Multi-parameter based diagnosis of lung disease

Abstract

A method and apparatus by which deceases are identified uses computer analysis of sound signals that are picked up from various locations on the chest walls of a subject by a modified stethoscope. The modification includes a small microphone in one of the hoses of the stethoscope. Signals from the microphone are input to a computer such as a personal computer or PC for processing. The computer extracts from these signals features which are dominant for particular diseases. A classifier classifies these features, determines if the lungs are diseased, and identifies the disease.

1 . A method of identifying a disease that produces sounds that are characteristic of the disease, comprising:
 subjecting the sounds to a plurality of multi-dimensional, multi-parameter feature space analyses to identify a plurality of multi-dimensional features associated with the sounds; and   selecting from among the plurality of multi-dimensional features, only dominant features that identify the disease by using discriminant analysis.   
   
   
       2 . A method according to  claim 1 , wherein the plurality fo multi-dimensional features associated with the sounds are selected from the group consisting of: linear predictive coding parameters, Fourier coefficients, partial autocorrelation coefficients, Cepstrum coefficients, and autocorrelation coefficients. 
   
   
       3 . A method according to  claim 1 , wherein the discriminant analysis is performed using neural network analysis of the dominant features for identifying the disease. 
   
   
       4 . A method according to  claim 1 , including projecting a plurality of the multi-dimensional features for the sounds into representations of the features, selected from among the representations of the features, those representation that provide relatively good correlation to the disease as compared to representations that provide relatively poor correlation to the disease, and identifying the disease by selection at least one pair of the projected features that separate clusters of samples in the representations that provide relatively good correlation to the disease. 
   
   
       5 . A method according to  claim 1 , including storing in a computer, a data base of dominant features from pre-recorded sounds from organs with at least one disease that produces sounds that are characteristic of the disease, during a training phase, obtaining sounds from an organ of a subject to be tested, conditioning the sounds to create a signal that can be subjected to analysis in the computer, and using the signal as the sounds to be subjected to the plurality of multi-dimensional, multi-parameter feature space analyses to identify the plurality of multi-dimensional features associated with the sounds. 
   
   
       6 . An apparatus for identifying a disease that produces sounds that are characteristic of the disease, the apparatus comprising:
 a stethoscope for acquiring the sounds from the subject;   a transducer operatively connected to the stethoscope for converting the sounds to signals; and   a computer-based analyzer for analyzing the signals to identify the disease, a computer-based analyzer identify the disease by subjecting the signals to multi-dimensional, multi-parameter feature space analysis to identify a plurality of multi-dimensional features associated with the sounds, and selecting from among the plurality of multi-dimensional features, only dominant multi-dimensional features that identify the disease by using discriminant analysis.   
   
   
       7 . An apparatus according to  claim 6 , wherein the stethoscope includes at least one hose for conveying the sounds, and the transducer comprises a microphone connected to the hose for picking up sounds in the hose. 
   
   
       8 . An apparatus according to  claim 6 , wherein the computer-based analyzer comprises a computer that is programmed for analyzing the signal to identify the disease using a plurality of modeling selected from the group consisting of: linear predictive coding parameters, Fourier coefficients, partial autocorrelation coefficients, Cepstrum coefficients, and autocorrelation coefficients. 
   
   
       9 . An apparatus according to  claim 6 , wherein the computer-based analyzer comprises a computer that is programmed with a program for analyzing the signal to identify a lung disease. 
   
   
       10 . An apparatus for identifying a disease in a subject, the disease producing sounds that are characteristic of the disease, the apparatus comprising:
 means for acquiring the sounds; and   a computer-based analyzer for analyzing the sounds to identify the disease by subjecting the sounds to multi-dimensional, multi-parameter feature space analysis to identify a plurality of multi-dimensional features associated with the sounds, and selecting from among the plurality of multi-dimensional features, only dominant multi-dimensional features that identify the disease by using discriminant analysis.   
   
   
       11 . An apparatus according to  claim 10 , wherein the computer-based analyzer comprises a computer that is programmed with a program for analyzing the sounds to identify the disease, the program using at least one of: Linear Predictive Coding Parameters, Fourier Coefficients, Partial Autocorrelation Coefficients, Cepstrum Coefficients, and Autocorrelation Coefficients. 
   
   
       12 . An apparatus according to  claim 10 , wherein the computer-based analyzer comprises a computer that is programmed with a program for analyzing the signal to identify a lung disease. 
   
   
       13 . An apparatus according to  claim 10 , wherein the program plots a plurality of two-dimensional representations of coefficients of modeling for the sounds, and selects features of the representations that provide relatively good correlation to the lung disease as compared to features that give relatively poor correlation to the lung disease, and identifies the disease by selection at least one pair of coefficients that separate clusters of samples in the representation for the features that provide relatively good correlation to the disease.