ML Prosthesis Shape Generation for Faster Socket Fit
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional prosthesis manufacturing is time-consuming and economically burdensome, requiring multiple visits and expert adjustments, and existing systems fail to provide a method for quickly and economically producing a properly fitting prosthesis.
Innovation Solution
A prosthesis shape data generation system utilizing machine learning to estimate and refine prosthesis shape data based on stump shape data, incorporating three-dimensional scanning and printing, and allowing for expert-driven modifications to improve fit and reduce expertise requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If conventional prosthesis manufacturing methods are used with expert customization, then the prosthesis fit quality is improved, but the production time and cost increase significantly
Solution Approach 1:
The system performs preliminary actions by acquiring stump shape data through 3D scanning and using machine learning models to pre-generate prosthesis shape data before actual manufacturing. This preliminary automated generation eliminates the need for multiple expert adjustment visits, resolving the contradiction by achieving good fit quality upfront without time-consuming manual iterations
Solution Approach 2:
The system creates a digital copy of the stump shape through 3D scanning and uses this copy to generate the prosthesis design. This digital copying and simulation allows the machine learning model to predict the optimal prosthesis shape without requiring physical trial fittings, thereby reducing production time while maintaining fit quality
2Manufacturing precision
If conventional prothesis manufacturing with multiple expert visits is used, then the prosthesis fit quality is improved, but the economic burden increases
Solution Approach 1:
The system enables self-service by allowing automated generation of prosthesis shape data using machine learning models trained on existing expert knowledge. The system serves itself by automatically processing stump shape data and generating optimized prosthesis designs without requiring continuous expert intervention, thereby reducing manufacturing costs while maintaining fit quality
Solution Approach 2:
The system changes parameters by transforming physical stump measurements into digital 3D shape data, then using machine learning to predict optimal prosthesis shape parameters. This parameter transformation from physical to digital domain enables automated processing and reduces the need for expensive manual expert adjustments
3Productivity
If automated shape generation without expert input is used, then the production time and cost are reduced, but the prosthesis fit quality may deteriorate
Solution Approach 1:
The system performs preliminary action by pre-training machine learning models using extensive expert-generated prosthesis data before deployment. This preliminary training embeds expert knowledge into the automated system, ensuring that when the system generates prosthesis shapes automatically, it produces results with quality comparable to expert-made prostheses while maintaining high productivity
Solution Approach 2:
The system implements feedback by using the generated prosthesis shape data to create sockets and potentially iteratively refining the design based on fit outcomes. This feedback mechanism ensures that automated generation maintains high fit quality by continuously improving based on actual performance data
Data Source
AI summary
To quickly and economically manufacture a prosthesis that fits the shape of a stump of a prosthesis user. A prosthesis shape data generation system provided includes: a stump shape data acquisition unit that acquires stump shape data, which is shape data on a stump of a living body; and an estimated prosthesis shape data generation unit that performs an estimation processing by inputting the stump shape data to a machine learning model that has previously learnt a correspondence between a stump shape and a shape of a part or a whole of a prosthesis that fits the stump shape, thereby generating estimated prosthesis shape data that is shape data on a part or a whole of a prosthesis that fits the stump of the living body.