Fault diagnosis in CPAP and NIPPV devices

Abstract

A ventilation device for non-invasive positive pressure ventilation (NIPPV) or continuous positive airway pressure (CPAP) treatment of a patient has a gas flow generator, a gas delivery circuit optionally including a humidifier, a controller and sensors monitoring values of operational parameters of the device. The device further includes one or more relationships stored in data storage of the controller relating combinations of parameter values as being indicative of fault conditions of the device operation, the sensors and/or the fault detection process.

Claims

1. A method of detecting a fault in an apparatus for supplying breathable gas to a patient at a positive pressure in connection with the treatment of sleep-disordered breathing, the apparatus including a gas flow generator comprising a servo controlled motor and impeller, the method comprising: monitoring first output from a breath detection module, the first output indicating whether the patient is breathing and being determined based at least in part on a patient respiratory flow signal calculated from parameters gathered from the apparatus; monitoring second output from an apnea detection module, the second output indicating whether the patient is experiencing an apneic event and also being determined based at least in part on the patient respiratory flow signal; determining whether the first output and the second output indicate that the patient is breathing and experiencing an apneic event at the same time; determining whether the first output and the second output indicate that the patient has neither breathed nor experienced an apneic event for a predetermined time period; and generating an alert of a likely fault in the apparatus (a) in response to a determination that the first output and the second output indicate that the patient is breathing and experiencing an apneic event at the same time, and (b) in response to a determination that the first output and the second output indicate that the patient has neither breathed nor experienced an apneic event for a predetermined time period.

2. The method of claim 1, wherein generated alerts are alerts indicative of likely software faults.

3. The method of claim 1, wherein the breath detection and apnea detection modules are software modules residing on and executed under control of the apparatus.

4. The method of claim 1, wherein the apparatus further comprises one or more of a motor speed transducer configured to produce a motor speed transducer signal, a pressure transducer configured to produce a pressure signal indicative of the pressure generated by the apparatus, and a flow transducer configured to produce a gas flow signal.

5. The method of claim 4, wherein the respiratory flow signal is calculated based at least in part on the motor speed transducer signal, the pressure signal, and/or the gas flow signal.

6. An apparatus for supplying breathable gas to a patient at a positive pressure in connection with the treatment of sleep-disordered breathing, comprising: a gas flow generator comprising a servo controlled motor and impeller; at least one sensor configured to monitor operating parameters of the apparatus; a controller configured to at least: selectively assert fault conditions based on analyses of monitored operating parameters; determine that a fault likely exists based at least in part on simultaneous assertion of two or more mutually exclusive fault conditions; and generate an alert responsive to the determination.

7. The apparatus of claim 6, further comprising: a breath detection module configured to generate first output indicating whether the patient is breathing based at least in part on a patient respiratory flow signal calculated from parameters gathered from the apparatus; and an apnea detection module configured to generate second output indicating whether the patient is experiencing an apneic event based at least in part on the patient respiratory flow signal; wherein the controller is further configured to at least: monitor the first output from the breath detection module; monitor the second output from the apnea detection module; and determine that the fault likely exists responsive to the first output and the second output indicating that the patient is breathing and experiencing an apneic event at the same time.

8. The apparatus of claim 6, further comprising: a breath detection module configured to generate first output indicating whether the patient is breathing based at least in part on a patient respiratory flow signal calculated from parameters gathered from the apparatus; and an apnea detection module configured to generate second output indicating whether the patient is experiencing an apneic event based at least in part on the patient respiratory flow signal; wherein the controller is further configured to at least: monitor the first output from the breath detection module; monitor the second output from the apnea detection module; and determine that the fault likely exists responsive to the first output and the second output indicating that the patient has neither breathed nor experienced an apneic event for a predetermined time period.

9. The apparatus of claim 6, wherein the generated alert is indicative of a likely software fault in a module associated with generation of first output indicating whether the patient is breathing and/or second output indicating whether the patient is experiencing an apneic event.

10. The apparatus of claim 6, wherein a respiratory flow signal is calculated based at least in part on parameters gathered from a motor speed transducer signal generated by a motor speed transducer of the apparatus, a pressure signal indicative of the pressure generated by the apparatus generated by a pressure transducer of the apparatus, and/or a gas flow signal generated by a flow transducer of the apparatus.

11. An apparatus for supplying breathable gas to a patient at a positive pressure in connection with the treatment of sleep-disordered breathing, comprising: a gas flow generator comprising a servo controlled motor and impeller; a breath detection module configured to generate first output indicating whether the patient is breathing based at least in part on a patient respiratory flow signal calculated from parameters gathered from the apparatus; an apnea detection module configured to generate second output indicating whether the patient is experiencing an apneic event based at least in part on the patient respiratory flow signal; and a controller configured to at least: monitor the first output from the breath detection module; monitor the second output from the apnea detection module; determine whether the first output and the second output indicate that the patient is breathing and experiencing an apneic event at the same time; determine whether the first output and the second output indicate that the patient has neither breathed nor experienced an apneic event for a predetermined time period; and generate an alert of a likely fault in the apparatus (a) in response to a determination that the first output and the second output indicate that the patient is breathing and experiencing an apneic event at the same time, and (b) in response to a determination that the first output and the second output indicate that the patient has neither breathed nor experienced an apneic event for a predetermined time period.

12. The apparatus of claim 11, wherein generated alerts are alerts indicative of likely software faults.

13. The apparatus of claim 11, wherein the breath detection and apnea detection modules are software modules residing on and executed under control of the apparatus.

14. The apparatus of claim 11, further comprising at least one of: a motor speed transducer configured to produce a motor speed transducer signal; a pressure transducer configured to produce a pressure signal indicative of the pressure generated by the apparatus; and a flow transducer configured to produce a gas flow signal.

15. The apparatus of claim 14, wherein the controller is further configured to calculate the respiratory flow signal based at least in part on the motor speed transducer signal, the pressure signal, and/or the gas flow signal.

16. An apparatus for supplying breathable gas to a patient at a positive pressure in connection with the treatment of sleep-disordered breathing, comprising: a gas flow generator comprising a servo controlled motor and impeller; a breath detection module configured to generate first output indicating whether the patient is breathing based at least in part on a patient respiratory flow signal calculated from parameters gathered from the apparatus; an apnea detection module configured to generate second output indicating whether the patient is experiencing an apneic event based at least in part on the patient respiratory flow signal; and a controller configured to at least: calculate a patient respiratory flow signal based at least in part on operating parameters of the apparatus; generate first output indicating whether the patient is breathing based at least in part on the calculated patient respiratory flow signal; generate second output indicating whether the patient is experiencing an apneic event based at least in part on the calculated patient respiratory flow signal; determine that a fault likely exists based at least in part on the generated first and second output; and generate an alert responsive to the determination.

17. The apparatus of claim 16, wherein the determination that the fault likely exists is made responsive to the first output and the second output indicating that the patient is breathing and experiencing an apneic event at the same time.

18. The apparatus of claim 16, wherein the determination that the fault likely exists is made responsive to the first output and the second output indicating that the patient has neither breathed nor experienced an apneic event for a predetermined time period.

19. The apparatus of claim 16, wherein the breath detection and apnea detection modules are software modules residing on and executed under control of the apparatus, the generated alert being indicative of a likely software fault in the breath detection module and/or the apnea detection module.

20. The apparatus of claim 16, further comprising at least one of: a motor speed transducer configured to produce a motor speed transducer signal; a pressure transducer configured to produce a pressure signal indicative of the pressure generated by the apparatus; and a flow transducer configured to produce a gas flow signal, wherein the respiratory flow signal is calculated based at least in part on parameters gathered from the motor speed transducer signal, the pressure signal, and/or the gas flow signal.

21. The apparatus of claim 6, wherein the generated alert is indicative of a likely software fault.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIGS. 1(a) to (d) illustrate possible configurations of a device employing a humidifier and a loop back box.

(2) FIG. 2 CPAP Apparatus

(3) FIG. 3(a) Desirable operating region for blower high speed

(4) FIG. 3(b) Desirable operating region for blower low speed

(5) FIG. 4 Computer software block diagram

(6) FIG. 5 Pressure transducer fault regions

(7) FIG. 6 Flowchart: Detect pressure transducer stuck low

(8) FIG. 7 Flowchart: Detect pressure transducer stuck high

(9) FIG. 8 Flowchart: Detect flow transducer stuck low

(10) FIG. 9 Flowchart: Detect flow transducer stuck high

(11) FIG. 10 Flowchart: Detect snore transducer stuck low

(12) FIG. 11 Flowchart: Detect snore transducer stuck high

(13) FIG. 12 Flowchart: Detect speed transducer stuck low

(14) FIG. 13 Flowchart: Detect speed transducer stuck high

(15) FIG. 14 Flowchart: Detect stalled motor

(16) FIG. 15 Flowchart: Detect restricted motor

(17) FIG. 16 Software fault diagnosis regions

(18) FIG. 17 Flowchart: Software fault diagnosis

(19) In the figures, the D symbol is a logical and, the rectangle symbol is a block of code. The triangle symbol is a comparator, if the conditions are met, then the output is a logical 1, otherwise 0.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

(20) An example of a problem solved by the present invention is provided by the Applicant's AutoSet. T device, which is provided with a loop back box which facilitates the optional connection of a humidifier into the air delivery circuit upstream of the pressure transducer. The arrangement is described in co-pending Australian patent application No. 71978/98 filed 18 Jun. 1998 and is schematically illustrated in FIGS. 1(a) to (d) hereof.

(21) FIGS. 1(a) and (b) show the correct manner in which the loop back box and the humidifier are intended to be connected. In FIG. 1(a) the humidifier is not used, and the loop back box connects the flow generator (F) to the internal conduit where the pressure sensor (T) is located. The mask (M) is connected to its outlet downstream of the pressure sensor.

(22) Typically the ventilatory assistance for CPAP or NIPPY treatment is delivered to the patient by way of nasal mask. Alternatively, a mouth mask, a full face mask or nasal prongs can be used. In this specification, any reference to a mask is to be understood as incorporating a reference to a nasal mask, mouth mask, full face mask or nasal prongs.

(23) In FIG. 1(b) the humidifier (H) replaces the loop back box and gas correctly flows from the flow generator, through the humidifier, past the pressure transducer and thus to the mask.

(24) In order that the Auto Set device be compatible with a large range of standard tubing, humidifiers and masks, the outlets all have the same size and shape. It is therefore possible to assemble the equipment incorrectly, as shown in FIGS. 1(c) and (d). In FIG. 1(c) neither the loop back box nor the humidifier is used, and the mask is attached directly to the outlet of the flow generator, thus by-passing the pressure sensor. In FIG. 1(d) the mask receives no pressurised air. In either of these situations the pressure transducer will give an incorrect indication of mask pressure and this may lead to a dangerous overpressure in an automatically adjusting device.

(25) By monitoring the pressure transducer output in conjunction with motor speed this dangerous condition can be detected and responded to. Boundary conditions of high motor speed and low pressure, and low motor speed and high pressure, may be chosen as indicators of a fault condition. The same conditions can also be used to assert a fault in the case of a defective pressure transducer, and are described later in relation to FIGS. 6 and 7.

(26) A similar approach enables a stalled motor condition to be catered for in the NIPPY device without the use of devices such as fuses. A stalled motor will not generate any pressure, and the feedback control loop to the servo unit will cause an increase in motor current. Conventionally, a fuse or other cut-out device will be required to protect the motor from overheating. If however, the motor parameters of motor speed and a motor drive parameter such as a function derived from current are monitored, a fault condition may be established. For example, if the motor parameter is greater than 80% and the motor speed remains below 4,500 rpm for 0.2 seconds, a Stalled Motor condition exists. A Restricted Motor condition can also be defined, for example where the motor parameter exceeds 95% and the motor speed is lower than 15,000 rpm, for at least 30 seconds.

(27) The corrective action in these cases is to disable the motor and otherwise disable operation of the machine until service can be performed, this mode being termed the service required mode.

(28) Malfunction of the motor speed transducer will impact on the detection of motor stall, and on other fault detection functions which will be described below. In this specification there is therefore also described the manner in which the invention may be applied to the detection of motor speed transducer failure.

(29) It will be seen that the invention can be put into effect with appropriate software, using the control electronics already provided in the machine, and therefore represents an economical solution to the diagnostic objectives.

Advantages of the Invention

(30) The invention has a number of advantages over the prior art. Firstly, safety can be improved. The greatest proportion of CPAP treatment apparatus are for use in a non-clinical setting, in which environment a user is untrained to detect conditions indicative of faults. Such fault conditions can lead to the CPAP treatment apparatus being ineffective or even dangerous. For example, the flow generators used in certain modern CPAP treatment apparatus are capable of delivering pressures in excess of 30 cm H.sub.2O, which may be required in certain situations, but excessive and potentially dangerous in others.

(31) The costs, both direct and associated, of CPAP treatment apparatus which include the invention can be reduced. Certain hardware such as fuses and other analogue circuitry may no longer be required. Furthermore, it becomes cheaper for technicians to diagnose faults since the device may be interrogated by interfacing with the controller, reducing the need to remove casing during service. Interfacing may be done locally or remotely, for example through a network. In a clinical setting, this may have the further advantage of reducing patient disturbance.

(32) In the CPAP treatment device with fault diagnosis, useability is improved. It becomes possible to provide fault diagnosis and rectification information to those without clinical or technical skills.

(33) The invention may also be used to predict more accurately when faults may occur, for example, to predict when a flow generator air inlet filter may need changing, based on measurements of the motor load.

(34) In the specification, any reference to operating parameters is to be understood to relate to any form of data or state signal, transducer or actuator, and the mechanical and electrical functions of component elements/apparatus of a CPAP apparatus. Any reference to process is to be understood to mean a unit of hardware and/or software which can perform a task or set of related tasks, for example, a fault detection process, a feedback control process, a pressure measurement process or a flow measurement process.

(35) Following the diagnosis of a current or potential fault occurring, the response may be one or more of the following: issuing a warning of the fault condition, recording a diary entry describing the fault condition, adjusting operating parameters and switching between the functional and stand-by or stop modes, or switching the device to a service-required mode. The response may be immediate, or at some later period, for example, the morning following the sleep period during which the device was used.

(36) FIG. 2 shows a simplified schematic of a typical CPAP treatment apparatus. An impeller (1) is powered by an electric motor (2) using a servo (3) under the direction of a microprocessor-based controller (4). The supply of breathable gas is carried to the mask (5) through a flexible conduit (6). The apparatus has various switches (7), displays (8) and a number of transducers. The transducers monitor a number of processes, for example: volumetric flow rate (10) (at a predetermined point in the flow path), pressure (11) (at a predetermined point downstream of the flow generator outlet or at the mask), snore (12), flow generator rotational speed (13) and motor parameter (14).

(37) The Concept of an Acceptable Region: Hardware Operation

(38) There may be a relationship between the measured blower flow rate, f, and the measured blower output pressure, p, such that f decreases when p increases. It may be desired that for acceptable operation of the device, the parameter values be kept in a certain region. The function R for that process may be written:
R=R(f,P)

(39) Distinct functions may be determined for different conditions, for example, high and low motor speeds, as shown in FIGS. 3a and 3b. Alternately, the acceptable region could be defined by a 3-parameter model, R=R (f, p, ), where is motor speed.

(40) Example of a Logical Test

(41) In a simple case, the acceptable region may be a rectangle, defined by two values of flow f.sub.1 and f.sub.2 and two values of pressure, p.sub.1 and p.sub.2. If the estimates of the parameter values were {f, p} then the test to diagnose a current fault may be, for example:

(42) If p>p.sub.1 and p<p.sub.2 and f>f.sub.1 and f<f.sub.2 then the current operation region lies within the acceptable range.

(43) Alternatively, the method may diagnose a fault if {f, p} lay outside the acceptable range for an instant, or lay outside the desirable range for some duration.

(44) Computer Software Block Diagram

(45) Advantageously, the invention is implemented in software. In this case, no additional hardware is needed. The fault detection software processes may be executed in conjunction with existing software. This is shown in FIG. 4. The inputs to the controller (4) are analogue electronic signals indicative of the value of various sensors, transducers and other electronic circuitry. These are converted to digital signals. The hardware parameter values (41) are passed to one or both of the normal computer software processes (44) (for example, feedback control processes) and the fault diagnosis processes (43). Further parameters may be generated (42) (for example, flags) indicative of the operation of each of the software processes, and passed back into one or both of the software processes. In addition, the processes may generate hardware instructions (45) (for example, information to be displayed, or directions to shut down the operation of the CPAP device) which are converted to analogue electronic signals and passed to the relevant hardware.

(46) Fault diagnosis of each apparatus process may be executed as a distinct software process, or several apparatus processes may be monitored within a single software process.

PREFERRED MODE OF OPERATION

(47) Processes Monitored

(48) In the preferred mode of carrying out the invention, in each device, a plurality of processes are monitored. These include:

(49) Pressure transducer operation

(50) Flow transducer operation

(51) Snore transducer operation

(52) Speed transducer operation

(53) Motor operation

(54) Breath and Apnea detection algorithms

(55) Fault diagnosis process operation

(56) Air filter operation

(57) Correct assembly of air delivery circuit

(58) Modes of fault diagnosis for these processes are described in the following sections.

(59) Pressure Transducer

(60) The invention may be used to detect whether the pressure transducers are correctly operating. A consequence of a failure of a pressure transducer (11) could be overpressure or under-pressure to the mask and thus the patient. To detect such a failure condition, the transducer pressure is monitored (11) together with motor speed (13). The regions of faulty device operation are shown in FIG. 5. Logical flowcharts indicating the decision process are shown in FIGS. 6 and 7.

(61) If the sensed pressure remains below 2.0 cm H.sub.2O (62) while the motor speed is above 12,000 rpm (61) for at least 0.3 seconds (65), a fault condition is signalled as a Pressure Transducer Low failure (66). If neither condition is satisfied, a timer is reset (64).

(62) In a similar way, if the pressure value remains above 15 cm H.sub.2O (72) while the motor speed is below 4,500 rpm (71) for at least 0.3 seconds (75), that corresponds to a Pressure Transducer High failure (76). On detection of either form of pressure transducer failure, corrective action is taken to disable power to the motor (2) and a notification is given on a display. The user may then check all components in the air delivery circuit for an obstruction or an incorrect connection.

(63) An additional embodiment is used for the pressure transducer fault diagnosis mode in conjunction with motor speed. The output from the pressure transducer is split into two signals. One signal, P.sub.high remains unfiltered, whilst the other signal is low-pass filtered to become P.sub.low. Each signal is tested to yield an output of Pressure Transducer High, Pressure Transducer Low or ok. The conditions for high and low are as described above. A signal which is neither high nor low is deemed to be ok. Unless the test result for both signals is ok, a fault condition will be signalled. If the test result for both signals is low, the fault condition is taken as an indication that the hoses are not connected.

(64) Flow Transducer

(65) A failure of the flow transducer (10) can cause errors in flow measurement, mask pressure measurement and signal processing (such as inspiratory flow flattening index, as taught in U.S. Pat. No. 5,704,345). This in turn may result in errors in the delivery pressure from the flow generator (1), compromising the efficacy of treatment. Logical flowcharts indicating the decision process are shown in FIGS. 8 and 9.

(66) To detect such a fault, the flow transducer (10) signal is monitored together with motor speed (13). If the flow remains below 5% scale deflection (f.s.d.) (82) while the motor speed is less than 4,800 rpm (81) for more than five seconds (85), the fault condition is signalled as Flow Transducer Low failure (86). If neither condition is satisfied the counter is reset (84). In a similar manner, if the flow generator flow remains above 95% f.s.d. (92) with the motor speed less than 4,800 rpm (91) for more than 5.0 seconds (95), the fault of Flow Transducer High is signalled (96).

(67) On occurrence of either fault condition, the patient is notified on a display, and in addition, the mask pressure will be increased over some time, for example, five minutes, to a predetermined level, in this case, the 95% centile of the previous session. However if that pressure value would be invalid by representing a dangerous overpressure, the pressure will be limited to 10 cm H.sub.2O.

(68) Snore Transducer

(69) A failure of the snore index transducer can cause errors in the snore measurement and the calculated snore index. Such errors result in consequential errors in the treatment pressure delivered by the flow generator (1), again possibly compromising the efficacy of treatment. Logical flowcharts indicating the decision process are shown in FIGS. 10 and 11.

(70) To detect a failed snore transducer the snore transducer signal (12) and motor speed (13) are sampled. If the snore index signal remains below 5% f.s.d. (102) with the motor speed greater than 11,000 rpm (101) for more than two seconds (105), a Snore Transducer Low failure is signalled (106). If neither condition is satisfied the counter is reset (104). If the snore transducer (12) signal remains above 95% f.s.d. (112) while the motor speed is less than 6,000 rpm (111) for more than two seconds (115) a Snore Transducer High failure mode occurs (116).

(71) The corrective action that can be taken is to notify the user of the occurrence and, if the flow generator (1) is generating a pressure, attempt to increase the pressure over some time, for example 5 minutes, to a predetermined level, for example, the 95% centile of the previous session. If that pressure value is invalid by representing a dangerous overpressure, the pressure will be limited to 10 cm H.sub.2O.

(72) Motor Speed Transducer

(73) An incorrect motor speed may impact upon functions arising from the pressure transducer (11), flow transducer (10) and snore transducer (12). Logical flowcharts indicating the decision process are shown in FIGS. 12 and 13. Detection of a failure of the motor speed transducer (13) is achieved by monitoring the speed (13) together with the sensed pressure (11) and flow (10) and the motor drive parameter (14). If the motor speed remains below 6,000 rpm (121) while the motor drive parameter is above 15% (122) and either the pressure is above 8.0 cm H.sub.2O (123) or (125) the flow is greater than 75% f.s.d. (124), this set of conditions being true for more than one second (128), then a Motor Speed Low failure is signalled as occurring (129). If any of the conditions are not met, the counter is reset (127).

(74) If the motor speed remains above 18,000 rpm (131) while the motor drive is below 5% (132) and (136) either the pressure is below 2.0 cm H.sub.2O (133) or (135) the absolute flow is less than 10 l/min (134) for more than one second (138), a Motor Speed High failure is signalled as occurring (139).

(75) The corrective action that can be taken is to notify the user of the occurrence and, if the flow generator (1) is generating a pressure, attempt to increase the pressure over some time, for example 5 minutes, to a predetermined level, for example, the 95% centile of the previous session. If that pressure value is invalid by representing a dangerous overpressure, the pressure will be limited to 10 cm H.sub.2O.

(76) Motor

(77) As noted earlier, a stalled motor will not generate any pressure. Such a condition can occur if the impeller (1) is jammed, for example. The feedback control loop from the flow generator (1) to the servo unit (3) will cause an increase in the power delivered to the motor (2), the increase continuing until maximum power is supplied. A consequence can be overheating of the motor to the extent that insulation or windings fail and the motor must be replaced. Flowcharts describing the operation for stalled and restricted motor are shown in FIGS. 14 and 15 respectively.

(78) A motor failure condition is detected by monitoring the motor speed (13) and motor drive parameter. If the motor parameter is greater than 80% (142) and (143) the motor speed remains below 4,500 rpm (141) for 0.2 seconds (145), a Stalled Motor condition exists (146). If any of the conditions are not met, the timer is reset (144). In a similar way, if the motor parameter exceeds 95% (152) and the motor speed is lower than 15,000 rpm (151), occurring for more than 30 seconds (155), then a Restricted Motor condition exists (156).

(79) The corrective action is to disable power to the motor and otherwise disable operation of the machine until a service can be performed, this mode being termed the service-required mode.

(80) Pneumatic Performance

(81) In another embodiment, the method may be used to check the overall pneumatic, or air delivery, performance of the apparatus. The relevant regions are shown in FIGS. 3(a) and 3(b), being the desired operating region .sub.d and fault regions .sub.f1, .sub.f2. The operating regions are a function of pressure (I), flow (f) and motor speed (). If, the operational state of the device lies in region .sub.f1, there are several possible causes:

(82) Dirty filter

(83) Internal air path blockage in the CPAP apparatus

(84) Leak in the internal air path of the CPAP apparatus

(85) Flow generator failure

(86) Faulty speed sensor

(87) Faulty flow sensor

(88) Faulty pressure sensor

(89) Possible responses include:

(90) Issuing a warning advising that the filter needs changing

(91) Switching the apparatus to service-required mode

(92) If the operational state of the device lies in region there are several possible causes:

(93) Faulty speed sensor

(94) Faulty flow sensor

(95) Faulty pressure sensor

(96) The response would be to switch the device to the service required mode.

(97) The embodiments described earlier may be used in conjunction with the current embodiment, to distinguish which of the sensors may be faulty.

(98) Fault Diagnosis Operation

(99) The events pressure transducer stuck high and pressure transducer stuck low are mutually exclusive. Similarly the flow, snore and motor speed transducers have mutually exclusive conditions. In another embodiment of the invention, the fault diagnosis processes tests whether mutually exclusive conditions have been deduced and signals an error in fault diagnosis if that event occurs.

(100) Breath and Apnea Detection Processes

(101) FIG. 16 shows the operating regions for the breath and apnea detection processes. These processes detect the presence of breathing by the user and the occurrence of apneas, these two events being mutually exclusive. During normal operation of the CPAP device, either one or the other of these events ought to be detected. In this case, these two events lie within non-intersecting regions of .sub.d (FIG. 16). If the current operating region of the device lies outside both these two regions, then some kind of error has occurred and an error is asserted. A flowchart for this fault detection process is shown in FIG. 17. While the device is operating the Breath Detection and Apnea Detection processes are monitored. The fault detection process will assert a fault when either of the following two conditions occur: Apnea and Breath are asserted simultaneously Neither Apnea nor Breath are asserted for a predetermined period.

(102) While particular embodiments of this invention have been described, it will be evident to those skilled in the art that the present invention may be embodied in other specific forms without departing from the essential characteristics thereof. The present embodiments and examples are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein.