Wearable Gesture Sensors for Autonomous Meal Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current insulin management systems for Type 1 diabetes patients are inefficient due to delays in glucose readings and insulin diffusion, requiring manual input for meal announcements, which can lead to poor glycemic control and compliance issues, especially in younger children.
Innovation Solution
A processor-implemented method and system that uses gesture sensors to detect eating or drinking activities, generating information to adjust the performance mode of a device monitoring physiological characteristics, such as a wearable device interacting with a digital health app to provide autonomous insulin dosing and reminders.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual meal announcement is used, then system response accuracy is improved, but patient burden and compliance issues worsen
Solution Approach 1:
The system automatically detects eating events using sensor data from wearables (accelerometers, gyroscopes, heart rate monitors) without requiring manual patient input. The device autonomously monitors physiological parameters and identifies eating patterns, eliminating the need for patients to manually announce meals while maintaining accurate detection.
Solution Approach 2:
The patent replaces manual mechanical input (patient pressing buttons or verbally announcing meals) with automated electronic sensing systems. Sensors detect eating events through subtle physiological changes, substituting the mechanical interaction between patient and system with passive, automated detection that improves compliance.
2Measurement precision
If continuous glucose monitoring is used, then real-time insulin dosing is improved, but system response speed worsens due to delays in glucose readings and insulin diffusion
Solution Approach 1:
The system performs preliminary detection of eating events using wearable sensors before glucose levels actually change. By detecting the act of eating itself (through accelerometer data, heart rate changes, or other physiological markers) and proactively triggering insulin delivery in anticipation of glucose absorption, the system overcomes the inherent delays in glucose monitoring and insulin diffusion.
Solution Approach 2:
The patent introduces an intermediary detection layer using wearable sensors that monitor eating behavior as a proxy for upcoming glucose changes. Instead of waiting for glucose readings to trigger insulin (which is too slow), the system uses eating detection as an intermediate signal to time insulin delivery optimally, accounting for the delays in glucose absorption and insulin diffusion.
3Extent of automation
If gesture sensors are used for activity detection, then autonomous meal detection is improved, but device complexity increases
Solution Approach 1:
The patent leverages existing multi-functional wearable devices (fitness trackers, smartwatches, medical monitors) that already contain accelerometers, gyroscopes, and heart rate sensors for general health monitoring. By making these devices also serve as eating detection tools, the system achieves autonomous meal detection without adding specialized hardware, thus avoiding increased device complexity.
Solution Approach 2:
The system achieves enhanced automation by changing the parameters and processing methods of existing sensors rather than adding new hardware. It analyzes patterns in accelerometer data, heart rate variability, and other physiological parameters already being collected for fitness tracking, transforming routine health monitoring data into meaningful eating event detection through software algorithms.
Data Source
AI summary
Disclosed herein are techniques related to device performance mode adjustment based on activity detection. In some embodiments, the techniques involve detecting, based on processing sensor data obtained from one or more gesture sensors, an activity in which a user of the one or more gesture sensors is engaged. The techniques further involve generating information corresponding to the detected activity in which the user of the one or more gesture sensors is engaged. The techniques also involve controlling, based on providing the generated information to a device for monitoring a physiological characteristic of the user, adjustment of a performance mode of the device for monitoring the physiological characteristic of the user.