DEVICE AND METHOD USING DAMPED HARMONIC ANALYSIS FOR AUTOMATED PULMONARY AND ABDOMINAL EXAMINATION

Abstract

An analyzer for diagnosing pulmonary and abdominals including, a pulsed force generator for outputting a mechanical disturbance to generate vibrations and reflected/return waves/vibrations in a patient's torso, and sensors for detecting the vibration/return wave signals. The apparatus compares the detected electrical signals with pre-stored reference wave profiles and based on the compared data generates an output signal indicative of a potential presence or absence of a pulmonary disease and/or condition.

Claims

1. An analyzer for diagnosing a pulmonary and/or abdominal condition of a patient, the analyzer comprising, a pulsed force generator operable to generate and transmit to a target area of the patient at least one preselected output pulsed force, each said output pulsed force being selected to generate at the target area at least one associated vibration or reflected return wave, a sensor assembly for detecting the at least one vibration or reflected return wave at said target area, and converting detected energy thereof into electrical signals, memory containing stored data representative of pre-stored vibration or reflected wave profiles indicative of the presence and/or absence of one or more pulmonary and/or abdominal conditions, p1 a processor containing programme instructions, whereby said processor is operable to, compare data representative of the detected energy of at least a part of the at least one associated vibration or reflected return wave generated with said stored data representative of at least one said pre-stored vibration or reflected wave profiles and based on the compared data generate an output signal indicative of a potential presence or absence of a pulmonary disease and/or condition in said patient, wherein said processor includes pre-stored software comprising programme instructions operable to compare data representative of an output parameters of said output pulsed force and the data representative of the detected vibration or reflected waves, and perform damped harmonic analysis on at least one of said data representative of said output disturbance and said data representative of said detected vibration or reflected waves.

2. (canceled)

3. The analyzer as claimed in claim 1, wherein the processor is operable to identify and select by at least one of signal filtering and signal truncation, an informative portion of the electrical signals as the part of the at least one associated return energy wave compared with said stored data.

4. The analyzer as claimed in claim 3, wherein the informative portion of the detected energy is identified and selected by filtering the electrical signals based on pre-identified background patterns.

5. The analyzer as claimed in claim 1, wherein the processor is operable to effect damped sinusoidal signal conditioning on data, representative of the detected energy of at least part of a plurality of said detected vibrations or reflected return waves; and output a classification of the detected energy as at least part of the output signal based on said signal conditioning, and wherein prior to damped harmonic analysis, the processor is operable to effect signal conditioning of the detected energy electrical signals by one or more steps selected from the group consisting of windowing the electrical signals, removing direct current (DC) offset from the electrical signals; and filtering preselected high and/or low frequency components from the electrical signals.

6. (canceled)

7. The analyzer as claimed in claim 1, wherein said damped harmonic analysis comprises nonlinear spectral fitting of the detected energy of a plurality of said associated return energy waves detected at the target area, with at least one of Lorentzian curves, Prony's method and Pisarenko method.

8. The analyzer as claimed in claim 7, wherein said processor is operable to classify said output signal based on at least one vibration or reflected return wave parameter selected from the group consisting of wave amplitude, phase, frequency and damping coefficient.

9. The analyzer as claimed in claim 1, wherein said processor is further operable to effect classification of the output signal based on at least one input patient parameter selected from the group consisting of patient age, sex, weight, and smoker status.

10. The analyzer as claimed in claim 1, wherein said pulsed force generator includes a selectively displaceable piston member which is movable between a forward position, where said piston member engages said target area or a plessimeter to transmit said output pulsed force thereto, and a retracted position spaced therefrom.

11. The analyzer as claimed in claim 10, wherein the output pulsed force comprises a pulsed impact force at said target area selected at between about 0.1 and 10N, preferably 0.2 to 5N and more preferably 0.5 to 4N, and said output pulsed force comprises a pulsed impact force having a repetition frequency selected between 0.1 and 10 hertz.

12. (canceled)

13. The analyzer as claimed in claim 1, wherein said target area is selected from the group consisting of a chest area, a back area and an abdoment area of the patient.

14. The analyzer as claimed in claim 10, wherein said analyzer comprises a portable hand-held pulmonary function tester for diagnosing a pulmonary injury or condition in said patient; said tester further comprising a portable power source for supplying electric power to said pulsed force generator and said sensor assembly, and an output display for displaying said output signal as a graphic output.

15. Apparatus for diagnosing a pulmonary or abdominal condition of a patient, the apparatus comprising, a pulsed force generator operable to generate and impart on a target area of the patient's torso a preselected output pulsed force, the input energy force being selected to generate in said patient's torso associated vibrations and/or return waves, a sensor assembly for detecting said vibrations and/or return waves at said target area, and converting said vibrations and/or return waves into sensed data signals, and a processing assembly having memory, a processor, and programme instructions, whereby said processor is operable to, perform damped harmonic analysis on said sensed data signals to identify a damped harmonic signal, and compare at least part of the sensed data signals with data representative of at least one of the output energy of said pulsed force generator and data stored in memory representative of one or more vibration and/or return wave profiles representative of a pre-identified pulmonary or abdominal condition or state, based on the compared data, generate at least one of an audible and a visual output signal indicative of the presence or absence of a pulmonary disease or condition in said patient, and wherein said pulsed force generator includes a selectively movable piston member which is reciprocally moveable between a forward position, where said piston member is moved into physical engagement with said target area or a plessimeter to transmit said pulsed force thereto, and a retracted position wherein said piston member is moved rearwardly to a position spaced from said target area.

16. The apparatus as claimed in claim 15, wherein said damped harmonic analysis comprises performing nonlinear spectral fitting of the sensed data signals with at least one of Lorentzian curves, Prony's method and Pisarenko method.

17. The apparatus as claimed in claim 15, wherein said processor is operable to compare the identified damped harmonic signal with at least one said preselected vibration and/or return wave profiles representative of a pre-identified pulmonary disease or condition stored in memory, said processor further outputting said output signal as a classified signal based on one or more sensed energy wave parameters selected from the group consisting of energy wave amplitude, phase, frequency and damping coefficient.

18. The apparatus as claimed in claim 15, wherein said processor is further operable to effect classification of said compared data based on at least one input patient parameter selected from the group consisting of patient age, sex, weight, and smoker status.

19. (canceled)

20. The apparatus as claimed in claim 18, wherein the pulsed force generator includes a motor operable to effect pulsed movement of said piston member to generate a pulsed impact force at said target area selected at between about 0.1 and 10N, and wherein said pulsed impact force has a repetition frequency selected at between 0.1 and 10 hertz.

21. (canceled)

22. The apparatus as claimed in claim 15, wherein said target area is selected from the group consisting of the patient's chest wall, stomach, a super-transpylonic planar region of the patient's back and a sub-transpyloric planar region of the patient's back.

23. The apparatus as claimed in claim 22, wherein said apparatus comprises a portable hand-held pulmonary function analyzer, and further includes a portable power source for supplying electric power to said pulsed force generator, said sensor assembly, and said processing assembly, and wherein the processor is operable to identify by at least one of signal filtering and signal truncation. an informative portion of the sensed data signals as the part of the sensed data signals compared with the data stored in memory.

24. (canceled)

25. The apparatus as claimed in claim 15, wherein the informative portion of the sensed data signals is identified by filtering the sensed data signals to remove pre-identified background features and/or patterns, and wherein prior to performing damped harmonic analysis, the processor is operable to effect signal conditioning of the sensed data signals by one or more steps selected from the group consisting of windowing the sensed data signals, removing direct current (DC) offset from the sensed data signals, and filtering preselected high and/or low frequency components from the sensed data signals.

26. (canceled)

27. A portable diagnosis analyzer for diagnosing a pulmonary function of a patient, the analyzer including, a display, a pulsed force generator operable to generate and transmit to a target area of the patient's torso a preselected output pulsed force selected to generate at said target area associated vibrations or reflected waves, a sensor assembly operable to detect energy of said vibrations or reflected waves at said target area and convert such detected energy into sensed data signals, and a processing assembly including memory and a processor containing programme instructions, said processor is operable to, perform damped harmonic analysis on at least part of said sensed data signals to generate a damped harmonic signal, and compare the damped harmonic signal with preselected harmonic signals stored in said memory, said preselected harmonic signals being representative of a pre-identified pulmonary state, disease or condition, and based on said comparison, generate an output a signal to a user indicative of the potential presence or absence of a pulmonary disease and/or condition in said patient, wherein the processor is operable to identify by at least one of signal filtering and signal truncation, an informative portion of the sensed data signals as the part of the sensed data signals compared with the data stored in memory, and wherein said processor is further operable to compare data representative of the output energy of said pulsed force generator, and the sensed data signals, and wherein the identification of the informative portion is based in part on said comparison.

28. (canceled)

29. (canceled)

30. The analyzer as claimed in claim 27, wherein said damped harmonic analysis comprises nonlinear spectral fitting of the informative portions of the sensed data signals with Lorentzian curves, Prony's method or Pisarenko method, and wherein the processor is operable to effect damped sinusoidal signal conditioning of the sensed data signals; and output a classification of the detected energy as at least part of the output signal based on said signal conditioning.

31. (canceled)

32. The analyzer as claimed in claim 27, wherein said processor is operable to classify said output signal based on one or more parameters selected from the group consisting of energy wave amplitude, phase. frequency and damping, coefficient.

33. The analyzer as claimed in claim 27, wherein said processor is operable to classify said compared data based on one or more input patient parameters selected from the group consisting of age, sex, weight, and smoker status.

34. The analyzer as claimed in claim 27, wherein said pulsed force generator includes a motor and selectively displaceable member, the motor being operable to activate the member in a reciprocal movement between a forward impact position, where said member is moved into engagement with said target, area to transmit said pulsed force thereto, and a rearward position wherein the member is moved to a position spaced from said target area, and wherein the motor is operable to activate the piston member in pulsed movement to provide said pulsed force as a pulsed impact force at said target area selected at between about 0.1 and 10N, and wherein said member is moved from said rearward position to said forward position, and then from said forward position to said rearward position at a repetition frequency selected between 0.1 and 10 hertz.

35. (canceled)

36. The analyzer as claimed in claim 27, wherein said processor is operable to classify the output signal whereby, signal classification parameters for the analyzer are pre-stored in memory, associating a multi-dimensional vector quantity to the sensed data signals, the vector quantity comprising pre-identified vector coordinates selected from the group consisting of signal amplitude, phase, frequency and damping factor, comparing the associated multi-dimensional vector quantity with one or more preselected signal classification parameters, and outputting said out signal based on the comparison.

37. The analyzer as claimed in claim 27, wherein said vibrations and/or return waves include a low frequency component selected at less than about 1000 Hz, preferably less than about 600 Hz, and most preferably between about 20 and 300 Hz, and wherein said sensor detects said low frequency component.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0059] Reference may now be had to the following detailed description taken together with the accompanying drawings in which:

[0060] FIGS. 1a and 1b illustrate schematically a portable analyzer for identifying and/or diagnosing a pulmonary or abdominal condition in a patient, in accordance with a preferred embodiment of the invention;

[0061] FIG. 2 illustrates schematically the use of the analyzer shown in FIG. 1a in the diagnosis of a potential pulmonary injury in the patient, in accordance with a preferred method of the invention;

[0062] FIG. 3 shows schematically a sectional view of the analyzer shown in FIG. 2;

[0063] FIG. 4 shows an enlarged cross-sectional view of the pulsed force generator and sensor assembly used in the analyzer of FIG. 2;

[0064] FIG. 5 illustrates the forward displacement of the energy generator piston in the output of a pulsed force into the patient's chest cavity;

[0065] FIG. 6 illustrates schematically the positioning of the pulsed force generator piston in a retracted position during the sensing of reflected return energy waves;

[0066] FIG. 7 illustrates a flow chart describing the operation steps of the analyzer shown in FIGS. 1a and 1b in accordance with a preferred method of operation;

[0067] FIG. 8 shows a flow chart illustrating the operation of the analyzer shown in FIG. 2 in detecting and analyzing reflected return energy sensed in the identification and/or confirmation of the presence or absence of a pulmonary disease and/or condition;

[0068] FIG. 9 illustrates graphically sample signal wave forms and spectrum of initially output and sensed pressure waves generated at the patient's colon, stomach, upper chest, liver, and muscle tissues;

[0069] FIG. 10 illustrates graphically a case where one Lorentzian curve provides adequately close fitting of a signal's Fourier power spectrum;

[0070] FIG. 11 illustrates graphically a case where three Lorentzian curves provide adequately close fitting of a signal's Fourier power spectrum;

[0071] FIG. 12 shows schematically a sample graphic output of an exemplary typical measured percussion and reflected energy signals corresponding to an abdominal air condition, and a generated analytical signal reconstructed from a single damped harmonic mode; and

[0072] FIG. 13 shows schematically a sample graphic output of an exemplary measured percussion and reflective signals from the patient's upper chest region, and a sample generated analytical signal reconstructed from three damped harmonic modes of the invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0073] The following description describes a preferred embodiment of the invention and is used for descriptive clarity and is not intended to limit the application and use of the invention.

[0074] Reference may be had to FIGS. 1a and 1b which illustrate a portable diagnostic device or analyzer 10 used in the diagnosis of a potential pulmonary injury or condition in a patient 6 (FIG. 2) in accordance with a preferred embodiment of the invention. As will be described, the analyzer 10 is operable to generate and impart mechanical disturbance as an output pressure wave (shown as 100 in FIG. 5) or force on the patient's chest 8; and then detect and analyze the surface vibrations or the return reflected wave (shown as 150 in FIG. 6) produced thereby.

[0075] As shown in FIGS. 1a and 1b, the analyzer 10 includes a pulsed force generator 12, for generating the output pulsed force, a sensor assembly 14 for detecting return wave energy 150 and converting the detected energy into electric signals, a processing assembly 16 and an output display 18 which is operable to output to a user, a visual graphic output 20 identifying the absence and/or presence of a potential pulmonary and/or abdominal condition under test. Optimally, the analyzer 10 may be further operational to output via the display 18 both instructions regarding the initial placement and operation of the analyzer 10, as well as directions as to possible treatment and patient care, having regard to any condition or trauma identified.

[0076] As shown in FIGS. 2 and 3, the analyzer 10 is adapted for portable use and includes an internal battery power supply 22 which is used to provide electrical power to the pulsed force generator 12, sensor assembly 14 and processing assembly 16. The processing assembly 16 electronically communicates with each of the pulsed force generator 12, sensor assembly 14 and output display 18 and, as shown in FIG. 4, includes memory 24 and a central processing unit (CPU) 26 for receiving and storing data signals from the sensor assembly 14 and the pulsed force generator 12.

[0077] FIGS. 1a and 4 show best the sensor assembly 14 as including a number of acoustic transducers 28. The transducers 28 are positioned in a circularly extending array about the periphery of an analyzer sensing head 30 which is adapted to be placed in direct juxtaposed contact against a selected target area A.sub.T of the patient's chest 8. Each of the acoustic transducers 28 is configured to detect and transmit to the CPU 26 data signals representative of sensed reflected return energy waves 150.

[0078] FIGS. 4 to 6 illustrate best the pulsed force generator 12 and its operation in generating at the target area A.sub.T on the patient's chest 8 the output pulsed force 100.

[0079] The pulsed force generator 12 includes an electro-mechanically displaceable piston 34 which is engagable with a plessimeter 35, a resiliently deformable biasing spring 36 and a selectively operable electro-magnet 38. As shown best in FIGS. 5 and 6, the piston 34 is reciprocally moveable within a piston chamber 40 between the forward impact position contacting the plessimeter 35 shown in FIG. 5, wherein the forward most end 42 of the piston 34 is moved into engaging contact there against, whilst the plessimeter 35 is pressed against chest 8 at the target area A.sub.T; and the retracted position shown in FIG. 6, where the piston 34 is moved with its forward end 42 spaced rearwardly therefrom.

[0080] When activated, the electro-magnet 38 operates to move the piston rearwardly against the bias of the spring 36 to the retracted position shown in FIG. 6. In operation of the analyzer 10, the electro-magnet 38 is selectively cycled to release the piston 34 so as to move rapidly under the force of the spring 36 into contact against the patient's chest 8, thereby generating at the target area A.sub.T, an output initial pulsed force 100 of between about 0.5 and 4N, and with a frequency of up to 10 Hz, and preferably 1 to 2 Hz. The pulsed force generator 12 is operable to produce and generate an output pulsed force 100 which is sufficient to generate at the target area A.sub.T one or more return energy waves 150 which are detectable by the analyzer sensor assembly 14, having regard to the particular pulmonary and/or abdominal condition under test.

[0081] Although not essential, in a most preferred mode of operation, the pulsed force generator 12 is operable to output to the patient 6 a preselected output pulsed force 100 and which optionally, may vary having regard to the specific target area A.sub.T of analyzer use on the patient's chest and/or abdomen.

[0082] The pulsed force generator 12 is preferably operable to impart at the target area A.sub.T a mechanical force having a preselected, and most preferably consistent magnitude. In an alternate embodiment, the analyzer 10 may allow for variable adjustment in the output force 100 applied to the patient 6, depending on subjective factors such as the user's body type, age, or weight, and/or depending on the specific use site of the analyzer 10.

[0083] Following the activation of the pulsed force generator 12, activation of the piston 34 and impart into the patient's chest 8 the output energy wave 100, the sensor assembly 14 is operated to detect one, and preferably a number of return energy waves 150 which are generated within the patient's chest 8 by the activation of the piston 34.

[0084] Most preferably, the acoustic transducers 28 electronically transmit signals to the CPU 26 and memory 24. The CPU 26 operates with the memory 24 to effect signal detection conditioning and damping, and to output via the display 18 a visual signal analysis. In this regard, the analyzer 10 may be used in the automated generation and analysis to effect the unbiased diagnosis of pulmonary trauma or disease. Furthermore, by the use of consistent automation, the analyzer 10 advantageously eliminates subjective factors associated with the conventional, manual percussion diagnosis, allowing the analyzer 10 to be used in the field by non-medical or casually-trained professionals.

[0085] As shown in FIG. 7, the analyzer 10 is operated to perform each of the steps of producing or generating the initial output pulsed force 100 or percussion signal, and therefore sensing and detecting the return energy waves 150 which are produced by the initial wave energy 100 reflecting and/or transitioning through patient organs, fluids and/or bones. The step of generating the output pulsed force 100 or percussion signal, and detecting and performing initial processing of the reflected return energy waves 150 as pulmonary or abdominal percussion signals produced thereby, most preferably includes a number of sub-steps. These include effecting the initial mechanical disturbance or force transfer of a selected pulsed force at the target area A.sub.T surface of the chest 8 of the patient 6 under examination; detecting the return energy waves 150 reflected as percussion signals generated internally by the transfer of output pulsed force 100 to patient organs thereby; converting the detected return energy waves 150 into electrical signals; and using the CPU 26, selecting an informative (carrying useful diagnostic information) portion of the signal.

[0086] As described, the initial percussion is performed using portable pulmonary injury diagnosis analyzer 10, with the piston 34 operable to impact the target area A.sub.T of the patient's chest 8 with a preselected pulsed force.

[0087] The initial step of producing the mechanical disturbance at the surface of the patient's chest 8 generates a response from the underlying organs and tissues, which effects the generation of return and reflected wave energy 150. The return wave energy signal will have a natural resonant frequency(ies), which is(are) dependent on organ/tissue anatomy and physical condition. Most preferably the return wave energy signal includes a low frequency vibration or wave energy component having a frequency less than 1000 Hz, preferably less than 600 Hz, and preferably ranging from about 20 to 100 Hz. The internally generated signals are detected with the acoustic transducers 28, and converted to analog percussion signals.

[0088] The sub-step of detecting return wave energy signals produced by the output disturbance preferably includes receiving and detecting the reflected return energy waves 150 with the sensors assembly 14, and recording the generated signal in the device memory 24. The return energy waves 150 typically will consist of vibrations of external and/or internal organs and tissues produced by the mechanical disturbance.

[0089] As shown in FIG. 8, preferably the analyzer CPU includes stored programme instructions for operations by the use of damped harmonic analysis of the sensed return wave/vibration in the automated diagnosing of pulmonary and abdominal injuries. In a preferred mode, the analyzer 10 performs the step of generating the output mechanical force or disturbance to the patient's chest 8 by percussion to effect transmission of the predetermined and preselected pulsed force 100. The exertion of pulsed force to the patient's chest cavity 8 in turn results in the generation vibrations 150 by the subject's organs as the input disturbance 100 passes therethrough. According to a preferred method, the vibrations/reflected waves 150 generated by organs and/or tissues in the patient's chest 8 or abdomen in response to the input mechanical disturbance 100 are detected by the sensor assembly 14 and transmitted to the CPU 26 for initial processing.

[0090] In the CPU 26 conditioning the signal is further performed. Preferably, data representing the sensed vibration/wave signal as a sum of one or more damped sinusoids (herein damped harmonic modes (DHMs)); and the DHMs are used to classify the signal and output signal classification parameters. More preferably, the analyzer memory 24 stores a number of separate predetermined signals which are representative of a signal indicating a normal or optimum physical state, and/or signals which are indicative of a compromised stressed or injured state which for example could represent as particular trauma, disease or other condition. Depending upon the results of the comparison, the CPU 26 may thus be used to activate the output display 18 to illustrate to the user a particular visual graphic display 20 correlated to the most proximate condition pre-stored within the memory 24.

[0091] The step of conditioning the detected return energy wave signal preferably includes the sub-steps of: selecting an informative portion of the signal; and preparing the signal for the damped harmonic analysis. Selecting the informative portion of the signal includes examination of the digitized signal in order to identify one or more of its parts containing information about the response of the patient's body to the percussion event. Once identified, the informative portion is separated from the rest of the signal, and the separated portion is used as “the signal” during subsequent processing steps. Typically the identification of the information portion of the signal is performed by filtering and/or signal truncation. In a simplified method, the detected return energy wave is filtered with respect to pre-identified known background parameters.

[0092] The sub-step of preparing the signal for damped harmonic analysis may include increasing or decreasing the number of samples in the signal, windowing the signal, removing direct current (DC) offset from the signal, and filtering the signal. Filtering of the signal may remove at least one of undesirable low frequency components and undesirable high frequency components. The conditioned signal may be either kept in the computer memory 24 for further processing or recorded externally using a hard drive, flash memory, or any other suitable storage medium (not shown).

[0093] The step of representing the signal as a sum of one or more DHMs is preferably carried out by means of an appropriate damped harmonic analysis algorithm. Each DHM is a damped sinusoid - an analytical function completely defined by its four parameters: amplitude, phase, frequency, and damping. The combination of these parameters for all DHMs representing the signal provides complete information about the signal and may be used, either fully or in its part, for the signal classification and diagnostic purposes. Examples of possible damped harmonic analysis algorithms include, but are not limited to, the nonlinear spectral fitting with multiple Lorentzian curves, Prony's damped harmonic analysis algorithm and its derivatives, the matrix pencil method, signal approximation by single or multiple DHMs derived from the spectral envelope, the Pisarenko method, the approach based on finding coordinates of the poles of the complex Laplace transform of the signal, or any other appropriate algorithm capable of representing the signal as a sum of damped sinusoids.

[0094] The step of representing the signal as a sum of one or more DHMs may be effected by either analog or digital decomposition of the signal into a sum of one or more damped sinusoids defined by their respective amplitudes, phases, frequencies and damping coefficients. This decomposition may be either exact or approximate. In the case of approximate decomposition, the error may be described in terms of the difference between the actual signal and the signal represented as a sum of one or more DHMs. The error may be evaluated to judge an accuracy of the damped harmonic analysis performed on each particular signal.

[0095] If a Fourier spectrum-based algorithm is used for the step of representing the signal as a sum of one or more DHMs, a Fourier Transform may be performed on the signal, producing a frequency spectrum. The frequency spectrum may further be smoothed. FIG. 9 shows exemplary conditioned waveforms and spectra of sensed return energy wave signals generated from different chest and abdominal areas. Peaks of the frequency spectrum may then be determined. A suitable curve-fitting procedure may be used to approximate the determined frequency peaks. Preferably but not exclusively, a set of Lorentzian curves may be used to carry out the fitting process, with each curve representing the spectrum of an independent exponentially damped harmonic oscillator. The amplitudes, peak frequencies f and widths at half height (WHH) w of the Lorentzian curves may be determined according to a selected curve-fitting procedure, thereby concluding the decomposition of the signal into a set of DHMs. (The damping constant of the oscillator corresponding to each DHM b=w/2.) FIGS. 10 to 13 illustrate this approach by examples of two typical percussion signals.

[0096] FIG. 10 provides an illustrative example of a typical Fourier spectrum of an abdominal air signal (the solid line) recorded by the system. In this case, one Lorentzian curve (the dashed line) with the peak frequency f and the WHH w suffices to provide adequately close fitting of the main spectral peak.

[0097] FIG. 11 provides an illustrative example of a typical Fourier spectrum of a pre-identified normal upper chest signal recorded by the system and stored in memory 24. In this case, three Lorentzian curves (the dotted lines) with peak frequencies f.sub.1-3 and the associated WHH w.sub.1-3 provide adequately close fitting of the signal spectrum.

[0098] FIG. 12 shows that the analytical signal reconstructed from a single DHM (bottom graph) closely matches the original abdominal air signal recorded by the system (top graph). FIG. 13 shows that the signal reconstructed from three DHMs (bottom graph) closely matches the original upper chest signal (top graph) recorded by the system. In this case, the amplitudes of the DHMs were derived from the amplitudes of the corresponding Lorentzian curves, while phases were found by a separate reconstruction program. In the preferred embodiment, all four parameters for each DHM may be determined automatically by the damped harmonic analysis algorithm.

[0099] If the spectral envelope-derived single damped harmonic mode is used for the step of representing the signal as a sum of one or more DHMs, then the signal is approximated by a single damped sinusoid that is derived from the parameters of the spectral envelope. According to this approach, the damped frequency Ω is equal to the peak frequency, and the damping factor b equals half the WHH of the spectral envelope. For example, each set of waveforms and spectra shown in FIG. 9 has a distinct (relative to the other waveforms and spectra) combination of Ω and WHH of the single major spectral peak, which may be used as a signal classifier either alone or combined with additional spectral or time-domain parameters.

[0100] If Prony's analysis or any of its derivative algorithms, or the matrix pencil method is used for the step of representing the signal as a sum of one or more DHMs, then the frequencies, damping factors, amplitudes, and phases may be computed for the individual DHMs.

[0101] If a complex Laplace transform is used for the step of representing the signal as a sum of one or more DHMs, then the real and imaginary coordinates of the poles of the complex Laplace transform of the signal may be computed for the one or more DHMs. The frequencies may then be identified as the real coordinates of the poles of the complex Laplace transform, while the damping factors may be identified as the imaginary coordinates of the poles of the complex Laplace transform. Alternatively, the Pisarenko method or any other damped harmonic analysis algorithm may be used for the step of representing the signal as a sum of one or more DHMs.

[0102] The fourth step of classifying the signal and outputting at least one of the signal classification parameters preferably includes the sub-steps of: computing signal classification parameters for the DHMs; associating a multidimensional vector quantity with each signal, the vector coordinates being the values of the signal classification parameters; evaluating the vector quantities of each signal and classifying the signal as one or more of “tympanic,” “resonant,” and “dull”; or using a different gradation scale; and outputting at least one of the diagnostic classification and the signal classification parameters in numerical, graphical, audible, or other form.

[0103] Signal classification parameters for each DHM include amplitude, phase, frequency, and damping factor. Additional classification parameters for each DHM may include damped frequency Ω and quality factor Q. Supplementary classification parameters derived from the signal waveform may include number of oscillations, number of zero crossings, zero-crossing rate, temporal envelope width at selected threshold levels, and other time- and frequency-domain parameters. The step of computing signal classification parameters from the DHMs may be carried out according to any acceptable method known in the art. The results of the damped harmonic decomposition of the signal may thereby be processed to reconstruct the main modes of the signal. The process produces quantitative information that may be used as a distinctive classifier of percussion signals and may be represented in a graphical, numerical, audible, or other form to facilitate interpretation by an examiner.

[0104] The sub-step of associating a multidimensional vector quantity with each signal is carried out based on the fact that the vector coordinates correspond to the values of the signal classification parameters. The sub-step of evaluating the vector quantities of each signal and classifying the signal results in attribution of the signal to one or more of the three conventional categories of acoustic signals, “tympanic,” “resonant,” and “dull”. Such simple classification, for example, could be useful for rapid identification of severe pulmonary conditions, such as pneumothorax, where a “tympanic” signal detected in the upper chest region instead of a typically observed “resonant” one would indicate the presence of anomaly. A more refined gradation of the signals, based on the values of the above vector quantities, is also possible with this method and can be used to build diagnostic images in cases when detailed percussion examinations are performed. The sub-step of outputting the signal classification parameters may be carried out by a numerical or text display, a graphical display, an audible output, or any other form of output intelligible to an examiner.

[0105] The analysis and classification of pulmonary and abdominal percussion signals utilizes a model based on the general concept of a multi-mode exponentially damped harmonic oscillator that, besides the abdomen, may be applied to both upper chest and lower chest percussion. The system is preferably configured to decompose an arbitrary percussion signal into a sum of a small number of damped sinusoids called here damped harmonic modes (DHM) with corresponding amplitudes, frequencies, phases, and damping factors. These parameters combined fully define the original signal and therefore can be used for classification purposes. For example, a combination of two of the parameters associated with each DHM, namely, quality factor Q and the damped frequency Ω, has been experimentally found to have high diagnostic classification potential. As shown in FIGS. 10 to 13, a general percussion signal may be represented as a sum of a small number of DHMs, each representing a particular oscillating subsystem. Since only a few DHMs are required to represent the signal, this is an advantage over conventional Fourier analysis where a much larger number of infinite-duration sinusoidal waves is used for the same purpose. Accordingly, real chest percussion signal may be synthesized from several damped oscillator modes that were extracted from it according to embodiments of the invention.

[0106] It is envisioned that the analyzer 10 of the present invention may be used in a variety of differing operational modes and/or applications. In one possible mode of operation, the actuator 12 may be operated to transmit pulsed or disturbance forces to the patient using comparatively low frequency signals, as for example, at a rate of 1 to 100 beats or impact forces per minute, whilst the sensor assembly 14 may be selected to detect return percussion signals in a less than 0 to 600 Hz range, which correlates to either a natural frequency or harmonic of the patient's internal organs and/or body parts. The analyzer 10 may further be operable to utilize a damped harmonics mode and/or analysis to determine the return signal frequency, amplitude, phase, and/or other signal classifications.

[0107] More preferably, the device processing assembly 16 is selected to separate individual signals from a mixed return signal and/or provide signal recognition and referencing correlated to selected patient organs. The processing assembly 16 may furthermore store a map of base signals used to parse, recognize and/or analyze mixed return signals generated in the patient's chest/torso. It is to be appreciated that by the use of handheld portable device 10, the device may be operated in a non-evasive manner as a method of determining the condition of pulmonary function for human, veterinary or other animal use.

[0108] In other non-limiting constructions, however, the apparatus could operate to generate a pressure wave by the use of air and/or gas pressure, electric stimulation, or other physical impact devices, while the preferred embodiment describes the sensor assembly 14 as including acoustic transducers 28, the invention is not so limited. In another possible embodiment, the signals generated by return waves 150 may be detected using other acoustic and/or audio sensors which convert the detected energy into sensed electric data signals. Such sensors would include other types of non-contact sensors, such as air microphones, laser vibrometers, and other suitable non-contact vibration or pressure sensors. Alternatively, the return energy waves 150 may be detected using direct contact pressure sensors or other suitable apparatus, including, for example, a microphone embedded into a stethoscope head, a contact accelerometer, a piezofilm sensor, or any other suitable contact vibration or pressure sensor. The sub-step of converting the generated return energy wave signal into electrical signals also include amplifying and preconditioning, for example, by analog filtering, an analog return percussion signal. In an alternate embodiment, the signal may be preprocessed, as for example, by digitizing the output percussion signal and/or one or more detected return energy waves.

[0109] Although the preferred embodiment describes the analyzer 10 as having an internal processor assembly 16, the detected signals alternating either transferred to computer memory which is external for further processing, or recorded using a hard drive, flash memory, or any other suitable storage medium.

[0110] Although the detailed description describes an electro-magnetically actuable piston 34 as used to generate the output energy force 100, the invention is not so limited. Other force inducing constructions may also be used. Such generators include other moveable impactors activated by pneumatic, hydraulic, electromechanical, or electromagnetic means. Alternatively, the analyzer 10 may be operable to effect percussion using pressurized fluid sources, such as pneumatic impact or electromechanical sources operable to provide a pulsed force designed to reproduce the effect of percussion impact without the moving impactor.

[0111] While the detailed description describes the best mode, the invention is not limited to the described embodiment. In other non-limiting applications, the method and apparatus may also be used in a variety of applications including without restriction, in veterinary diagnostics; material analysis; change of state analysis in multiphase materials; and viscosity analysis of liquids, jells and semisolids.

[0112] A person skilled in the art will recognize from the previous detailed description and from the figures and claims, modifications and changes can be made to the preferred embodiments of the invention without departing from the scope of this invention defined in the following claims.