Multi-modal cardiac diagnostic decision support system and method

• US 6,572,560 B1
• Filed: 09/25/2000
• Issued: 06/03/2003
• Est. Priority Date: 09/29/1999
• Status: Expired due to Term

• Current Assignee: 3M Innovative Properties Company (3M Company)
• Original Assignee: Zargis Medical Corp.
• Inventors: Watrous, Raymond L., Reichek, Nathaniel
• Primary Examiner(s): Schaetzle, Kennedy
• Assistant Examiner(s): DROESCH, KRISTEN L
• Application Number:
  US09/670,284
• Time in Patent Office: 981 Days
• Field of Search: 128/906, 128/920, 128/923-925, 600/300-301, 600/508, 600/513, 600/528, 600/586
• US Class Current: 600/528
• CPC Class Codes: A61B 5/0002 Remote monitoring of patients using telemetry, e.g. transmission of vital signals via a communication network A61B5/07 takes precedence; transmission systems for measured values G08C; transmission H04B; transmission of digital information H04L; wireless communication networks H04W
• CPC Class Codes: A61B 5/30 Input circuits therefor
• CPC Class Codes: A61B 5/308 for electrocardiography [ECG]
• CPC Class Codes: A61B 5/726 using Wavelet transforms
• CPC Class Codes: A61B 5/7264 Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems neural networks per se G06N3/00; expert systems per se G06N5/00
• CPC Class Codes: A61B 5/7267 involving training the classification device
• CPC Class Codes: A61B 5/7285 for synchronising or triggering a physiological measurement or image acquisition with a physiological event or waveform, e.g. an ECG signal A61B6/5205, A61B6/541, A61B8/5207, A61B8/543 take precedence
• CPC Class Codes: A61B 7/00 Instruments for auscultation
• CPC Class Codes: A61B 7/04 Electric stethoscopes microphones, acoustic transducers therefor H04R
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems
• CPC Class Codes: Y10S 128/92 Computer assisted medical diagnostics
• CPC Class Codes: Y10S 128/925 Neural network

Abstract

A method for extracting features from cardiac acoustic signals includes the steps of obtaining a cardiac acoustic signal, and extracting physiologically significant features from the cardiac acoustic signal using a neural network. A method for evaluating cardiac acoustic signals includes the steps of obtaining a cardiac acoustic signal, analyzing the cardiac acoustic signal with a wavelet decomposition to extract time-frequency information, and identifying basic heart sounds using neural networks applied to the extracted time-frequency information. A method for determining cardiac event sequences from cardiac acoustic signals includes the steps of obtaining a cardiac acoustic signal, and processing a sequence of features extracted from the cardiac acoustic signal by a probabilistic finite-state automaton to determine a most probable sequence of cardiac events given the cardiac acoustic signal. A method for extracting findings from cardiac acoustic signals includes the steps of obtaining a cardiac acoustic signal, processing the cardiac acoustic signal to determine a most probable sequence of cardiac events given the cardiac acoustic signal, and extracting the clinical findings from the sequence of cardiac events. A method for determining a status of heart murmurs includes the steps of obtaining a cardiac acoustic signal, detecting a murmur, if any, from the cardiac acoustic signal, and determining whether the murmur is one of functional and pathological based upon expert rules.

Claims

1. A system for determining the presence of cardiac diseases, comprising:

a sensor adapted to obtain a cardiac acoustic signal of a patient;

an acoustic signal processor adapted to process the cardiac acoustic signal to find evidence of cardiac diseases;

an input device adapted to receive data corresponding to a medical record of the patient; and

a diagnostic decision support circuit adapted to provide a diagnostic recommendation corresponding to a probability of the cardiac diseases being present in the patient, using Bayesian networks comprising a unified model that integrates statistical signal processing with probabilistic reasoning to provide recommendations based on the results of the acoustic signal processing and on the medical record data.

2. The system according to claim 1, wherein said support circuit is further adapted to determine whether to refer the patient for echocardiography, based upon the diagnostic recommendation.

3. The system according to claim 1, wherein said support circuit is further adapted to determine whether to refer the patient for further treatment by a cardiac specialist, based upon the diagnostic recommendation.

4. The system according to claim 1, wherein the acoustic signal processor comprises:

a feature extraction module for extracting features from the cardiac acoustic signal, wherein the extracted features comprise physiologically significant events; and

a probabilistic finite-state automaton for processing a sequence of extracted features to determine a most probable sequence of cardiac events given the cardiac acoustic signal.

5. The system according to claim 4, wherein the feature extraction module performs a wavelet decomposition of the cardiac acoustic signal to extract time-frequency information, and processes the extracted time-frequency information using a neural network to extract cardiac acoustic features comprising physiologically significant events.

6. The system according to claim 4, wherein the physiologically significant events correspond to at least one of basic heart sounds and components of the basic heart sounds.

7. The system according to claim 4, wherein the probabilistic finite-state automaton comprises a hidden markov model.

8. The system according to claim 4, further comprising a module for automatically determining clinical findings from the sequence of cardiac events.

9. A program storage device readable by a machine, tangibly embodying a program of instructions executable by the machine to perform method steps for determining the presence of cardiac diseases, the method steps comprising:

obtaining a cardiac acoustic signal of a patient;

signal processing the cardiac acoustic signal to find evidence of cardiac diseases;

receiving data corresponding to a medical record of the patient; and

providing a diagnostic recommendation corresponding to a probability of the cardiac diseases being present in the patient using Bayesian networks comprising a unified model that integrates statistical signal processing with probabilistic reasoning to provide recommendations based on the results of the signal processing and on the medical record data.

10. The program storage device according to claim 9, further comprising instructions for performing the step of determining whether to refer the patient for echocardiography, based upon the diagnostic recommendation.

11. The program storage device according to claim 9, further comprising instructions for performing the step of determining whether to refer the patient for further treatment by a cardiac specialist, based upon the diagnostic recommendation.

12. The program storage device according to claim 9, wherein the instructions for performing the step of signal processing the cardiac acoustic signal comprise instructions for performing the steps of:

extracting features from the cardiac acoustic signal, wherein the extracted features comprise physiologically significant events; and

processing a sequence of extracted features using a probabilistic finite-state automaton to determine a most probable sequence of cardiac events given the cardiac acoustic signal.

13. The program storage device according to claim 12, wherein the instructions for performing the step of extracting features from the cardiac acoustic signal comprise instructions for performing the steps of:

performing a wavelet decomposition of the cardiac acoustic signal to extract time-frequency information; and

processing the extracted time-frequency information using a neural network to extract cardiac acoustic features comprising physiologically significant events.

14. The program storage device according to claim 12, wherein the physiologically significant events correspond to at least one of basic heart sounds and components of the basic heart sounds.

15. The program storage device according to claim 12, wherein the probabilistic finite-state automaton comprises a hidden markov model.

16. The program storage device according to claim 12, further comprising instructions for performing the step of automatically determining clinical findings from the sequence of cardiac events.

17. A method for determining the presence of cardiac diseases, comprising the steps of:

obtaining a cardiac acoustic signal of a patient;

signal processing the cardiac acoustic signal to find evidence of cardiac diseases;

receiving data corresponding to a medical record of the patient; and

providing a diagnostic recommendation corresponding to a probability of the cardiac diseases being present in the patient, using Bayesian networks comprising a unified model that integrates statistical signal processing with probabilistic reasoning to provide recommendations based on the results of the signal processing and on the medical record data.

18. The method according to claim 17, further comprising the step of determining whether to refer the patient for echocardiography, based upon the diagnostic recommendation.

19. The method according to claim 17, further comprising the step of determining whether to refer the patient for further treatment by a cardiac specialist, based upon the diagnostic recommendation.

20. The method according to claim 17, wherein the step of signal processing the cardiac acoustic signal comprises the steps of:

extracting features from the cardiac acoustic signal, wherein the extracted features comprise physiologically significant events; and

processing a sequence of extracted features using a probabilistic finite-state automaton to determine a most probable sequence of cardiac events given the cardiac acoustic signal.

21. The method according to claim 20, wherein the step of extracting features from the cardiac acoustic signal comprises the steps of:

performing a wavelet decomposition of the cardiac acoustic signal to extract time-frequency information; and

processing the extracted time-frequency information using a neural network to extract cardiac acoustic features comprising physiologically significant events.

22. The method according to claim 20, wherein the physiologically significant events correspond to at least one of basic heart sounds and components of the basic heart sounds.

23. The method according to claim 20, wherein the probabilistic finite-state automaton comprises a hidden markov model.

24. The method according to claim 20, further comprising the step of automatically determining clinical findings from the sequence of cardiac events.