Medication Dose Prediction with Adaptive Models for Incomplete Test Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing medication management systems struggle to accurately predict the influence of medication on patients due to incomplete test parameter data, limiting the versatility of predictive models, especially in intensive care units where heparin dosing is complex and not optimally managed.
Innovation Solution
A medication management system utilizing a machine learning model that generates multiple learning models trained on different test parameters, allowing for prediction of medication influence and dose recommendations, even with incomplete data, by selecting the appropriate model based on the available test parameters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a predictive system uses test parameters such as APTT as explanatory variables, then the prediction accuracy for heparin dosing is improved, but the versatility of the system deteriorates because it cannot be applied to patients whose data are not complete
Solution Approach 1:
The patent divides the predictive system into multiple independent prediction models, each trained on a specific subset of test parameters (e.g., APTT-based model, ACT-based model, or combined model). This segmentation allows the system to select and apply the most appropriate model based on which test parameters are available for a given patient, thereby maintaining high prediction accuracy while improving versatility across different clinical settings and data completeness scenarios.
2Adaptability or versatility
If multiple learning models are generated for different test parameters, then the adaptability to different medical institutions is improved, but the device complexity increases
Solution Approach 1:
The patent implements a dynamic model selection mechanism that automatically chooses the appropriate prediction model based on the availability and type of test parameters for each patient. The system dynamically adjusts which model (APTT-based, ACT-based, or combined) is applied to the current patient case, allowing the system to adapt to different medical institutions and data scenarios without requiring manual configuration or increasing operational complexity for the user.
Data Source
AI summary
A medication management system includes a display device, a memory, and a processor configured to acquire patient information about a patient, a first test value of a laboratory test performed on the patient, and medicine information indicating a medicine to be administered, execute a call to a machine learning model with the patient information, the first test value, and the medicine information to determine a second test value expected at a predetermined time after the medicine is administered, the model trained with test values obtained before and after administration of the medicine to different patients, determine a recommended dose or a preferred range of doses for the medicine based on the second test value, generate a graph showing a relationship between doses of the medicine and test values, and display the graph and the recommended dose or the preferred range on the graph.