Methods of biomimetic finger control by filtering of distributed forelimib pressures

• US 6,660,042 B1
• Filed: 02/02/2001
• Issued: 12/09/2003
• Est. Priority Date: 02/02/2001
• Status: Expired due to Fees

• Current Assignee: Rutgers University
• Original Assignee: Rutgers University
• Inventors: Flint, James A., Curcie, David J.
• Primary Examiner(s): Snow, Bruce
• Assistant Examiner(s): MILLER, CHERYL L
• Application Number:
  US09/776,100
• Time in Patent Office: 1,040 Days
• Field of Search: 623/24, 623/25, 623/64
• US Class Current: 623/24
• CPC Class Codes: A61F 2/586 Fingers
• CPC Class Codes: A61F 2/68 Operating or control means
• CPC Class Codes: A61F 2/72 Bioelectric control, e.g. myoelectric
• CPC Class Codes: A61F 2002/701 operated by electrically controlled means, e.g. solenoids or torque motors
• CPC Class Codes: A61F 2002/704 computer-controlled, e.g. robotic control
• CPC Class Codes: A61F 2002/705 Electromagnetic data transfer
• CPC Class Codes: A61F 2002/7635 for measuring force, pressure or mechanical tension

Abstract

A method for controlling digit movement of a prosthesis based on residual forelimb volitional pressures is provided. A filter is provided for discriminating from the distributed pressures to provide reliable digit control. The extracted commands can be used to control actuators in a hand prosthesis to restore biomimetic finger control to the user. Feature extraction algorithms can be based on signal amplitudes (peak amplitude or root mean square). Training signal validation can be based on cepstral encoding. The filter is trained under user supervision.

Claims

1. A method for controlling digit movement on a prosthesis comprising:

interconnecting a plurality of mechanical sensors at locations on a forelimb;

acquiring data from the sensors during a training phase by flexing phantom digits while pressure signals are digitally sampled and stored;

filtering the data;

processing the data; and

individually controlling digits of the prosthesis in accordance with the processed data.

2. The method of claim 1 wherein filtering is performed by an absolute peak method.

3. The method of claim 1 wherein filtering is performed by a root mean square method.

4. The method of claim 1 wherein training involves quantifying degradation of volitional commands through a limb sensor interface.

5. The method of claim 1 wherein training includes at least one full flexion of each digit over a short time period.

6. The method of claim 1 wherein the step of processing the data further comprises cepstral encoding the data.

7. The method of claim 1 wherein cepstral distance is used to validate training signal quality.

8. A method of training a filter for discriminating distributed pressures in a forelimb and controlling movements of digits on a prosthesis comprising:

positioning a plurality of sensors on a forelimb;

sampling pressure signals at the sensors on the forelimb corresponding to flexing of selected phantom digits; and

extracting an amplitude feature for each sensor resulting in a feature vector for each digit of the prosthesis.

9. The method of claim 8 wherein the step of extracting the amplitude feature comprises using a peak amplitude method to determine a largest signal deviation.

10. The method of claim 8 wherein the step of extracting the amplitude feature comprises using a root mean square method.

11. The method of claim 8 wherein a feature matrix is formed from feature vectors for each digit.

12. The method of claim 11 wherein a pseudoinverse of the feature matrix is calculated.