MONITORING THE OPERATION OF RESPIRATORY SYSTEMS

Abstract

There is provided a method of detecting a fault in a breathing system. The method comprises the steps of (a) taking a series of measurements of a first parameter of the breathing system; and (b) setting a fault boundary for the first parameter, the fault boundary being dependent on a plurality of the measurements of the first parameter. The method further includes at least one update procedure comprising the steps of (c) taking one or more further measurements of the first parameter; and (d) updating the fault boundary, the updated fault boundary being dependent on an updated set of measurements of the first parameter, the updated set of measurements of the first parameter including at least one of the further measurements of the first parameter.

Claims

1-42. (canceled)

43. A respiratory apparatus for supplying gas to a breathing system, the respiratory apparatus comprising: at least one sensor configured to take a series of measurements of a first parameter of the breathing system and configured, in at least one update procedure, to take one or more further measurements of the first parameter; a controller configured to set a fault boundary for the first parameter, the fault boundary being dependent on a plurality of the measurements of the first parameter, and configured, in the at least one update procedure, to update the fault boundary, the updated fault boundary dependent on an updated set of measurements of the first parameter, the updated set of measurements of the first parameter including at least one of the further measurements of the first parameter.

44. The respiratory apparatus according to claim 43, wherein the controller is configured to set the fault boundary for the first parameter to be offset from a measurement parameter that is dependent on a plurality of the measurements of the first parameter.

45. The respiratory apparatus according to claim 44, wherein the offset of the fault boundary from the measurement parameter is a predetermined amount or an amount that is dependent on a variance factor of a plurality of the measurements.

46. The respiratory apparatus according to claim 44, wherein after taking one or more further measurements of the first parameter, the controller is configured to update the measurement parameter, the updated measurement parameter being dependent on at least the one or more further measurements of the first parameter.

47. The respiratory apparatus according to claim 46, wherein the controller is configured to set the updated fault boundary to be offset from the updated measurement parameter.

48. The method according to claim 47, wherein the offset of the fault boundary from the measurement parameter is an amount that is dependent on a variance factor of a plurality of the measurements, and after the one or more further measurements of the first parameter are taken, the controller is configured to update the variance factor, the updated variance factor being dependent on the updated set of measurements of the first parameter, and the offset of the updated fault boundary from the updated measurement parameter is an amount that is dependent on the updated variance factor.

49. The respiratory apparatus according to claim 48, wherein the updated variance factor is dependent on at least one of the further measurements of the first parameter.

50. The respiratory apparatus according to claim 44, wherein the measurement parameter is an average of a plurality of the measurements and/or wherein the updated measurement parameter is an average of a plurality of the updated set of measurements.

51. The respiratory apparatus according to claim 43, wherein the updated set of measurements comprises at least one of the measurements upon which the fault boundary being updated is dependent and at least one of the further measurements, and wherein the at least one of the further measurements replaces an equivalent number of the earliest measurements upon which the fault boundary being updated is dependent.

52. The respiratory apparatus according to claim 43, wherein the controller is further configured to compare a measurement of the first parameter with either the fault boundary, or once the fault boundary has been updated, the updated fault boundary, and determine a fault with the operation of the breathing system in response to the measurement being outside of the fault boundary or the updated fault boundary.

53. The respiratory apparatus according to claim 52, wherein in response to determining a fault with the operation of the breathing system, the controller is configured to indicate the fault to a user.

54. The respiratory apparatus according to claim 43, wherein the sensor and/or the controller are configured to repeat the update procedure at least once, wherein the fault boundary being updated in each subsequent update procedure is the updated fault boundary from the previous update procedure.

55. The respiratory apparatus according to claim 54, wherein the sensor and/or the controller are configured to repeat the update procedure at intervals during operation of the breathing system, including at least a period during which gas is supplied to the breathing system, and/or the period during which treatment is being provided to a patient, wherein at least one of the intervals is less than 10 seconds.

56. A respiratory apparatus according to claim 43, wherein the at least one sensor is configured to take the series of measurements of the first parameter of the breathing system during operation of the breathing system, and/or the controller is configured to set the fault boundary for the first parameter during operation of the breathing system.

57. A respiratory apparatus according to claim 43, wherein the respiratory apparatus is configured to commence supplying gas to the breathing system before the at least one sensor takes the series of measurements of the first parameter of the breathing system.

58. A respiratory system comprising the respiratory apparatus of claim 43 and a breathing system in fluid communication with the respiratory apparatus, the breathing system comprising a patient interface arranged to supply gas to a patient in use, and a breathing tube arranged to deliver the gas supply from the respiratory apparatus to the patient interface.

59. A method of detecting a fault in a breathing system, the method comprising the steps of: (a) taking a series of measurements of a first parameter of the breathing system; and (b) setting a fault boundary for the first parameter, the fault boundary being dependent on a plurality of the measurements of the first parameter; wherein the method includes at least one update procedure comprising the steps of: (c) taking one or more further measurements of the first parameter; and (d) updating the fault boundary, the updated fault boundary being dependent on an updated set of measurements of the first parameter, the updated set of measurements of the first parameter including at least one of the further measurements of the first parameter.

60. A method of detecting a fault in a breathing system, the method comprising the steps of: (a) supplying a gas to the breathing system; (b) taking a series of measurements of a first parameter of the breathing system; (c) setting a fault boundary for the first parameter, the fault boundary being dependent on a plurality of the measurements of the first parameter, wherein a second parameter of the breathing system is controlled during supply of the gas to the breathing system, and the first parameter of the breathing system is dependent on the second parameter and the breathing system, such that a change in the first parameter may be indicative of a fault in the breathing system.

61. A method as claimed in claim 60, wherein the second parameter is flow rate or pressure within the device or the breathing system.

62. A method as claimed in claim 60, wherein the first parameter is any of flow rate, back pressure within the breathing system, or patient pressure.

Description

[0094] Practicable embodiments of the invention will now be described, by way of example only, with reference to the accompanying drawings, of which:

[0095] FIG. 1 is a flow diagram illustrating a method of dynamically updating the fault boundaries for a gas parameter in a constant flow therapy device and testing for a fault according to a first embodiment;

[0096] FIG. 2 is a flow diagram illustrating a method of dynamically updating the fault boundaries for a gas parameter in a constant flow therapy device and testing for a fault according to a second embodiment;

[0097] FIG. 3 is a flow diagram illustrating initialisation of the parameters associated with the method of FIG. 2.

[0098] FIG. 4 is a first worked example of the setting of fault boundaries for a gas parameter in a constant flow therapy device;

[0099] FIG. 5 is a second worked example of the setting of fault boundaries for a gas parameter in a constant flow therapy device;

[0100] FIG. 6 is a third worked example of the setting of fault boundaries for a gas parameter in a constant flow therapy device; and

[0101] FIG. 1 illustrates a method by which the fault boundaries for a gas parameter in a constant flow therapy device are dynamically updated during operation of the device according to a first embodiment, in order to determine whether there is an error in the operation of the device and then testing for a fault.

[0102] Steps 100-120, indicated by the dashed box of FIG. 1, illustrate an initialisation process by which the long-term average of the sensor is determined.

[0103] At step 100 it is determined whether the device is in a start-up mode. The start-up mode is defined as being when the device has just been turned on, when the treatment provided by the device has just been changed, or when the parameters of the device have just been altered by an operator. When the device is in the start-up mode, the initialisation process indicated by the dashed box runs as a continuous loop, for 45 seconds. It will be appreciated that the exact value of this time period is not essential, and may vary, within reason.

[0104] If at step 100 it is determined that the device is in the start-up mode, the method progresses to step 120, where a measurement of a parameter is taken and the long-term average of that parameter (ie over the 45 second period) is updated.

[0105] The measured parameter is one that varies as a result of another parameter that is predefined by a user, ie as a result of using a different patient interface or as a result of variations in the therapy delivered. The measured parameter is also one that varies as a result of faults in a breathing system to which the device is connected. In the following embodiments, this measured parameter will be exemplified as pressure at the device. The pressure at the device is known to be dependent on the back pressure within the breathing system to which the device is connected, and it is known that the back pressure within the breathing system is dependent on any faults in that breathing system. Hence, a change in the pressure at the device indicates a change in the back pressure within the breathing system, which if large enough, may indicate a fault in the breathing system.

[0106] Once the long-term average has been updated, the method progresses to step 130. At step 130, it is determined whether the device is still in the start-up mode. If yes, the method returns to the start to continue with the initialisation process.

[0107] Once the 45 second initialisation process has been undertaken, at step 130 it will be determined that the device is no longer in the start-up mode, and the method will progress to steps 140 and 150 simultaneously. At steps 140 and 150, an upper fault boundary and a lower fault boundary are defined. These are defined as being 20% higher than the long-term average pressure and 20% lower than the long-term average pressure respectively. It will be appreciated that the exact value of these percentages are not essential, and may vary, within reason.

[0108] At step 160, the offset of the lower fault boundary from the long-term average defined in step 150 is compared with a predetermined minimum offset. The predetermined minimum offset is associated with the measuring accuracy of the sensor measuring the pressure at the device. That is, the predetermined minimum offset equates to the minimum difference required between any two measurements for the sensor to distinguish the two values for certain, based on its measuring accuracy. If the lower fault boundary offset is lower than the predetermined minimum offset assigned to the sensor, then at step 170, this is adjusted. This ensures that the upper fault boundary and the lower boundary do not become too narrow about the long-term average, which could lead to a large number of measurements being falsely determined as error measurements, particularly where the precision or accuracy of the sensor is relatively low.

[0109] Where it is determined that the offset of the lower fault boundary from the long-term average is lower than the predetermined minimum offset, both the offset of the lower fault boundary and the higher fault boundary are increased to the predetermined minimum offset. As an example, if the long-term average at step 120 is 3, then the 20% fault boundaries will result in an upper boundary of 3.6 and a lower boundary of 2.4. The fault boundary offset of 0.6 will be compared with a predetermined minimum offset assigned to the sensor. If the predetermined minimum offset assigned to the sensor is 2 (ie the sensor is accurate to ±2), then the lower fault boundary offset and the higher fault boundary offset will be increased to 2, such that the lower fault boundary becomes 1 (ie 3−2) and the higher fault boundary becomes 5 (ie 3+2). It will be appreciated that in other examples, only one of the fault boundaries may be offset in response to the offset of the lower fault boundary from the long-term average being lower than the predetermined minimum offset

[0110] If the lower fault boundary offset is higher than the predetermined minimum offset assigned the sensor, then the method continues and progresses to step 180.

[0111] A maximum offset may also be defined in a similar manner such that the fault boundaries do not become too wide about the long-term average.

[0112] At step 180, the sensor takes a pressure measurement. This measurement is compared with the defined fault boundaries at step 190. If the measured pressure is found to be outside of those fault boundaries, then an adaptive alarm fault is raised at step 200 and an alarm signal is raised to signal to an operator that there is a fault in operation of the device. The alarm signal is a combination of visual and sound signals that alerts an operator that there is a fault with the configuration of the device. If the measured pressure is found to be inside of those fault boundaries, then no adaptive alarm fault is raised at step 210. This represents the end of the algorithm. In either case, the algorithm ends at this point (220) and restarts at step 100.

[0113] Since the device is no longer in the start-up mode, at step 100 the method will move to step 110. At step 110, if it was determined in the previous method loop that there was an adaptive alarm fault, then the pressure measured during the previous method loop will not be used to update the long-term average of the pressure. Instead, the method will move straight to step 130 and determine again whether there is still a fault in operation of the device, or if the fault has been solved.

[0114] If It was determined in the previous method loop that there was not an adaptive alarm fault, then the pressure measured during the previous method loop will be used to update the long-term average of the pressure at step 120. The method will then move on to step 130, to update the upper and lower limits at steps 140 and 150 based on the updated long-term average of the pressure, and to measure and monitor the pressure again at steps 180-210.

[0115] Steps 100 to 220 are repeated as a continuous loop until interrupted by turning off of the device, changing of any attached equipment that requires a resetting of the method, changing of parameters implemented in the treatment, an initiation of a change in the treatment carried out by the device, or a user acknowledging an existing alarm fault signal. That is, in response to any of these events, the initialisation process will need to be repeated for 45 seconds once more.

[0116] FIG. 2 illustrates a method by which the fault boundaries for a gas parameter in a constant flow therapy device are dynamically updated during operation of the device according to a second embodiment, in order to determine whether there is an error in the operation of the device and then testing for a fault. It is worth noting that a long-term average of the sensor and an average deviation of the sensor are firstly determined by an initialisation process which is described in relation to FIG. 3 below.

[0117] Although illustrated differently, for simplicity of explanation, the method will be described here as beginning at step 340, since in reality the method is a continuous loop.

[0118] At step 340, an average reading is taken from a plurality of sensors measuring the same parameter. At step 350, it is calculated how far the average reading taken at step 340 deviates from the long-term average of the sensor, in order to determine a short-term deviation of the sensors. It is then determined whether the average reading taken at step 340 falls outside fault boundaries which are set as being three times the average deviation from the long-term average of the sensor. That is, it is determined whether the average reading taken at step 340 is greater than a first fault boundary or lower than a second fault boundary.

[0119] As with the method described in relation to FIG. 1, the method of this embodiment also employs the same compensating technique if it is determined that the fault boundaries resulting from three times the average deviation from the long-term average are too narrow. This shall not be repeated here for conciseness.

[0120] If the average reading taken at step 340 falls within the fault boundaries, then the method moves on to step 360, and the test is passed. That is, it is determined that there is no fault in operation. If the average reading taken at step 340 falls outside the fault boundaries, then the method moves on to step 370, and the test is failed. That is, it is determined that there is a fault in operation.

[0121] In either case, the method reverts to step 300 to perform another iteration of the method. At step 300, it is determined whether there is a currently a fault in the operation signalled by an alarm. If so, the method moves on to step 340, which allows updating of the long-term average of the sensor and the average deviation of the sensor to be skipped. This is implemented so that the long-term average of the sensor and the average deviation of the sensor, which are used for determining whether a fault occurs, are not skewed by the addition of a sensor value that itself has led to an error alarm signal.

[0122] Where there is no current alarm signal, the method moves on to step 310. At step 310 it is determined whether a second has elapsed since the long-term average of the sensor and the average deviation of the sensor were updated. This allows multiple sensor readings to be taken per second without using the processing power of updating the long-term average of the sensor and the average deviation of the sensor at every reading. Where one second has not elapsed, the method performs steps 340-370 again as previously described. It will be appreciated that in alternative embodiments, the long-term average of the sensor and the average deviation of the sensor may be updated each time a sensor reading is taken.

[0123] Where one second has elapsed since the long-term average of the sensor and the average deviation of the sensor were updated, the method proceeds to step 320 where the long-term average of the sensor is updated. Here, an average of the short-term average readings taken in the one second period since the long-term average was last updated is added to the long-term average of the sensor. The long-term average of the sensor comprises 32 readings which are made up of the previous 32 short-term averages, and the latest value replaces the oldest value of those 32 readings.

[0124] The method then proceeds to step 330, an average of the short-term deviations calculated in the one second period since the average deviation was last updated is added to the average deviation of the sensor. The average deviation of the sensor comprises 64 readings which are made up of the previous 64 averaged short-term deviations, and the latest value replaces the oldest value of those 64 readings.

[0125] Steps 300 to 370 are repeated as a continuous loop until interrupted by turning off of the device, changing of any attached equipment that requires a resetting of the method, changing of parameters implemented in the treatment, a change in the treatment carried out by the device is initiated, or user acknowledgement of an existing alarm fault signal. That is, in response to any of these events, the initialisation process illustrated in FIG. 3 will need to be repeated for 45 seconds once more.

[0126] At step 370, where it is determined that there is a fault in operation, an alarm signal is produced. The alarm signal is a combination of visual and sound signals that alerts an operator that there is a fault with the configuration of the device. FIG. 3 illustrates the initialisation process by which the long-term average of the sensor and the average deviation of the sensor are determined before the continuous method of FIG. 2 is implemented.

[0127] At step 400, an operator defines a second parameter, in this case the flow rate through the device, via a user interface associated with the device. At step 410, measurements of a first parameter, in this case the pressure at the device, are taken at 1 second intervals over a 45 second period by a pressure sensor positioned within the device. An average of the pressure readings taken is calculated and is set as a reference level for the pressure through the breathing system, the reference level being the long-term average of the sensor referred to in step 320.

[0128] Simultaneously, at step 420, the deviation of the pressure measurements taken over that 45-second period is determined relative to the average pressure calculated at step 410. The deviation of the measurements taken over that 45-second period may be an average of the deviation of each reading taken over that 45-second period (ie an average deviation of all the readings compared with the average reference level determined at step 410), or a maximum deviation from the reference level experienced over the duration of the 45-second period.

[0129] At step 430, the fault boundaries are defined as a function of the deviation of the pressure determined at step 420. Two fault boundaries are defined, a first fault boundary greater than the average pressure calculated in step 410 and a second fault boundary lower than the average pressure calculated in step 410. Worked examples of the calculations for defining the error parameter boundaries will be shown in more detail below.

[0130] In alternative embodiments, it is anticipated that alternative parameters may be set at step 400 and measured at steps 410, 420 and 340. In one example, the pressure may be set at step 400 and the flow rate may be measured at steps 410, 420 and 340.

[0131] In alternative embodiments, it is anticipated that the fault with the configuration of the device may be solved by the device itself, in which case the alarm signal may be an internal signal sent from a processor of the device to a controller of the device.

[0132] In alternative embodiments, only one fault boundary may be set at step 430. In this instance, the fault boundary may be greater or lower than the average pressure determined at step 410.

[0133] In alternative embodiments, in response to determining that there is a fault in operation of the device at step 370, operation of the device may cease until an operator has fixed the fault.

[0134] In alternative embodiments, the deviation recorded as a reference at step 330 may be a maximum deviation from the long-term average of the sensor.

[0135] Worked examples of steps 410 to 430 of FIG. 4 are presented in FIGS. 4, 5 and 6. For these examples, the values are given in arbitrary units for the sake of simplicity.

[0136] According to a first example, illustrated in FIG. 4, the pressure is measured at step 410 at 1 second intervals for 45 seconds, and the average pressure is determined to be 20. The pressure long-term average is therefore set as 20. The average deviation of the pressure measurements from the long-term average over that 45-second period is 2.4.

[0137] The fault boundaries in this example are set as an offset of three times the average deviation from the long-term average. The upper boundary is therefore set as 27.2, +7.2 from the long-term average, and the lower boundary is set as 12.8, −7.2 from the long-term average.

[0138] In the second and third examples, illustrated in FIGS. 5 and 6, rather than using the average deviation of the pressure measurements from the long-term average to calculate the upper and lower boundaries, a maximum deviation is used. Although not illustrated in FIG. 5 or 6, this may require the upper and lower boundaries to be offset by a lesser multiple of the calculated deviations.

[0139] In FIG. 5, the pressure is measured at step 410 for 45 seconds, and the average pressure is determined to be 20. The long-term average of the sensor is therefore set as 20. The pressure measurements fluctuate over that 45-second period between values of 18 and 23. The maximum deviation of the pressure measurements from the long-term average is therefore recorded as 3.

[0140] The fault boundaries in this example are set as an offset of three times the maximum deviation from the long-term average. The upper boundary is therefore set as 29, +9 from the long-term average, and the lower boundary is set as 11, −9 from the long-term average.

[0141] In FIG. 6, the pressure is measured at step 410 for 45 seconds, and the average pressure is determined to be 20. The pressure long-term average is therefore set as 20. The pressure measurements fluctuate over that 45-second period between values of 18 and 23. In this example, the maximum deviation of the pressure measurements from the long-term average is recorded as 2 in the negative direction and 3 in the positive direction.

[0142] The fault boundaries in this example are set as an offset of three times the maximum deviation in that specific direction from the long-term average. The upper boundary is therefore set as 29, +9 (3×3) from the long-term average, and the lower boundary is set as 14, ×6 (3×−2) from the long-term average.