Instrumented Mouthguard Impact Classification Without Sensor Calibration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing monitoring systems for head impacts in sports, particularly contact sports, lack accuracy and consistency due to variations in sensor placement and orientation in instrumented mouthguards, leading to subjective assessments and high costs for calibration.
Innovation Solution
A method for generating a feature array and training a classifier using signal data from instrumented mouthguards, involving data rotation and alignment, noise filtering, and feature extraction to create a customizable training dataset, independent of sensor alignment, followed by machine learning model training for accurate impact classification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sensor calibration and alignment procedures are implemented to improve measurement precision, then measurement precision improves, but device complexity and manufacturing cost increase
Solution Approach 1:
The patent transforms the coordinate system and rotates the gyroscope data to align with the accelerometer axes, changing the parameter representation rather than physically realigning sensors. This mathematical transformation achieves alignment without adding physical calibration mechanisms.
Solution Approach 2:
The patent replaces mechanical alignment procedures with computational methods. Instead of physically calibrating and aligning sensors during manufacturing, the system uses software-based coordinate transformation and data rotation to achieve the same effect, eliminating complex mechanical calibration steps.
2Reliability
If spotter monitoring is used to assess head impact events, then reliability of impact detection improves, but loss of time and productivity decrease due to manual assessment
Solution Approach 1:
The system enables self-service by allowing the mouthguard device to automatically classify and assess impact events using embedded sensors and machine learning algorithms, eliminating the need for external spotter intervention for initial assessment and classification.
Solution Approach 2:
The patent replaces manual spotter assessment with automated machine learning classification. The system uses trained classifiers to automatically evaluate impact severity and type, substituting human judgment with computational analysis that is both faster and more consistent.
3Measurement precision
If multiple sensors are placed in the mouthguard to improve measurement accuracy, then measurement precision improves, but device complexity and manufacturing cost increase
Solution Approach 1:
The patent makes the classification system universal by training machine learning models that can accurately classify impacts regardless of specific sensor placement variations. The system is designed to work with standard off-the-shelf sensors in various configurations, making the manufacturing process simpler while maintaining measurement accuracy through robust algorithmic handling of sensor data.
Data Source
AI summary
A computer implemented method of generating a feature array configured for training a classifier of impact data measured by a mouthguard includes receiving signal data, receiving impact classification data representative of feature information that specifies the class of impact of each received signal data and storing the impact classification in a response array, extracting one or more features from the signal data and storing in an array, comparing each extracted feature to the corresponding response array element to select the set of features that respectively satisfy a classification relevance threshold indicative of a classification relevance, and, responsive to the number of selected features being less than a feature threshold, iteratively rotating the direction of respective x, y, z components of the rotational velocity time series data relative to corresponding x, y, z components of the linear acceleration time series data.