermined level over a predetermined length of time.

72. A breath monitoring and treating method as in claim 3 including the step of:

making no changes of pressure applied to the patient when no changes are detected in the patient and when no rules for changing the pressure apply.

73. A breath monitoring and treating method as in claim 1 including the step of:

determining a flattening index for measuring patient inspiration flow limitations such as apneas and hypopneas.

74. A breath monitoring and treating method as in claim 1 including the step of:

determining a straight line flow approximation of the flow rate data between an inspiration start time and an inspiration end time by a least square error method,

determining the least error of the flow signal relative to the straight line flow approximation to calculate the flattening index.

75. A breath monitoring and treating method as in claim 74 including the step of:

selecting a set of flow signal data having an inspiration period length of 60% of the time between the start and end times of an inspiration, and testing for the least square error between the 5% and 95% of the inspiration period length to perform horizontal searching for a reference straight line flow approximation.

76. A breath monitoring and treating method as in claim 75 including the step of:

detecting the crossing points of the straight line flow approximation on the flow rate data and only using the period between the crossing points and the flow rate date from above the straight line flow approximation and below the flow rate data to calculate the errors and using a vertical line displacement of the straight line flow approximation line to find the least error.

77. A breath monitoring and treating method as in claim 75 including the step of:

calculating the flattening index by the error between the flow signal data and the flat signal level.