Using speech analysis to assess a speaker's physiological health

• US 9,070,357 B1
• Filed: 05/11/2012
• Issued: 06/30/2015
• Est. Priority Date: 05/11/2011
• Status: Expired due to Fees

• Current Assignee: Brian K. Buchheit
• Original Assignee: Brian K. Buchheit
• Inventors: Buchheit, Brian K., Kennedy, Peter
• Primary Examiner(s): Han, Qi
• Application Number:
  US13/469,616
• Time in Patent Office: 1,145 Days
• Field of Search: 704/270, 704/246, 704/250, 704/251, 704/255, 704/270.1, 704/274
• US Class Current: 1/1
• CPC Class Codes: G06Q 10/10 Office automation; Time management
• CPC Class Codes: G06Q 30/0271 Personalized advertisement
• CPC Class Codes: G06Q 40/08 Insurance
• CPC Class Codes: G10L 15/00 Speech recognition G10L17/00 takes precedence
• CPC Class Codes: G16H 40/63 for local operation
• CPC Class Codes: G16H 40/67 for remote operation
• CPC Class Codes: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems
• CPC Class Codes: G16Z 99/00 Subject matter not provided for in other main groups of this subclass

Abstract

A method for using speech analysis to detect speech pathologies can begin with registration of a patient with a speech-based health monitor. A speech segment baseline representing an initial state of the patient'"'"'s speech system can be established for the patient. When prompted, the patient can submit a speech segment representing a current state of the patient'"'"'s speech system to the speech-based health monitor. The speech-based health monitor can analyze the submitted speech segment using the established speech segment baseline and/or a speech segment history that comprises speech segments previously submitted by the patient. Based upon said analysis, satisfaction of a health alert definition can be determined. A health alert definition can define an action performed by the speech-based health monitor when its associated triggering conditions are satisfied. The action associated with the at least one satisfied health alert definition can then be executed.

Claims

1. A method comprising:

indexing, by a computer comprising hardware and software executing on the hardware, a plurality of health conditions of a human against a plurality of different health-related speech characteristics;

identifying, by the computer, a speech segment of a speaker;

analyzing, by the computer, the speech segment to determine a presence or absence of any of the health-related speech characteristics;

upon determining a presence of one of the health-related speech characteristics, determining, by the computer, a corresponding one of the health conditions consistent with the indexing; and

generating, by the computer, an alert indicating that the speaker has a statistically significant likelihood of having the determined health condition based on results of the analyzing, wherein the speech segment is received from mobile computing device of the speaker, wherein the alert triggers an advertisement specific to the determined health condition to be sent over a network for presentation on the mobile communication device.

2. The method of claim 1, further comprising:

registering the speaker with a speech-based health monitor; and

establishing a speech segment baseline representing an initial state of the speaker, said state being a state where the speaker is in relatively good health, wherein the health-related speech characteristics represent speaker specific deviations from the speech segment baseline.

3. The method of claim 1, further comprising:

storing a plurality of audio segments for the speaker while the speaker is in an initial state of health, said audio segments being part of a speech history specific to the speaker, wherein the determining of the presence of the health-related speech characteristics is based on deviations between the health-related speech characteristics in the speech segment and the audio segments of the speech history.

4. The method of claim 1, wherein the alert is one of a set of different health alerts, said health alert being triggered responsive to conditions of a corresponding health alert definition being satisfied, said method further comprising:

executing at least one action associated with the at least one satisfied health alert definition.

5. The method of claim 1, wherein the analyzing compares respiratory specific characteristics of the speech segment against an excepted baseline for respiratory specific characteristics anticipated for a healthy respiratory state of the human, wherein the determined health condition is one that has altered a respiratory state of the human as determined by analyzing the speech segment to an unhealthy respiratory state as contrasted with the healthy respiratory state.

6. The method of claim 1, wherein the speaker is an individual insured by or seeking to be insured by a health insurance provider, wherein the generated alert triggers an event in a computing system of the health insurance provider that:

raises a cost of health insurance for the speaker in a manner consistent with the likelihood that the speaker has the determined health condition;

denies health insurance for the speaker based on the likelihood that the speaker has the determined health condition;

or

requires the speaker to undergo testing by a medical professional to assess whether or not the speaker has the determined health condition, wherein in absence of the determined health condition from the analyzing no testing by medical professionals would be required by the health insurance provider.

7. The method of claim 1, further comprising:

responsive to the alert, providing a computing device of the speaker with a request for additional data relating to the health of the speaker;

receiving the additional data; and

analyzing the additional data to make a more accurate assessment regarding whether the speaker has the determined health condition or not.

8. The method of claim 7, wherein the additional data comes from an image or video capture mechanism of a device used by the speaker, said device being a device from which audio for the speech segment was acquired.

9. The method of claim 1, further comprising:

processing the speech segment of the speaker to obscure spoken words contained in the speech segment while retaining content for detecting the health-related speech characteristics.

10. The method of claim 1, wherein the detected health-related speech characteristic indicates that the speaker is a smoker, is pregnant, or has a sinus blockage.

11. The method of claim 1, wherein the analyzing determines from the speech segment that there has been a weakening in the lavatory palatine of the speaker.

12. The method of claim 1, wherein the analyzing determines there is a problem relating to the primary muscle for moving a velum of the speaker.

13. The method of claim 1, wherein the health condition is one indicated by a respiratory or laryngeal problem ascertained by the analyzing.

14. The method of claim 1, wherein the health condition indicates a presence of mucus in the body of the speaker, or wherein the health condition is that the speaker has the statistically significant likelihood of having a tumor in their respiratory system.

19. The method of claim 1, wherein determining of the satisfaction of at least one health alert definition further comprises:

identifying the satisfaction of only a default health alert definition, wherein a situation is created where a health alert definition to handle said analysis is undefined;

indicating that the speech segment is unable to be handled by current health alert definitions, wherein said speech segment is marked within the speech segment history with an identifier that instructs the speech-based health monitor to ignore the speech segment for analysis purposes; and

requesting resolution of the marked speech segment from a designated entity; and

upon receipt of resolution for the marked speech segment, removing the identifier from the speech segment in the speech segment history.

15. A method for using speech analysis to detect speech pathologies comprising:

registering a patient with a speech-based health monitor;

establishing a speech segment baseline representing an initial state of the patient'"'"'s speech with the speech-based health monitor;

when prompted by the speech-based health monitor, submitting of a speech segment to the speech-based health monitor by the patient, wherein the submitted speech segment represents a current state of the patient'"'"'s speech system;

analyzing of the submitted speech segment by the speech-based health monitor with respect to at least one of the established speech segment baseline and a speech segment history associated with the patient, wherein the speech segment history comprises speech segments previously submitted by the patient;

based upon said analysis, determining of a satisfaction of at least one health alert definition by the speech-based health monitor, wherein a health alert definition defines at least one action performed by the speech-based health monitor when its associated triggering conditions are satisfied by the analysis; and

executing the at least one action associated with the at least one satisfied health alert definition.

16. The method of claim 15, wherein establishing of the speech segment baseline further comprises:

storing the speech segment baseline within a data store of the speech-based health monitor, wherein the stored speech segment baseline is associated with the patient.

17. The method of claim 15, wherein submitting of the speech segment to the speech-based health monitor further comprises:

adding the submitted speech segment to the speech segment history of the patient within a data store of the speech-based health monitor.

18. The method of claim 15, wherein analyzing the submitted speech segment further comprises:

comparing the submitted speech segment with the established speech segment baseline;

when results of said comparison indicate at least one difference between the submitted speech segment and the established speech segment baseline, determining whether each indicated difference is within a corresponding predefined speech segment tolerance, wherein said predefined speech segment tolerance expresses a maximum value that a speech characteristic of the submitted speech segment is allowed to deviate from the established speech segment baseline;

when each indicated difference of the speech segment is within its corresponding predefined speech segment tolerance, canceling the determination of the satisfaction of at least one health alert definition and the execution of the at least one action associated with the at least one satisfied health alert definition, wherein the speech-based health monitor awaits a new submission of the speech segment from the patient; and

when at least one indicated difference of the speech segment is outside its corresponding predefined speech segment tolerance, proceeding with the determination of the satisfaction of at least one health alert definition.

20. A computer program product comprising a non-transitory computer readable storage medium having computer usable program code embodied therewith and executed by one or more processors of the computer, the computer usable program code comprising:

computer usable program code configured to establish a speech segment baseline representing an initial state of a patient'"'"'s speech system;

computer usable program code configured to receive a speech segment from the patient, wherein the speech segment represents a current state of the patient'"'"'s speech system;

computer usable program code configured to analyze the received speech segment with respect to at least one of the established speech segment baseline and a speech segment history associated with the patient, wherein the speech segment history comprises speech segments previously submitted by the patient;

computer usable program code configured to, based upon said analysis, determine satisfaction of at least one health alert definition, wherein a health alert definition defines at least one action to be performed when its associated triggering conditions are satisfied by the analysis; and

computer usable program code configured to execute the at least one action associated with the at least one satisfied health alert definition.