Device for analyzing auscultation sounds, in particular respiratory sounds

• US 6,699,204 B1
• Filed: 03/05/2002
• Issued: 03/02/2004
• Est. Priority Date: 03/24/1999
• Status: Expired due to Fees

• Current Assignee: Georges Kehyayan, Ivan, Jointly Kehayoff, Ivan Kehayoff
• Original Assignee: Georges Kehyayan, Ivan Kehayoff
• Inventors: Roux, Bruno, Kehayoff, Ivan, Kehyayan, Georges, Avedissian, Christian, Lenoir, Gerard
• Primary Examiner(s): Nasser; Robert L.
• Assistant Examiner(s): MALLARI, PATRICIA C
• Application Number:
  US09/937,424
• Time in Patent Office: 728 Days
• Field of Search: 600/300,529,533,586 381/71.1,94.1,94.2,94.3
• US Class Current: 600/533
• CPC Class Codes: A61B 7/003 Detecting lung or respiration noise

Abstract

The invention concerns a device for analysing medical sounds, in particular respiratory sounds, comprising an input receiving a sound signal sampled in intensity levels each associated with a selected time, and storage means (2, 4, 5) comprising a processing module for evaluating, in cooperation with computing means (3), a set of transformed intensity levels, each associated with a predetermined sound frequency. It further comprises an output (26) connected to the storage means (2) for delivering each transformed intensity level in correspondence with an associated frequency, and means for representing (6, 90, 91) intensity levels transformed on the basis of frequencies, to obtain a spectral representation of the auscultation sound. The invention is characterised in that the processing module is further capable of evaluating a set of transformed intensity levels for each selected time, and the storage means (2) store each intensity level transformed on the basis of an associated frequency, and of a selected time, thereby providing a three-dimensional representation in time and in frequencies of the auscultation sound.

Claims

1. A device for analyzing mediate-auscultation noises, in particular respiratory noises, comprising:

an input for receiving a time-domain signal (S(t)) representative of an acoustic noise, sampled in a multiplicity of intensities (I(m)) each associated with a chosen instant (m),

storage means including, a processing module configured to work together with calculating means with a view to evaluating a set of transformed intensities (I_T(f_k)), each associated with a predetermined audio frequency (f_k), while the storage means are further configured to store said transformed intensities in memory, and

an output (26) linked, on the one hand, to the storage means in order to deliver each transformed intensity in correspondence with an associated frequency, and, on the other hand, to means for representing the transformed intensities as a function of the frequencies, with a view to obtaining a spectral representation of the auscultation noise, characterized in that the processing module is configured to evaluate a phase for each intensity (Φ

[I(m)]), as a function of a difference between the associated chosen instant (m) and a chosen start instant (m₀), with a view to attributing, to each intensity, a complex value (I_c(n)) and its conjugate value (I_c*(n)),

in that the processing module is further configured to evaluate a set of transformed intensities (I_T(m, f_k)) for each chosen instant (m), by producing, for each frequency (f_k), a sum over a multiplicity of current instants (n) of terms resulting from a product involving, for each current instant (n),

a) the complex intensity (I_c(m+n)) associated with the sum of the chosen instant (m) and of the current instant (n),

b) the conjugate complex intensity (I_c*(m−

n)), associated with the difference in the chosen instant and in the current instant, and

c) a complex exponential of a pure imaginary product of the frequency (f_k) and the current instant (n),

and in that the storage means are configured to store in memory each transformed intensity (I_T) as a function of an associated frequency (f_K), on the one hand, and of a chosen instant (m), on the other hand,

which makes it possible to obtain a three-dimensional representation in terms of time (m) and of frequencies (f_k) of the auscultation noise, while preserving a real-time scale.

2. The device as claimed in claim 1, characterized in that the storage means further comprise chosen coefficients (h(n)), for weighting of the terms of said sum, each associated with a current instant.

3. The device as claimed in claim 2, characterized in that the auscultation noise to be analyzed is an adventitious noise from among a set comprising sibilants, severe crepitation and mild crepitation, and in that the storage means include one set of weighting coefficients (h(n)) per type of noise to be analyzed.

4. The device as claimed in any one of the preceding claims, characterized in that it includes visual-display means of the type comprising a monitor and/or a printer while the representation means include a graphics-representation module, able to work together with the visual-display means in order to represent said transformed intensities (I_T(m, f_k)) graphically, as a function of the associated frequencies (f_k), on the one hand, and of the chosen instants (m), on the other hand, forming three-dimensional spectra in terms of time and of frequencies.

5. Device as claimed in claim 2, characterized in that it further includes input means able to work with the representation means in order to refine the time and frequency scales and/or in order to refine the choice of coefficients (h(n)) for weighting the terms of said sum.

6. Device as claimed in claim 1, characterized in that it includes a respiratory-rate probe able to define said chosen start instant (m₀).

7. Device as claimed in claim 1, characterized in that said storage means include a memory area for holding said transformed intensities (I_T), as a function of the frequencies (f_k) and of the associated chosen instants (m).

8. Device as claimed in claim 1, characterized in that said storage means include a memory area for holding said multiplicity of intensities (I_T), as a function of the associated chosen instants (m).

9. Device as claimed in claim 7 or 8, characterized in that it includes a hard disk and/or a removable medium for storing the content of said memory area.

10. Device as claimed in claim 7, characterized in that it further includes a communications interface for transmitting the content of the memory area to a remote station and/or in order to receive samples of time-domain signals (S(t)), representative of auscultation noises.

11. Device as claimed in claim 7, characterized in that it includes a comparison module, able to work together with said memory area in order to compare a number of spectral representations of auscultation noises, with different chosen start instants (m₀), with a view to detecting a progressive change in said auscultation noises.

12. Device as claimed in claim 1, characterized in that it includes an electronic stethoscope equipped with at least one audio sensor configured to convert acoustic noise into a time-domain electrical signal (S(t)), as well as an interface equipped with a sampler (Ech), connected to the sensor and linked furthermore to said input.

13. Device as claimed in claim 12, characterized in that the electronic stethoscope includes means for frequency filtering of the time-domain signal at the output of the sensor with a view to limiting the frequency components of the time-domain signal to a useful spectral band.

14. A method of analyzing mediate-auscultation noises respiratory of the type comprising the following stages:

a) sampling a time-domain signal (S(t)), representative of an acoustic noise, in a multiplicity of intensities (I(m)) each associated with a chosen instant (m),

b) evaluating a set of transformed intensities (I_T(f_k)), each associated with a predetermined audio frequency (f_k),

c) storing said transformed intensities in memory, and

d) delivering each transformed intensity in correspondence with an associated frequency, with a view to obtaining a spectral representation of the auscultation noise, characterized in that stage a) further includes the choice of a start instant (m₀), in that stage b) comprises the following operations;

b1) evaluating a phase for each intensity (Φ

[I(m)]), as a function of a difference between the associated chosen instant (m) and the start instant (m₀), with a view to attributing, to each intensity, a complex value (I_c(n)) and its conjugate value (I_c*(n)),

b2) evaluating a product involving, for a current instant (n);

the complex intensity (I_c(m+n)) associated with the sum of the chosen instant (m) and of the current instant (n),

the conjugate complex intensity (I_c*(m−

n)) associated with the difference in the chosen instant and in the current instant, and

a complex exponential of a pure imaginary product of the frequency (f_k) and the current instant (n),

b3) evaluating a set of transformed intensities (I_T(m, f_k)) for each chosen instant (m), by forming, for each frequency (f_k), a sum over a multiplicity of current instants (n) of terms resulting from the products obtained at operation b2),

in that, in the course of stage c), each transformed intensity (I_T) is stored in memory as a function of an associated frequency (f_k), on the one hand, and of a chosen instant (m), on the other hand,

and in that, in the course of stage d), each transformed intensity is delivered in correspondence with an associated frequency (f_k), on the one hand, and with a chosen instant (m), on the other hand, which makes it possible to obtain a three-dimensional representation in terms of time (m) and of frequencies (f_k) of the auscultation noise, while preserving a real-time scale.

15. A representation of transformed intensities in terms of time and of frequencies of a mediate-auscultation noise, which is obtained by an analysis method as claimed in claim 14.