Assessing joint condition using acoustic sensors

• US 10,448,919 B2
• Filed: 02/02/2018
• Issued: 10/22/2019
• Est. Priority Date: 12/08/2015
• Status: Active Grant

• Current Assignee: Kneevoice, Inc.
• Original Assignee: Kneevoice, Inc.
• Inventors: Leal, Carlos, Rigby, Felipe, De Greiff, Gustavo
• Primary Examiner(s): Fernandes, Patrick
• Application Number:
  US15/887,638
• Publication Number:
  US 20180153501A1
• Time in Patent Office: 627 Days
• CPC Class Codes: A61B 5/107 Measuring physical dimensions, e.g. size of the entire body or parts thereof
• CPC Class Codes: A61B 5/4514 Cartilage
• CPC Class Codes: A61B 5/4528 Joints A61B5/4533, A61B5/4538 take precedence
• CPC Class Codes: A61B 7/00 Instruments for auscultation
• CPC Class Codes: A61B 7/006 Detecting skeletal, cartilage or muscle noise A61B7/003 and A61B7/008 take precedence
• CPC Class Codes: A61B 8/0875 for diagnosis of bone A61B5/4504 takes precedence
• CPC Class Codes: G16H 20/30 relating to physical therapies or activities, e.g. physiotherapy, acupressure or exercising
• CPC Class Codes: G16H 50/30 for calculating health indices; for individual health risk assessment

Abstract

A new non-invasive tool for cartilage assessment, exercise and sports management, and prevention of osteoarthritis is provided. In various embodiments, cartilage condition is assessed using audible signals from joints. Assessment test results are used to provide feedback regarding joint stress and friction that is related to physiological or pathological loads. Data obtained from audible signals are processed to provide an index that can be interpreted by a user or third parties. The index is useful as a baseline for exercise practices, training routines, wellness programs, or rehabilitation protocols.

Claims

1. A system comprising:

a contact microphone;

a processor operatively connected to the contact microphone;

a computer readable storage medium having program instructions embodied therewith, the program instructions configured to cause the processor to perform a method comprising;

receiving from the contact microphone audio comprising sounds emanating from a human patellofemoral joint of a subject during knee flexion;

extracting a plurality of features from the audio by the processor;

providing the plurality of features to a trained classifier;

obtaining from the trained classifier a first numeric score indicative of patellofemoral joint health;

obtaining a pain value reported by the subject or a characteristic of the subject;

computing a second score indicative of joint health from the first score, wherein computing the second score comprises weighting the first score according to the pain value reported by the subject or according to the characteristic of the subject, wherein said weighting comprises applying a mathematical function to the first score; and

displaying the second score with an indication to perform follow-up when the second score is below a predetermined threshold.

2. The system of claim 1, the method further comprising:

capturing the audio via the contact microphone in contact with the exterior of a human knee.

3. The system of claim 2, wherein the contact microphone is in contact with an anterior patellar surface.

4. The system of claim 1, the method further comprising:

canceling noise from the audio prior to extracting the plurality of features.

5. The system of claim 4, wherein canceling noise comprises:

capturing an ambient audio signal by an outward facing microphone;

removing the ambient audio signal from the audio.

6. The system of claim 4, wherein canceling the noise comprises:

applying a bandpass filter to the audio.

7. The system of claim 1, wherein the plurality of features comprises zero-crossing rate, entropy of energy, spectral centroid, spectral spread, mel-frequency cepstral coefficients, or chroma vector.

8. The system of claim 1, wherein the trained classifier comprises a random decision forest.

9. The system of claim 1, wherein the trained classifier comprises a neural network.

10. The system of claim 1, wherein the trained classifier comprises a support vector machine.

11. The system of claim 1, wherein the characteristic comprises body mass index, age, gender, existing medical condition, or frequency of physical activity.

12. The system of claim 1, the method further comprising outputting the first score or the second score to a user.

13. A method comprising:

receiving from a contact microphone audio comprising sounds emanating from a human patellofemoral joint of a subject during knee flexion;

extracting a plurality of features from the audio by a processor;

providing the plurality of features to a trained classifier;

obtaining from the trained classifier a first numeric score indicative of patellofemoral joint health;

obtaining a pain value reported by the subject or a characteristic of the subject;

computing a second score indicative of joint health from the first score, wherein computing the second score comprises weighting the first score according to the pain value reported by the subject or according to the characteristic of the subject, wherein said weighting comprises applying a mathematical function to the first score; and

displaying the second score with an indication to perform follow-up when the second score is below a predetermined threshold.

14. The method of claim 13, further comprising:

capturing the audio via a contact microphone in contact with the exterior of a human knee.

15. The method of claim 14, wherein the contact microphone is in contact with an anterior patellar surface.

16. The method of claim 13, further comprising:

canceling noise from the audio prior to extracting the plurality of features.

17. The method of claim 15, wherein canceling noise comprises:

capturing an ambient audio signal by an outward facing microphone;

removing the ambient audio signal from the audio.

18. The method of claim 15, wherein canceling the noise comprises:

applying a bandpass filter to the audio.

19. The method of claim 13, wherein the plurality of features comprises zero-crossing rate, entropy of energy, spectral centroid, spectral spread, mel-frequency cepstral coefficients, or chroma vector.

20. The method of claim 13, wherein the trained classifier comprises a random decision forest.

21. The method of claim 13, wherein the trained classifier comprises a neural network.

22. The method of claim 13, wherein the trained classifier comprises a support vector machine.

23. The method of claim 13, wherein the characteristic comprises body mass index, age, gender, existing medical condition, or frequency of physical activity.

24. The method of claim 13, further comprising outputting the first score or the second score to a user.