Hearing Aid Shell Fit via Machine Learning Prototype Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current procedures for fitting hearing assistance devices are lengthy and complex, requiring multiple steps and taking an average of three weeks, with limited variability in device design that prevents a snug fit deep in the ear canal.
Innovation Solution
The system generates semi-customized 3D shell models using machine learning techniques, allowing for a set of prototype shells to be determined from a database of custom-made shells, enabling users to select a best fit without professional help, and mass-producing optimized designs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If custom-made shells are produced for each patient using traditional procedures, then the device can be fitted to the patient's anatomy, but the process takes an average of three weeks and is lengthy and complex
Solution Approach 1:
The system pre-generates a set of prototype shell models using machine learning techniques before patient fitting. These prototypes are created in advance from a database of custom-made shells, allowing patients to select from pre-fabricated options rather than waiting for custom manufacturing. This preliminary preparation reduces the fitting time from three weeks to one hour or less.
Solution Approach 2:
The system creates simplified 3D copies of custom-made shells as prototype models that can be digitally manipulated and selected. These copied models retain the essential fit characteristics of custom shells but can be rapidly generated and modified, enabling fast patient selection without full custom manufacturing cycles.
2Adaptability or versatility
If traditional custom-making procedures are used, then device design variability is maintained, but the process complexity increases and requires professional assistance
Solution Approach 1:
The system creates a universal set of prototype shell models that can serve multiple patients with similar anatomical characteristics. These prototypes are generated using machine learning to capture the essential variations needed for different patient types, providing adaptability without requiring complex custom manufacturing for each individual case.
Solution Approach 2:
The system enables patients to independently select their best-fit shell from the generated prototypes using a simple interface, eliminating the need for complex professional fitting procedures. Patients can view, compare, and select from the prototype options without requiring expert intervention in the selection process.
3Ease of manufacture
If prototype shells are mass-produced, then production costs are lowered, but the fit may be less precise compared to custom-made shells
Solution Approach 1:
The system uses machine learning to identify and adjust key geometric parameters of shell models when generating prototypes from the database. By optimizing these parameters based on aggregated data from multiple custom shells, the system maintains fit precision while enabling mass production of standardized prototype models.
Data Source
AI summary
Systems and methods may be used to determine a fit for a hearing assistance device shell model. For example, a method may include receiving an image of anatomy of a patient including at least a portion of a canal aperture of an ear of the patient, generating a patient model of a portion of the anatomy of the patient, the patient model indicating at least one of a height or width of the canal aperture, and determining, using the patient model, a best fit model from a set of hearing assistance device shell models generated using a machine learning technique. The method may include outputting an identification of the best fit model.