Automatic Insulin Dosage via Meal Purpose Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current automatic insulin delivery systems struggle with accurately detecting meals and determining their purpose, leading to potential under-delivery or over-delivery of insulin due to user error or difficulty in estimating carbohydrate intake.
Innovation Solution
A system that utilizes a non-transitory computer readable medium with programming code executable by a processor to receive data from various data sources, including direct and indirect sources, to determine the purpose of a detected meal and generate appropriate insulin dosage instructions for a drug delivery device.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If users manually input carbohydrate intake into the device, then the device can control therapy based on accurate meal information, but the process is time-consuming and complex for users
Solution Approach 1:
The system automatically detects meals using multiple data sources (glucose sensor data, activity data, location data) without requiring user input. The processor evaluates data from these sources to autonomously determine meal occurrence and purpose, making the system self-sufficient and eliminating the need for manual user operation.
Solution Approach 2:
The patent replaces the mechanical/manual input system with an automated electronic detection system. Instead of users physically entering carbohydrate information, the system uses sensors, processors, and algorithms to automatically detect and analyze meal-related data from multiple indirect sources, substituting electronic automation for manual mechanical input.
2Loss of information
If users estimate carbohydrate content of meals, then they can provide meal information to the device, but estimation errors lead to under-delivery or over-delivery of insulin
Solution Approach 1:
The system introduces multiple intermediary data sources between the user and the insulin delivery decision. Instead of relying directly on user estimation, the processor evaluates glucose sensor data, activity data, location data, and other indirect measurements as intermediaries to infer meal occurrence and characteristics, reducing reliance on potentially inaccurate user estimates.
Solution Approach 2:
The system uses glucose sensor data as feedback to verify and refine meal detection. By monitoring glucose level changes and comparing them against expected patterns, the system can confirm or adjust its meal detection conclusions, creating a feedback loop that improves measurement accuracy over time.
3Extent of automation
If the device controls all therapy automatically without meal purpose consideration, then therapy management is simplified, but the system cannot differentiate between different types of meals (e.g., rescue meals vs. regular meals)
Solution Approach 1:
The system dynamically adjusts its meal detection and classification based on real-time data evaluation. The processor analyzes multiple data sources and adapts its determination of meal purpose (rescue meal, preemptive meal, regular meal) based on the specific pattern of glucose changes, activity levels, and contextual information, allowing flexible adaptation to different meal scenarios.
Solution Approach 2:
The patent segments the meal detection process into distinct analytical components: detecting glucose changes, evaluating activity data, analyzing location data, and determining meal purpose. This segmentation allows the system to automatically control therapy while simultaneously differentiating between various meal types by evaluating each segment independently and combining the results.
Data Source
AI summary
Disclosed are examples of a device, a system, methods and computer-readable medium products operable to implement functionality to determine and respond to a purpose of a meal. An algorithm or application may receive data that may include data related to a meal purpose from data sources and determine whether any of the data received from the plurality of data sources was received from a direct data source or an indirect data source. The data may be evaluated to determine a purpose of the meal. Based on the results of the evaluation, instructions may be generated to provide an appropriate response based on the determined purpose of the meal. The generated instructions to provide the appropriate response based on the determined purpose of the meal may be output.