CVD Risk Prediction System Using Minimal Multi-Omics Factors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for predicting the risk of atherosclerotic cardiovascular disease (CVD) are inaccurate and unreliable, focusing on correlations rather than causations, and require a large number of irrelevant variables, making them impractical for widespread implementation and failing to provide actionable insights for preventive measures.
Innovation Solution
A system and method using a minimal viable set of human-curated risk factors across multiple biological levels, including demographic, biomarker, comorbidity, genetic, and lifestyle factors, to predict the risk of atherosclerotic plaque build-up, integrating multi-omics data and clinical information for actionable and interpretable results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a purely ML-driven approach with many variables is used, then the model can capture correlations, but the prediction accuracy and reliability for atherosclerotic CVD risk is improved only marginally while the complexity and data requirements increase significantly
Solution Approach 1:
The patent extracts and removes irrelevant variables from the prediction system. It identifies and eliminates variables with no direct link to atherosclerotic CVD pathophysiology, such as infections, allergies, fractures, glasses worn, and speed of sound through heel. This extraction process reduces the variable set from 473 to a focused subset of clinically relevant risk factors, thereby reducing system complexity while maintaining prediction accuracy.
Solution Approach 2:
The patent changes the parameters by redefining the CVD outcome variable to specifically focus on atherosclerotic plaque build-up rather than a broad unspecific CVD definition that includes strokes more likely due to physical trauma. This parameter change ensures the model predicts the correct pathological process, improving reliability without requiring additional complexity.
2Loss of information
If 473 variables are used for prediction, then more correlations can be captured, but the availability of required variables for large numbers of patients decreases
Solution Approach 1:
The patent extracts only the essential risk factors needed for atherosclerotic CVD prediction, removing the 473-variable requirement down to a manageable subset. This extraction enables the system to be applied to large patient populations where collecting 473 variables per patient would be impractical, while still capturing the critical information needed for accurate risk stratification.
3Adaptability or versatility
If an unspecific CVD definition is used including strokes, then more cases are covered, but the prediction reliability for atherosclerotic plaque build-up decreases
Solution Approach 1:
The patent changes the outcome parameter from a broad unsspecific CVD definition to a specific definition focused on atherosclerotic plaque build-up. This parameter change excludes strokes more likely due to physical trauma and other non-atherosclerotic conditions, thereby improving the reliability and clinical actionability of the risk predictions for preventive interventions.
Data Source
AI summary
Techniques are described for predicting the risk of a patient to develop an atherosclerotic cardiovascular disease within a predefined time interval. A system receives a current set of risk factors associated with the risk of atherosclerotic plaque build-up in the patient's arteries. The current set comprises a set of human-curated risk factors covering multiple biological levels of the patient to capture the potential interactions or correlations detected between molecules in different biological levels. A predictor receives the current set as test input, wherein the predictor has been trained with a training data set including a number of training patient records. The predictor provides, in response to the test input, a predicted risk for developing an atherosclerotic cardiovascular disease for said patient within the predefined time interval.