RESPIRATORY CONDITION DIAGNOSIS AND APPARATUS THEREFOR This invention concerns respiratory condition diagnosis and apparatus therefor.
The invention has been conceived and developed to date primarily for the diagnosis of asthma, particularly in distinction from an alternative diagnosis of bronchitis, and especially In relation to children. An earlier study (British Medical Journal, 29 July 1978, pp.331-332) showed that asthma in children can be undetected and subject to delay 1n diagnosis, with the most common cause of delay, particularly in the pre-school age group under 4-years old, being an alternative diagnosis of bronchitis. This can result in a lack of treatment or an unnecessary use of antibiotics, the latter being rarely appropriate in asthma, and consequent further bronchial damage leading to hyper-responsive airways. The traditional use of the stethoscope and use of the more recent instruments for measuring peak expiratory flow rate, PFR, do not allow a satisfactory resolution of this situation. For example in the case of school children simple bronchitis can lead to a drop 1n PFR without necessarily exhibiting any degree of bronchial muscle spasm to signal the relevant dysfunction, and in the case of pre-school children there is an inability to monitor PFR accurately due to a lack of compliance.
Various proposals have been made suggesting that an Improved respiratory condition diagnosis may be facilitated by a procedure involving use of a transducer to provide electrical signals representative of breath sounds and instrumentation to analyse such signals for features characteristic of dysfunction of interest. However, these proposals have not given results indicative of routine clinical utility. An object of the present invention is to Improve this overall situation and to this end follows a procedure of this last form, but with signal provision being effectively continuous during
respiration extending through at least one phase of an inspiration/expiration cycle and signal analysis being of a spectral form to indicate frequency components in a range up to at least about 5 kHz and the relative intensities of such components.
This combination of signal provision and analysis has, during development of the invention, been found to be necessary to the provision of results which are consistent to a degree appropriate to routine clinical utility. At the same time this finding indicates why prior proposals having a similar objective have in fact not proved satisfactory. The prior proposals have, for example, variously entailed application of a transducer to the chest wall with consequent attenuation of breath sounds down to about 2 kHz, the provision of signals from a transducer in response to sound pressure level and so not representative of the sound frequency spectrum, and the analysis of signals representing breath sounds at only one instant in time.
Preferably, according to the present invention signal provision is effective during an expiration phase as breath sounds are generally more pronounced then than during inspiration. More preferably, the sounds in question are those arising from forced expiration as this will tend to enhance signal characteristics of interest.
Also the transducer is preferably applied in the near vicinity of the mouth for effectively direct response to breath sounds to the extent that no significant attenuation occurs by way of sound transmission through body tissue.
For the purposes of these preferred facets of the invention the transducer is suitably sited in a mouthpiece of predetermined form. Clearly this will serve to lcc?te the transducer appropriately- Also, the use of a mouthpiece will assist in normalising the procedural conditions from patient to patient and from one occasion to another for the same patient, and particularly so in the case of forced expiration.
The signal analysis preferably provides a visual display and this is conveniently of a form in which time and frequency serve
as respective coordinates in a 2-dimensional plot with frequency component intensity being depicted by way of density, colour or like variation' effectively serving as a further coordinate in the display. It may be appropriate, such as in the case where intensity is represented by a visual density variation, to display components only when above a predetermined threshold representing significance relative to noise.
While the invention has been discussed this far largely in terms of a diagnostic procedure, equally well it contemplates apparatus for carrying of the procedure.
In any event, a fuller understanding of the invention may be assisted by the following further description, given by way of example, with reference to the accompanying drawings, in which:-
Figure 1 schematically illustrates an apparatus embodiment according to the invention and employed during initial development thereof, and
Figures 2-9 respectively show different display plots obtained with use of the apparatus of Figure 1.
The apparatus of Figure 1 includes a mouthpiece 10, and a microphone 11 connectable to the input of a tape recorder 12. The tape recorder output is applicable by way of an analogue- to-digital converter 13 to the store of a computer 14 under the control of a record program denoted at 15. The stored data in the computer is applicable, in turn, under the control of an analysis program denoted at 16, to an X-Y plotter 17 of pen recorder form.
The mouthpiece is of disposable form such as used in peak expiratory flow meters and the microphone is of hand-held form for location near, but not in line with, the expired breath from the mouthpiece. Further development is likely to involve incorporation of a miniature microphone or other suitable electroacoustic transducer in the side wall of a tubular instrument for use with a separable disposable mouthpiece, the instrument being adapted for appropriate cleaning. Such a development is indicated in Figure 1 by way of transducer 11a.
Development may well also involve the use of a dedicated microprocessor or, more economically, a suitable commercially- available personal computer with dedicated software for direct handling of the microphone output. In any event, the microphone output is converted to digital form for the purposes of computer handling, including analysis, and the analysis entails fast Fourier transformation to convert the data from the time domain to the frequency domain. More particularly the transformation provides signals indicating, at each of a plurality of closely spaced successive instants of time, the microphone output intensity at each of a plurality of similarly spaced successive frequencies in the range 0-5 kHz. The last signals are applied, in the extent to which they exceed a predetermined threshold representative of noise, to the pen recorder to produce a so-called Z-plot. This plot involves a plurality of traces extending in registered manner along the paper to represent time t in seconds as the X-ordinate, with the traces being successively spaced across the paper to represent frequency f in kHz as the Y-ordinate, and each trace including transverse oscillations of amplitude proportionate -to frequency component intensity whereby to effect a visually perceived density variation as a Z-ordinate. In initial development such analysis and plots have been produced with 72 traces for respiration periods of 1.95 records including a forced expiration. Figure 2 shows one such plot for a normal adult male (40-years old). A characteristic of significance is seen to be the presence of bands of relatively high spectral energy well distributed over the whole frequency range of the plot. These bands are identified by the addition of horizontal bounding lines after generation of the plot anrf **■ further arrowed line below the time axis indicates the period of about 0.5 s during which the bands occurred, this period in fact indicating the period of peak expiration. The bands have in fact been calculated to correspond with the natural acoustic frequencies of the bronchial tree in simple organ pipe modes of oscillation and the bands have been found to be consistently repeated on successive expirations.
Figures 3-5 show similar plots for an asthmatic female adult (38-years old). These plots individually relate to expiration in circumstances involving no medication, 20 minutes following self- administration of a bronchodilatory drug ("Sal utamol") via an aerosol spray, and 2-months following routine use of such therapy.
The characteristics of significance in these plots Indicate a marked reduction of spectral energy bands in Figure 3 compared to
Figure 2, with only two bands occurring in the vicinity of 1 kHz and 2.5 kHz, but with these bands being of high intensity. In Figure 4 these two bands recur, but at reduced intensity, and a further band appears just below 4 kHz. In Figure 5 the plot is yet more comparable with that of Figure 2.
A point of interest arising from these plot comparisons so far is that improvement in an asthmatic's condition is associated with a shift of spectral energy upwards in terms of frequency. This contrasts with the commonly perceived view that asthma produces a high frequency wheeze, whereas in fact the frequency shift is down. This perception is now seen to arise from increased intensity of breath sounds at lower frequency. Again, the plots of Figures 3-5 are found to be repeatable.
Figures 6 and 7 show plots for an asthmatic female child (9-years old) in circumstances corresponding to those for Figures 3 and 4, and similarity is seen with a shift in spectral energy from lower to high frequencies. These plots are, again, repeatable. It is to be noted that it was found appropriate to increase the gain in the recorder to allow for weak expiration, but this does not affect the relative nature of the bands.
The remaining Figures 8 and 9 show plots for an adult male (58-years old) with a mixture of asthma and bronchitis in circumstances also corresponding to those of Figures 3 and 4. It can be seen that there are similarities with and differences from the earlier plots which is consistent with bronchodilation influencing the asthmatic component of the patient's condition but being unable to reverse the airways obstruction related to
bronchitic component. Thus, bronchodi lation reduces the spectral intensity in Figure 9 compared to Figure 8, but in neither case can one readily perceive discrete bands of higher energy through the frequency range. The overall results as shown by the plots of Figures 2-9 show a clear basis on which diagnosis can be aided in detecting asthma relative to Its absence and, moreover, in distinguishing from bronchitis.
It has been made clear in the above description that the invention is open to practical implementation in a varied manner, particularly in respect of apparatus. The apparatus of Figure 1 was chosen to allow flexibility whereby the underlying concept of correlation between breath sound characteristics and respiratory condition could be more thoroughly examined than in the past. In the result such a correlation has now been established and apparatus suited to routine usage can be developed on the basis of the present findings, which apparatus is likely to be of a dedicated form with consequent simplication and increased convenience in use. Some facets of such development have been indicated and other possibilities include, for example, the provision of an averaging facility whereby display results take account of plural operations by the patient, and extension of the frequency range for the purposes of analysis.