Wrist-Worn Sensor Fusion for Real-Time Cognitive Load Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for assessing cognitive load are subjective, costly, or not suitable for real-time, continuous monitoring, and low-cost wearable sensors provide lower quality data, making accurate classification challenging.
Innovation Solution
A method and system using multi-modal wearable sensors to collect physiological signals, apply a multi-level feature extraction, select optimal features, augment training data, and use synthetic minority over-sampling to train a classification model for real-time cognitive load classification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If expensive physiological sensors are used for cognitive load assessment, then measurement precision is improved, but device cost increases
Solution Approach 1:
The patent employs low-cost wearable sensors (such as wrist-worn devices with skin conductance, heart rate, and temperature sensors) instead of expensive physiological sensors. These affordable sensors are sufficient when combined with multi-level feature extraction and machine learning models, achieving comparable or superior classification accuracy while dramatically reducing system cost and complexity.
Solution Approach 2:
The patent combines multiple low-cost wearable sensors (skin conductance, heart rate, respiration rate, skin temperature) to create a multi-modal sensing system. By merging data from these inexpensive sensors and applying multi-level feature extraction, the system achieves comprehensive cognitive load assessment without requiring any single expensive sensor.
2Device complexity
If low cost wearable sensors are used for cognitive load assessment, then device cost is reduced, but measurement precision deteriorates
Solution Approach 1:
The patent applies multi-level feature extraction that transforms raw sensor signals into multiple dimensions of information. Level 1 extracts basic signal features, Level 2 extracts temporal and spectral features, and Level 3 extracts non-linear dynamics features. This dimensional transformation creates rich feature representations from low-cost sensor data, enabling accurate cognitive load classification.
Solution Approach 2:
The patent employs multiple feature extraction methods (time-domain, frequency-domain, wavelet transform) that transform sensor parameters into different representations. By changing the parameter space and extracting features at multiple levels, the system maximizes the information content from low-cost sensor signals, improving measurement precision without increasing hardware cost.
3Device complexity
If single physiological signal is used for cognitive load assessment, then device complexity is reduced, but measurement precision deteriorates
Solution Approach 1:
The patent combines four different physiological signals (skin conductance, heart rate, respiration rate, skin temperature) measured by multiple wearable sensors. This multi-modal approach captures complementary information about cognitive load, and the combined features achieve superior classification accuracy compared to any single signal alone.
Solution Approach 2:
The patent uses a unified machine learning classification model that processes features from multiple physiological domains (electrical, temporal, spectral, non-linear). This universal model handles diverse sensor types and feature representations, extracting meaningful patterns that indicate cognitive load states across all measured parameters.
4Ease of operation
If traditional self-report questionnaires are used for cognitive load assessment, then ease of operation is improved, but measurement precision deteriorates and real-time capability is lost
Solution Approach 1:
The patent replaces manual self-report questionnaires with an automated physiological monitoring system. Instead of requiring participants to verbally report their cognitive load, the system automatically measures physiological signals through wearable sensors and classifies cognitive load states using machine learning models, providing continuous objective monitoring.
Solution Approach 2:
The patent enables continuous real-time cognitive load assessment through wearable sensors that continuously monitor physiological signals. Unlike questionnaire-based methods that provide only snapshots at specific moments, the system continuously tracks cognitive load changes throughout the entire task duration, providing ongoing feedback.
Data Source
Figure 1
Figure 2
Figure 3A
AI summary
This disclosure relates generally to a method and system for classification of cognitive load (CL) using data obtained from wearable sensors. The disclosed method uses a multi-modal based approach using wrist-worn sensors for real time monitoring of CL in real world scenarios and improves the accuracy of detection of CL. A set of distinguishing features are selected from physiological signals received from the wrist-worn sensors. These features are used for training a classification model for classifying the CL of a patient into a no load or a high load. The set of distinguishing features are selected from domain specific features and signal property based generic features of the physiological signals. The disclosed method is used for classification of CL in scenarios such as to check how the cognitive load of a candidate varies during interviews, to assess the participants workload during online meetings and so on.