GB2329966A - Breathing pattern monitor - Google Patents

Abstract

A monitoring apparatus comprises a sensor device (10,11) arranged to produce a pulsed output signal having a frequency which varies with a parameter being monitored, a counter (12) for counting the pulses produced within a predetermined time period, and means (14) for analysing the number of pulses which are counted in successive time periods. The embodiment of the invention comprises a means to monitor breathing patterns in infants so that apnoeic attacks may be detected. The sensor consists of a variable capacitor forming part of an oscillator circuit, such that respiratory action causes changes in value of the capacitor, thus affecting the output of the oscillator. A frequency spectrum of the output signal may be obtained.

Description

Monitorinq Apparatus This invention relates to a monitoring apparatus and more particularly but not solely to an apparatus for monitoring breathing patterns of infants and other persons.

The breathing pattern of premature and sick infants often becomes irregular and life threatening. Hitherto, it has been proposed to monitor for these so-called apnoeic attacks, in order to provide an alarm signal in the event that an attack is detected.

Apnoea can be classified into three main types depending on whether inspiratory muscle activity is present.

The first type of apnoea is known as central apnoea, which is considered a nervous system disorder and accounts for between 10 to 25% of all apnoea in premature infants. The condition occurs when the respiratory centre within the brain stem fails to send the necessary impulses via the nervous system to the intercostal muscles and diaphragm. Therefore, no breathing effort is made and movement of air within the lungs ceases to take place. Central apnoea is particularly common in the neonatal period, especially in the case of preterm infants.

It is believed that these infants have a reduced sensitivity to hypoxia (low blood oxygen levels) and hypercapnia (high blood carbon dioxide levels) . This is due to an under developed respiratory centre which results in immature reflex responses and poor regulation of breathing.

The second type of apnoea is known as obstructive apnoea, which occurs from a physical obstruction within the infant's upper airway. Respiration signals continue to be sent from the brain stem and inspiratory muscle activity is present, yet no effective breathing can take place. Obstructive apnoea accounts for 12 to 20% of apnoeic cases in premature infants.

It is not often as serious as central apnoea and can be called hypopnoea, when the airflow is only partially blocked. Normal breathing is usually resumed independently of external intervention. Obstructive apnoea usually occurs from the loss of muscle tone within the tongue, throat or larynx and causes a constriction as the airway collapses from negative pressure generated during inspiration. As infants are obligate nose breathers, obstructions in the nasal passage can also promote apnoeic episodes.

Finally, mixed apnoea is a combination of both central and obstructive apnoea. Mixed apnoea can account for between 53 to 71% of all apnoea in premature infants. The symptoms usually start with a central apnoeic attack, followed by a period of obstructive apnoea.

It is not uncommon for an infant to suffer from any of the above types of apnoea, and usually the child will spontaneously re-initiate its own breathing. If the event becomes more serious and the period of not breathing increases above 20 seconds, the condition is called pathological apnoea, or abnormal apnoea. Of all infant apnoea cases lasting for period of more than 20 seconds, obstructive apnoea accounts for 13% and mixed apnoea for 82%, whilst central apnoea is as little as 5%.

After apnoea periods greater than 20 seconds the infant's blood oxygen starts to fall to a dangerous level and hypoxia will occur, depriving the heart, brain and other vital tissues of oxygen. This can also cause a fall in the child's normal heart rate, leading to bradycardia and an increase in blood pressure. For short periods even these episodes are not considered uncommon for a premature infant. However, if persistent and accompanied by a colour change in the skin, marked changes in muscle tone, choking or gagging, the child is said to be have an "Apparent Life-Threatening Event".

Monitoring infants for apnoea is well known and has been used in the past both in the hospital environment, and more recently in the home. In the United States well over 50,000 apnoea monitors are used daily at home for infants with varying success. Two problems which all monitors can suffer to a certain degree are caused by false negatives and false positives.

False negatives occur when body movements unrelated to breathing, such as twitching or increased heart beat are sufficient to register on a monitor as breathing and prevent it from alarming even though breathing movements have ceased.

False positives occur when the alarm gives a positive warning yet the child is perfectly healthy. This is often dus to sensors becoming detected or from external noisE interference.

One method of monitoring for apnoea comprises attachinc a variable capacitance pressure transducer to the patient's abdomen. The capacitance of the transducer varies as the patient's abdomen moves whilst breathing. The transducer forms a part of a tuned circuit and thus the resonant frequency of the tuned circuit increases as the capacitance increases an vice-versa. The varying output signal of the tuned circuit is then analysed to detect apnoea.

The change in capacitance whilst breathing is small compared with the overall capacitance of the transducer anc thus a high sensitivity circuit is needed to detect for an change in frequency caused by apnoea.

A disadvantage of having a high sensitivity circuit is that many false alarms can occur. Another disadvantage of known systems is that the output of the tuned circuit is connected to a frequency-to-voltage convertor and it will b appreciated that this conversion process introduces quantisation noise. Furthermore, the varying output voltagE is then applied to an analogue-to-digital converter, which again introduces more noise.

We have now devised a monitoring apparatus which alleviates the above-mentioned problems.

In accordance with this invention there is provided a monitoring apparatus comprising a monitoring device arrange to produce a pulsed output signal having a frequency which varies in accordance with a parameter being monitored, a counter for counting the number of pulses within a predetermined time period and means for analysing the number of pulses that are counted in successive time periods.

In use, the count value of the counter will remain the same if the frequency does not change. However, if the frequency increases then the count value will increase and vice-versa. This change in count value can be detected tc signal any change in frequency.

Preferably the analysing means comprises means for incrementing and decrementing the counter over successive time periods and means for determining the residual count value of the counter after alternate time periods.

Thus, if the frequency does not change during successive count periods, the residual count after the clock has incremented and decremented will be zero. However, the residual count of the clock will be positive if the frequency increases over the successive count periods and vice-versa.

The apparatus is more reliable than known apparatus for monitoring apnoea because conversion errors and noise are not introduced. Furthermore, the apparatus is more reliable than known apparatus because any gradual changes in output frequency caused by temperature drift or component ageing will cause negligible changes in the residual count value.

Preferably the pulsed output signal comprises a square wave, thereby making the apparatus simple to implement using digital techniques.

Preferably the counter is reset once the residual count value has been read.

Preferably the frequency spectrum of the residual count valves is obtained, say by Fourier Transform means, since it has been found that the three different types of apnoea have a characteristic distribution of frequency components within the frequency spectrum.

For central apnoea, the counter values are almost time invariant and produce a frequency spectrum which contains a high proportion of low frequency components. In the case of obstructive apnoea, where the counter value is rapidly varying in time, a frequency spectrum with peaks in the high frequency range will be produced.

An embodiment of this invention will now be described by way of example only and with reference to the accompanying drawings, in which: FIGURE 1 is a block diagram of a monitoring apparatus for detecting apnoea; FIGURE 2 is a flow chart to illustrate the principle of operation of the apparatus of Figure 1; FIGURE 3 is a graph of counter value against time for a central apnoea attack; FIGURE 4 is a graphical representation of the Fourier Transformation of the counter values of Figure 3; FIGURE 5 is a graph of counter value against time for an obstructive apnoea attack; and FIGURE 6 is a graphical representation of the Fourier Transformation of the counter values of Figure 5.

Referring to Figure 1 of the drawings, there is showr a monitoring apparatus for detecting apnoea. The apparatus comprises a variable capacitance pressure transducer 10 for attaching to the abdomen of the patient being monitored. The transducer 10 is connected to a variable frequency oscillator 11, which is arranged to produce a square-ware output signs having a 50% duty cycle. The frequency of the square wavc signal is directly proportional to the capacitance of ths transducer 10.

The digital output signal from the variable frequency oscillator 11 is connected to a 12-bit synchronous binary up/down counter 12. The output of the up/down counter 12 is connected to a personal computer 14 via an 8-bit parable input/output circuit 13.

The up/down counter 12 has its count enable pir connected to a monostable 16 and its count direction pir connected to the monostable 16, via a divide by two circuit 17.

The monostable 16 is arranged to produce a count enable pulsE of approximately 50ms. The monostable 16 is activated by a lms pulse, which is output from the computer 16 via thc input/output circuit 13. The monostable 16 also has an output connected to the input/output circuit 13, to informs thc computer 14 that the count period has ended and that thc residual count value can be read.

An audible alarm 15 is connected to the computer via the input/output circuit 13.

In use, the variable oscillator is set to produce e square wave output having a central frequency of 30 kHz, which varies by about +/- 250 Hz as the breathing of the patient causes the capacitance of the transducer 10 to vary.

Referring to Figure 2 of the drawings, the computer 14 outputs a start signal to the input/output circuit 13, which activates the monostable 16, as hereinbefore described, tc reset and activate the counter 12. The counter 12 then counts the number of pulses output from the oscillator 11. After some, the output of the monostable 16 falls and the count stops.

The computer 14 then outputs another start pulse, which activates the monostable 16 as before but this time the output of the divide by two counter 17 changes, so that the counter 12 is enabled to count down.

The counter 12 then decrements from the previous count value until the output of the monostable 16 falls. The computer 14 then reads the residual count value after the counter 12 has counted up and down in successive 50ms monostable periods.

If the residual value of the counter is zero then it will be appreciated that the frequency of the oscillator 11 did not change in the successive periods. However, if the residual count value is positive it will be appreciated that the frequency has decreased and vice-versa.

Referring to Figure 3 of the drawings, when central apnoea occurs no breathing effort is made and chest movement ceases, thereby causing the output frequency of the oscillator to stabilise at about 30kHz. Accordingly, the residual count values are close to zero.

The computer 14 is arranged to apply a Fast Fourier Transform (FFT) to the residual count values of the counter 12.

Referring to Figure 4, it can be seen that for central apnoea the counter values are almost time invariant and produce a frequency spectrum which contains a high proportion of low frequency components.

Referring to Figure 5 of the drawings, when obstructive apnoea occurs, rapid chest movements occur, thereby producing residual count values which vary greatly with time. Referring to Figure 6 of the drawings, it can be seen that a Fourier Transformation of the counts resulting from obstructive apnoea shows a large number of peaks in the high frequency range.

Mixed apnoea comprises a period of central apnoea followed by a period of obstructive apnoea, which will produce a combination of the two results.

The output from the Fourier Transformation program can be applied to a peak detection circuit or filtered in order to determine whether it matches the characteristics of one of the aforementioned apnoea attacks.

The audible alarm 15 is then activated, once apnoea has been detected.

Claims (7)

1. Claims 1) A monitoring apparatus comprising a monitoring device arranged to produce a pulsed output signal having a frequency which varies in accordance with a parameter being monitored, a counter for counting the number of pulses produced within a predetermined time period and means for analysing the number of pulses which are counted in successive time periods.
2. 2) A monitoring apparatus as claimed in claim 1, in which the analysing means is arranged to cause the counter to increment and decrement its count over successive time periods and is further arranged to read the residual count of the counter after alternate said time periods.
3. 3) A monitoring apparatus as claimed in claim 2, arranged so that, in use, the counter is reset once each residual count is read.
4. 4) A monitoring apparatus as claimed in claim 2 or 3, in which the analysing means is arranged to form the frequency spectrum of the residual counts read from the counter.
5. 5) A monitoring apparatus as claimed in claim 4, in which the analysing means is arranged to perform a Fourier Transform to form said frequency spectrum.
6. 6) A monitoring apparatus as claimed in claim 4 or 5, in which the analysing means is arranged to compare said frequency spectrum with at least one prestored characteristic indicative of a predetermined conditions.
7. 7) A monitoring apparatus substantially as herein described with reference to the accompanying drawings.