Hierarchical Pattern Recognition for Simultaneous Prosthetic Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current myoelectric prostheses are limited to sequential control of movements, restricting the ability to perform simultaneous multi-degree of freedom movements, which are essential for everyday activities, and existing pattern recognition systems are not effective in enabling fluid, life-like movements with prosthetics.
Innovation Solution
A hierarchical pattern recognition control scheme using a hierarchy of classifiers, including linear discriminant analysis, to classify and control simultaneous movements by interpreting electromyography signals, allowing for independent control of movement speed and direction across multiple degrees of freedom.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If sequential control methods are used, then device complexity is reduced, but productivity is limited due to inability to perform simultaneous movements
Solution Approach 1:
The control system is segmented into multiple independent classifiers, each responsible for a specific degree of freedom. This allows simultaneous processing of multiple movements while maintaining manageable complexity through modular architecture.
Solution Approach 2:
The system transitions from sequential one-dimensional control to simultaneous multi-dimensional control by processing multiple degrees of freedom in parallel, enabling complex coordinated movements that were previously impossible.
2Measurement precision
If pattern recognition algorithms are used, then measurement precision of movement intent is improved, but device complexity increases due to sophisticated processing requirements
Solution Approach 1:
The pattern recognition system is divided into separate classifiers for each degree of freedom, each processing specific movement patterns independently. This segmentation maintains high measurement precision while reducing overall system complexity through distributed processing.
Solution Approach 2:
Feature extraction acts as an intermediary layer that transforms raw EMG signals into meaningful patterns before classification. This intermediary processing improves measurement precision by focusing computational resources on discriminative features rather than raw signals.
3Adaptability or versatility
If independent control of each degree of freedom is implemented, then adaptability is improved, but device complexity increases due to multiple classifiers required
Solution Approach 1:
The control system is segmented into independent classifiers for each degree of freedom, allowing flexible and adaptive control of individual movements while maintaining overall system manageability through modular architecture.
Solution Approach 2:
The classifier hierarchy dynamically adapts to different movement combinations, enabling versatile control of simultaneous movements. Each classifier can be independently activated based on the specific movement intent, providing dynamic adaptability without requiring a completely reconfigured system.
Data Source
AI summary
This disclosure relates to a hierarchy of classifiers, including linear discriminant analysis (LDA) classifiers, arranged to provide simultaneous control of multiple degrees of freedom of a prosthetic. The high accuracy of the hierarchical approach allows pattern recognition techniques to be extended to permit simultaneous control, potentially allowing amputees to produce more fluid, life-like movements, ultimately increasing their quality of life. The hierarchy may also be used to control a biological interface that allows input to computers for persons with disabilities, used as a potential video-game controller, and used as an input interface for tablets, and phones.