MUSCULOSKELETAL ACTIVITY RECOGNITION SYSTEM AND METHOD

• US 20150045699A1
• Filed: 03/15/2013
• Published: 02/12/2015
• Est. Priority Date: 04/03/2012
• Status: Active Grant

• Current Assignee: Carnegie Mellon University
• Original Assignee: Carnegie Mellon University
• Inventors: Mokaya, Frank, Zhang, Pei
• Granted Patent: US 10,130,298 B2
• Time in Patent Office: Days
• US Class Current: 600/595
• CPC Class Codes: A61B 2562/0219 Inertial sensors, e.g. accelerometers, gyroscopes, tilt switches
• CPC Class Codes: A61B 5/0004 characterised by the type of physiological signal transmitted
• CPC Class Codes: A61B 5/0024 for multiple sensor units attached to the patient, e.g. using a body or personal area network
• CPC Class Codes: A61B 5/1101 Detecting tremor
• CPC Class Codes: A61B 5/1107 Measuring contraction of parts of the body, e.g. organ, muscle apparatus for measuring work or force in general G01L
• CPC Class Codes: A61B 5/1126 using a particular sensing technique
• CPC Class Codes: A61B 5/224 Measuring muscular strength
• CPC Class Codes: A61B 5/4519 Muscles A61B5/389, A61B5/224 take precedence
• CPC Class Codes: A61B 5/7203 for noise prevention, reduction or removal
• CPC Class Codes: A61B 5/725 using specific filters therefor, e.g. Kalman or adaptive filters A61B5/316 take precedence
• CPC Class Codes: A61B 5/7257 using Fourier transforms
• CPC Class Codes: A61B 5/7267 involving training the classification device
• CPC Class Codes: A61B 5/7278 Artificial waveform generation or derivation, e.g. synthesising signals from measured signals A61B5/327 takes precedence
• CPC Class Codes: A61B 5/7425 Displaying combinations of multiple images regardless of image source, e.g. displaying a reference anatomical image with a live image geometric transforms for image registration G06T3/14; enhancement using image fusion G06T5/50; algorithms for registration of images G06T7/30

Abstract

A muscle activity and skeletal monitoring system and method are disclosed. A network of at least two distributed inertial sensor nodes are configured to communicate with each other over a first interface. Each of the inertial sensor nodes comprises at least one sensor configured to sense muscle vibrations and monitor body motion. A muscle activity recognition and motion tracking and visualization methods also are disclosed.

Claims

1. A muscle activity and skeletal monitoring apparatus comprising:

a network of at least two distributed inertial sensor nodes configured to communicate with each other over a first interface, wherein each of the inertial sensor nodes comprises at least one sensor configured to sense muscle vibrations and monitor body motion

2. The apparatus of claim 1, wherein the least one sensor comprises at least one of accelerometer, gyroscope, or magnetometer, and any combination thereof.

3. The apparatus of claim 1, wherein at least one inertial sensor node is configured to communicate over a wired channel or a wireless channel, and any combination thereof.

4. The apparatus of claim 3, wherein at least one inertial sensor node is configured to communicate to a wireless data aggregator.

5. The apparatus of claim 3, wherein at least one inertial sensor node is configured to communicate to a server.

6. The apparatus of claim 1, wherein the at least one sensor is configured to communicate over a wired channel or a wireless channel, and any combination thereof.

7. The apparatus of claim 6, wherein the at least one sensor is configured to communicate to a wireless data aggregator.

8. The apparatus of claim 6, wherein the at least one sensor is configured to communicate to a server.

9. The apparatus of claim 1, wherein the at least one sensor comprises a microelectromechanical systems (MEMS) based device.

10. The apparatus of claim 1, wherein the at least one sensor is a triple axis device.

11. A method comprising monitoring muscle activity through mechanomyography (MMG) and tracking skeletal motion using a common sensor, sensor node or any combination thereof.

12. A method comprising monitoring muscle activity through mechanomyography (MMG) and tracking skeletal motion using a common sensor.

13. The method of claim 12, wherein the common sensor is an accelerometer.

14. The method of claim 13, wherein the accelerometer comprises a microelectromechanical systems (MEMS) based device.

15. The method of claim 13, wherein monitoring muscle activity comprises incorporating data from a gyroscope or a magnetometer, and any combination thereof, with data from the accelerometer.

16. A muscle activity recognition method, comprising:

determining vibration signature data from data received from an accelerometer, wherein the vibration signature data is associated with a muscle in proximity of the accelerometer; and

extracting features from the vibration signature data.

17. The muscle activity recognition method of claim 16, further comprising training a classifier on the extracted features.

18. The muscle activity recognition method of claim 16, comprising filtering the accelerometer data.

19. The muscle activity recognition method of claim 16, comprising determining a time domain feature vector using a sliding window technique with either fixed or variable overlap on the accelerometer data.

20. The muscle activity recognition method of claim 16 comprising determining a selected set of feature vectors.

21. The muscle activity recognition method of claim 20, comprising determining a frequency domain feature vector calculating a discrete Fourier transform on the filtered accelerometer data contained in respective fixed sliding window.

22. The muscle activity recognition method of claim 21, comprising combining the time domain feature vector and the frequency domain features vector into a single matrix, wherein the matrix constitutes the unique vibration signature of a predetermined muscle group.

23. The muscle activity recognition method of claim 17, wherein training the classifier comprises building a machine learning technique to classify and distinguish different instrumented muscle groups from each other.

24. The muscle activity recognition method of claim 22, wherein the machine learning technique is an implementation of a J48 decision tree.

25. The muscle activity recognition method of claim 16, comprising building another machine learning technique to infer a level of exertion or fatigue in a given instrumented muscle group.

26. The muscle activity recognition method of claim 16, comprising filtering the sensor orientation data through a range of motion limiter.

27. The muscle activity recognition method of claim 2526, comprising:

limiting offset errors in the sensor orientation data; and

restricting the sensor orientation data obtained from the fusion of sensor data.