Telephone pathology assessment

• US 20070005357A1
• Filed: 06/29/2005
• Published: 01/04/2007
• Est. Priority Date: 06/29/2005
• Status: Active Grant

• Current Assignee: University College Dublin
• Original Assignee: University College Dublin
• Inventors: Moran, Rosalyn, Lacy, Peter, Reilly, Richard, Chazal, Philip, O' Mullane, Brian
• Granted Patent: US 7,457,753 B2
• Time in Patent Office: Days
• US Class Current: 704/246
• CPC Class Codes: G10L 25/66 for extracting parameters related to health condition detecting or measuring for diagnostic purposes A61B5/00

Abstract

A system for remote assessment of a user is disclosed. The system comprises application software resident on a server and arranged to interact across a network with a user operating a client device to obtain one or more sample signals of the user'"'"'s speech. A datastore is arranged to store the user speech samples in association with details of the user. A feature extraction engine is arranged to extract one or more first features from respective speech samples. A comparator is arranged to compare the first features extracted from a speech sample with second features extracted from one or more reference samples and to provide a measure of any differences between the first and second features for assessment of the user.

Claims

1. A system for remote assessment of a user comprising:

application software resident on a server and arranged to interact across a network with a user operating a client device to obtain one or more sample signals of the user'"'"'s speech;

a datastore arranged to store said one or more user speech samples in association with details of the user;

a feature extraction engine arranged to extract one or more first features from respective speech samples; and

a comparator arranged to compare said first features extracted from a speech sample with second features extracted from one or more reference samples and to provide a measure of any differences between said first and second features for assessment of said user.

2. A system as claimed in claim 1 wherein said client is a cellular phone, said network includes the Global System for Mobile Communications (GSM) network, wherein said application software comprises an interactive voice recognition (IVR) application and wherein said server includes a voice mark-up language (VML) gateway.

3. A system as claimed in claim 1 wherein said client is a telephone handset, said network includes the public switched telephone network (PSTN), wherein said application software comprises an interactive voice recognition (IVR) application and wherein said server includes a voice mark-up language (VML) gateway.

4. A system as claimed in claim 1 wherein said client is a computing device, said network includes a packet switched network, wherein said application software comprises one or more web pages and wherein said server includes a web server.

5. A system as claimed in claim 1 wherein said reference samples comprise a database of speech samples, each sample having a pathology associated therewith, and wherein said feature extraction engine is arranged to said extract said second features from respective samples and to store said second features in association with respective pathologies.

6. A system as claimed in claim 5 wherein said pathologies comprise:

normal, neuromuscular pathologic, physical pathologic and mixed pathologic.

7. A system as claimed in claim 5 wherein said first and second features comprise one or more of pitch perturbation, amplitude perturbation and harmonic-to-noise ratio features.

8. A system as claimed in claim 7 wherein said pitch perturbation features include a mean frequency measure for a sample signal.

9. A system as claimed in claim 5 wherein said sample signals comprise a sustained phonation of the vowel sound /a/.

10. A system as claimed in claim 9 wherein said sample signals are between 2 and 5 seconds in length.

11. A system as claimed in claim 1 wherein said reference samples are limited in bandwidth to the bandwidth of said sampled signals.

12. A system as claimed in claim 1 wherein prior to operation of said feature extraction engine, said reference samples are distorted in a manner similar to any distortion involved in acquiring said sampled signals across said network.

13. A system as claimed in claim 5 wherein said comparator is arranged to aggregate second features for reference samples associated with like pathologies and to provide respective measures of the difference between said first features and respective aggregated second features for use in assessment of said user.

14. A system as claimed in claim 13 wherein said measures are stored in a datastore in association with user details and wherein said application software is arranged to interact with a clinician to provide respective measures for a speech sample in relation to any pathology having an associated reference sample.

15. A system as claimed in claim 1 wherein said one or more reference samples comprise a sample signal for said user, and wherein said sample signal is associated with a user state.

16. A system as claimed in claim 15 wherein said user state comprises one of:

forced respiration;

spontaneous respiration;

resting;

or after bronchodilator inhalation.

17. A system as claimed in claim 15 wherein said sample signals are between 20 and 120 seconds in duration.

18. A system as claimed in claim 15 wherein sample signals comprise at least one user respiratory cycle.

19. A system as claimed in claim 18 wherein said feature extraction engine is arranged to divide said sample signals into a sequence of one or more inspiration and expiration phases and wherein said first and second features comprise one of a mean or a peak valued frequency component of an expiration phase of a respiratory cycle.

20. A system as claimed in claim 19 wherein said frequency component is calculated based on a temporal sub-interval of said expiration phase.

21. A system as claimed in claim 15 wherein said sample signals and said reference samples are band limited between 100 and 2000 Hz.

22. A method operable in a server of remotely assessing a user comprising the steps of:

interacting with a user operating a client device connected to the server across a network to obtain one or more sample signals of the user'"'"'s speech;

storing said one or more user speech samples in association with details of the user;

extracting one or more first features from respective speech samples; and

comparing said first features extracted from a speech sample with second features extracted from one or more reference samples; and

providing a measure of any differences between said first and second features for assessment of said user.

23. A computer program product comprising a computer readable medium comprising computer code which when executed on a server device is arranged to perform the steps of claim 22.