Prosthesis (i.e. – artificial body members) – parts thereof – or ai – Having electrical actuator
Reexamination Certificate
2001-02-02
2003-12-09
Snow, Bruce (Department: 3738)
Prosthesis (i.e., artificial body members), parts thereof, or ai
Having electrical actuator
C623S025000, C623S064000
Reexamination Certificate
active
06660042
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention generally relates to methods for moving individual fingers on a prosthetic hand based on volitional pressures within the residual forelimb, and more particularly to a method of extracting individual finger commands from volitional forelimb pressures utilizing training and extraction algorithms based on signal amplitudes.
2. Related Art
Many upper limb amputees can readily learn to control a prosthetic hand by flexing and extending their residual forearm muscles. Their dexterity, however, is limited by current myoelectric technology to simple grasping, with one degree of freedom (DOF). Advanced electromyographic (EMG) processing methods may allow enhanced grasping control Okuno, et al., “Development of Biomimetic Prosthetic Hand Controlled by Electromyogram,” Int. Workshop on Advanced Motion Control, Vol. 1, pp. 103-108, 1996; Vuskovic, et al., “Blind Separation of Surface EMG Signals,” 18
th
Ann. Intl. Conf. of the IEEE Med. Biol. Soc., pp. 1478-1483, 1996, but remain limited for more dexterous motions due to the inherent noisiness and complexity of EMG signals. Alternatives to EMG-based control use muscle and tendon forces generated within the prosthetic socket. An early demonstration of this method was the French Electric Hand, which used pneumatic pressure generated voluntarily within the socket to actuate a 1-DOF hand. Lucaccini, et al., “The French Electric Hand: Some Observations and Conclusions,”
Bulletin of Prosthetics Research
, vol. 10, no. 6, pp. 31-51, 1966.
Recently, Abboudi, et al., “A Biomimetic Controller for Multifinger Prosthesis,” IEEE Trans. Rehab. Eng., Vol. 7, No. 2, pp. 121-129, 1999, showed that many amputees could express control over their extrinsic finger muscles through dynamic pressure and shape changes in their residua. These three-dimensional mechanical dynamics can be measured with myo-pneumatic (M-P) sensors within the socket. Phillips, et al., “A Smart Interface for a Biomimetic Upper Limb Prosthesis,” Proc. 6
th
World Biomaterials Congr., May, 2000. A fine degree of residual limb muscle control was also demonstrated by amputees in a study using Hall effect magnetic movement sensors. Kenney, et al., “Dimensional Change In Muscle As a Control Signal for Powered Upper Limb Prostheses: A Pilot Study,” Med. Eng. & Phys., vol. 21, pp. 589-597, 1999.
Previous studies have proposed methods to control prosthetic upper limbs based on EMG signals from one or more sites on the upper body. Kang, et al., “The Application of Cepstral Coefficients and Maximum Likelihood Method in EMG Pattern Recognition,” IEEE Trans. Biomed. Eng., Vol. 42, No. 8, pp. 777-784, 1995; Park, et al., “EMG Pattern Recognition Based on Artificial Intelligence Techniques,” IEEE Trans. Rehab. Eng., Vol. 6, No. 4, pp. 400-405, 1998; Chang, et al., “Real-Time Implementation of Electromyogram Pattern Recognition as a Control Command of Man-Machine Interface,” Med. Eng. Phys., Vol. 18, No. 7, pp. 529-537, 1996; and Gallant, et al., “Feature-Based Classification of Myoelectric Signals Using Artificial Neural Networks,” Med. Biol.
Eng. Comput., Vol. 36, pp. 485-489, 1998. These methods can give the user control over basic limb movements such as elbow flexion, and wrist pronation. Decoding spatially distributed EMG signals is relatively difficult, however, requiring an initial pattern extraction from each EMG site, followed by a second extraction of the coupled patterns. Decoding of individual finger extensor EMG signals has been attempted, and may provide enhanced grasping control Vuskovic, et al. EMG signals, however, are inherently more complex than mechanical signals, and remain subject to problems such as relative motion between the skin and electrode, skin moisture, electrode corrosion, and noise.
Accordingly, what is needed, but has not heretofore been provided, is a method for sensing distributed mechanical forces in a limb corresponding to digit movement, and discriminating to provide reliable digit control. Herein, a pressure vector decoder (PVD) is developed that extracts multiple DOF from distributed pressures in the residual forelimb. The extracted commands can control fingers in real time, thus restoring a degree of biomimetic dexterity to the user.
The present invention relates to a user-trained filter for decoding volitional commands acquired from pressure sensors on a residual limb. The method is biomimetic, since the user controls the movement of mechanical fingers by activating appropriate finger flexor muscles. Pseudoinverse filtering has not been previously applied to biomechanical signals, but has been used for image processing and audio source separation. Cepstral analysis has been successfully used for pattern recognition applications in speech, EMG signals, and heart rate variability, knee vibrations, and arterial pulse velocity measurements.
OBJECTS AND SUMMARY OF THE INVENTION
It is a primary object of the present invention to provide a filter for extracting individual finger commands from volitional pressures within the residual forelimb.
It is an additional object of the present invention to discriminate individual finger commands from distributed pressures in the residual forelimb for controlling digit movement on a prosthesis.
It is even an additional object of the present invention to provide a method which utilizes linear filter parameters using the pseudoinverse of a feature matrix obtained from the forelimb signals, for extracting finger commands from volitional pressures within the residual forelimb.
It is a further object of the present invention to provide a method for controlling digit movement for a prosthesis wherein the prosthesis is trained under user supervision.
It is an additional object of the present invention wherein the feature extraction algorithm used in the present invention is based on signal amplitudes, derived from mechanical activity within the subject's anatomy.
It is an additional object of the present invention wherein the signal validation algorithm used in the present invention is based on cepstral encoding.
The present invention relates to filters for extracting finger commands from volitional pressures within the residual forelimb. The filters discriminate from the distributed pressures to provide reliable digit control. Linear filter parameters are trained using the pseudoinverse of the feature matrix obtained from the forelimb signals. Feature extraction algorithms are based on signal amplitudes (peak amplitude or root mean square). Cepstral distance analysis can be performed as a quality validation step for the training data, but is not necessary for the pseudoinverse filter method to work. Results with an amputee showed that multiple finger commands could be decoded reliably, representing real-time individual control over three prosthetic fingers. The filter is a critical component for biomimetic control and enhanced dexterity.
REFERENCES:
patent: 3652980 (1972-03-01), Goupillaud
patent: 3820168 (1974-06-01), Horvath
patent: 4650492 (1987-03-01), Barkhordar et al.
patent: 4748672 (1988-05-01), Nevill, Jr. et al.
patent: 4770662 (1988-09-01), Giampapa
patent: 4858124 (1989-08-01), Lizzi et al.
patent: 5336269 (1994-08-01), Smits
patent: 5367475 (1994-11-01), White
patent: 5413611 (1995-05-01), Haslam, II et al.
patent: 5447403 (1995-09-01), Engler, Jr.
patent: 5480454 (1996-01-01), Bozeman, Jr.
patent: 5598505 (1997-01-01), Austin et al.
patent: 6321195 (2001-11-01), Lee et al.
patent: 6416480 (2002-07-01), Nenov
Abboudi, et al., “A Biomimetic Controller for a Multifinger Prosthesis,” IEEE Trans. Rehab. Eng., vol. 7, No. 2, pp. 121-128, 1999.
Atal, “Automatic Recognition of Speakers From Their Voices,” Proc. IEEE, vol. 64, No. 4, pp. 460-475, 1976.
Chang, et al., “Real-Time Implementation of Electromyogram Pattern Recognition as a Control Command of Man-Machine Interface,” Med. Eng. Phys., vol. 18, No. 7, pp. 529-537, 1996.
Curcie, et al., “Biomimetic Finger Control by Filtering of Distributed Forelimb Pressures,” IEEE Transactions on Neural Syst
Curcie David J.
Flint James A.
Miller Cheryl
Rutgers The State University of New Jersey
Snow Bruce
Wolff & Samson PC
LandOfFree
Methods of biomimetic finger control by filtering of... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Methods of biomimetic finger control by filtering of..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Methods of biomimetic finger control by filtering of... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3154000