Systems and methods for approximating musculoskeletal dynamics
Inventors
Sobinov, Anton • Yakovenko, Sergiy • Gritsenko, Valeriya • Boots, Matthew
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Assignees
Member
West Virginia UniversityWest Virginia University is a public R1 research institution offering diverse undergraduate and graduate programs across science, engineering, business, creative arts, and media. The university emphasizes experiential learning, research, and innovation, with notable strengths in academic program development, research excellence, community engagement, and a commitment to affordability, career readiness, and student success. WVU supports a vibrant campus environment, industry partnerships, and impactful scholarship, preparing students for careers through hands-on education, research, and applied learning.
West Virginia University is a public R1 research institution offering diverse undergraduate and graduate programs across science, engineering, business, creative arts, and media. The university emphasizes experiential learning, research, and innovation, with notable strengths in academic program development, research excellence, community engagement, and a commitment to affordability, career readiness, and student success. WVU supports a vibrant campus environment, industry partnerships, and impactful scholarship, preparing students for careers through hands-on education, research, and applied learning.
Publication Number
US-12508138-B2
Publication Date
2025-12-30
Expiration Date
Abstract
A system and method for controlling a device, such as a prosthetic limb are provided. A biomimetic controller of the system comprises a signal processor and a musculoskeletal model. The signal processor processes M biological signals received from a residual limb to transform the M biological signals into N activation signals, where M and N are integers and M is less than N. The musculoskeletal model transforms the N activation signals into intended motion signals. A prosthesis controller transforms the intended motion signals into three or more control signals that are outputted from an output port of the prosthesis controller. A controlled device receives the control signals and performs one or more tasks in accordance with the control signals.
Core Innovation
A controller receives at least one biological signal from a residual limb and expands the at least one biological signal into one or more activation signals based at least in part on a polynomial associated with musculoskeletal dynamics of the residual limb. The controller transforms the one or more activation signals into one or more intended motion signals based at least in part on a plurality of approximated muscle dynamics, and further transforms the one or more intended motion signals into one or more control signals.
The polynomial associated with musculoskeletal dynamics is based on approximated muscle dynamics, including approximations of muscle-length and moment-arm polynomials fitted from datasets. Polynomial structures are incrementally expanded using candidate-term selection and a corrected Akaike Information Criterion (AICc), and the polynomials can be constrained via a differential relationship between muscle length and moment arms.
Similarity and structure analyses describe non-random organization of the polynomial vectors, including DOF-independent muscle polynomial vectors and organization analyses using PCA/clustering and statistical tests. The approximation is integrated into real-time control loops for myoelectric prostheses and virtual reality training, including feedback and switching between virtual reality and physical prosthesis control modes.
Claims Coverage
The partial content provides three independent claims: a system claim (clm-00001), a method claim (clm-00007), and an additional system claim (clm-00013). Across these independent claims, the main inventive features are the same pipeline: biological-signal reception, polynomial-based expansion to activation signals, transformation to intended motion signals using approximated muscle dynamics, and transformation to control signals.
Polynomial-based activation-signal expansion from residual-limb biological signals
Expand at least one biological signal received from a residual limb into one or more activation signals based at least in part on a polynomial associated with musculoskeletal dynamics of the residual limb.
Approximated muscle-dynamics transformation to intended motion signals
Transform the one or more activation signals into one or more intended motion signals based at least in part on a plurality of approximated muscle dynamics.
Intended-motion to control-signal transformation
Transform the one or more intended motion signals into one or more control signals.
Controller-operated system architecture for biological-to-control pipeline
A controller comprising a processor and a memory executes machine-readable instructions that cause the controller to receive biological signals from a residual limb, expand the biological signals into activation signals based at least in part on a polynomial associated with musculoskeletal dynamics, transform activation signals into intended motion signals based at least in part on approximated muscle dynamics, and transform intended motion signals into control signals.
Method for controlling a device using the biological-to-control pipeline
Receive at least one biological signal from a residual limb, expand the biological signal into one or more activation signals based at least in part on a polynomial associated with musculoskeletal dynamics of the residual limb, transform activation signals into one or more intended motion signals based at least in part on a plurality of approximated muscle dynamics, and transform the intended motion signals into one or more control signals.
Signal processing, modeling, and control algorithms
Execute a signal processing algorithm to expand at least one biological signal into one or more activation signals based at least in part on a polynomial associated with musculoskeletal dynamics of the residual limb; execute a modeling algorithm to transform the one or more activation signals into one or more intended motion signals based at least in part on a plurality of approximated muscle dynamics; and execute a control algorithm to transform the one or more intended motion signals into one or more control signals.
Across the independent claims, the core claim coverage centers on producing one or more control signals by using a polynomial associated with musculoskeletal dynamics to expand residual-limb biological signals into activation signals, using approximated muscle dynamics to produce intended motion signals, and using a controller or algorithms to map intended motion signals to control signals.
Stated Advantages
Achieves low approximation error compared to spline approximations, while being far faster and lower-memory than spline approximations.
Documented Applications
Real-time control loops for myoelectric prostheses.
Virtual reality training with switching between virtual reality and physical prosthesis control modes.
Interested in licensing this patent?