Prescriptive Therapy Instruction Sets Using Tolerance-Aware Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for selecting prescriptive therapies are unreliable, leading to potential adverse reactions and increased healthcare costs due to the lack of accurate consideration of individual responses to treatments.
Innovation Solution
A system and method using a computing device to train a prescriptive therapy instruction set machine learning model with subject prescription and tolerance data, allowing for the generation and retraining of therapy instructions based on updated tolerance data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional methods are used for selecting prescriptive therapies, then the selection process is simple, but the accuracy and reliability of therapy selection deteriorates
Solution Approach 1:
The patent replaces traditional mechanical/manual therapy selection methods with an artificial intelligence-based machine learning system. The machine learning model automatically analyzes patient data, prescription data, and tolerance data to generate therapy recommendations, substituting human decision-making processes with computational algorithms that provide more accurate and reliable selections.
Solution Approach 2:
The patent introduces a machine learning model as an intermediary between patient data and therapy selection. This intermediary processes and correlates multiple data sources (patient profiles, prescription histories, tolerance data) to generate informed therapy recommendations, acting as a bridge that transforms raw data into actionable medical decisions.
2Reliability
If individual patient data is not considered, then the therapy selection process is faster, but adverse reactions increase
Solution Approach 1:
The patent performs preliminary actions by collecting and storing patient data, prescription data, and tolerance data before therapy selection is needed. The machine learning model is pre-trained on this accumulated data, enabling it to quickly generate accurate therapy recommendations when needed, thus reducing adverse reactions without sacrificing selection speed during actual use.
Solution Approach 2:
The patent establishes continuous data collection and model retraining processes. As new patient data, prescription data, and tolerance data become available, the system continuously updates its knowledge base and retrainsthe machine learning model, ensuring that therapy selections always benefit from the latest information while maintaining efficient selection processes.
3Measurement precision
If machine learning models are trained and retrained with updated data, then therapy selection accuracy improves, but computational resources and time increase
Solution Approach 1:
The patent applies partial retraining rather than complete retraining of the machine learning model. When new data becomes available, only specific components or layers of the model are updated rather than retraining the entire model from scratch. This reduces computational resource consumption and energy usage while still incorporating new information to maintain or improve recommendation accuracy.
Data Source
AI summary
A system for determining a prescriptive therapy instruction set may include a computing device configured to receive a first subject prescription datum associated with a first subject; receive a first subject tolerance datum associated with the first subject; determine a prescriptive therapy instruction set by training a prescriptive therapy instruction set machine learning model on a training dataset including a plurality of example subject prescription data and subject tolerance data as inputs correlated to a plurality of example prescriptive therapy instruction sets as outputs; and generating the prescriptive therapy instruction set as a function of the subject prescription datum and the subject tolerance datum using the trained prescriptive therapy instruction set machine learning model; receive a second subject tolerance datum associated with the first subject; and retrain the prescriptive therapy instruction set machine learning model as a function of the second subject tolerance datum.