Personalized Biological Age Prediction Model Using Segmented Logistic Regression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing biological age prediction models, such as multivariable linear regression and principal component analysis, fail to accurately represent individual aging status due to overestimation in younger individuals and underestimation in older individuals, and they present biological age as a single numerical value without clear interpretation.
Innovation Solution
A method and system for generating a personalized biological age prediction model using binary logistic regression models, which calculate an 'excess age' by dividing individuals into age groups and using medical checkup data to predict a biological age probability spectrum, rather than relying solely on chronological age.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multivariable linear regression or principal component analysis is used to predict biological age, then a single numerical value is obtained, but the prediction accuracy deteriorates with overestimation in younger individuals and underestimation in older individuals
Solution Approach 1:
The patent divides the continuous age range into discrete age groups (e.g., 20-29, 30-39, 40-49 years old) and creates separate prediction models for each group. This segmentation allows each model to be optimized for specific age-related physiological characteristics, eliminating the overestimation in younger and underestimation in older individuals that plagues single-model approaches.
Solution Approach 2:
The patent applies different prediction models (binary logistic regression, Cox proportional hazards model, or random survival forest) to different age groups based on their specific characteristics. Each age group receives a locally optimized model rather than a universal model, improving prediction accuracy for each specific age range while maintaining overall reliability across the full age spectrum.
2Ease of operation
If a single numerical value is presented for biological age, then the result is simple, but the interpretation becomes unclear and subjective
Solution Approach 1:
The patent transforms the single-dimensional biological age value into a multi-dimensional output that includes both a numerical biological age and a probability spectrum showing the distribution across different age groups. This dimensional expansion provides clear interpretation by showing not just the predicted age but also the confidence distribution, making the results both simple to understand and rich in information.
3Measurement precision
If binary logistic regression models are generated for each age unit, then prediction accuracy improves, but the model complexity increases
Solution Approach 1:
The patent combines multiple binary logistic regression models (one for each age group) into a unified prediction system. While individual models are simple, their integration creates a comprehensive system that maintains the accuracy benefits of age-specific modeling while providing a streamlined interface for users. The system automatically selects and applies the appropriate model based on the input data's age characteristics.
Data Source
AI summary
A method and a system thereof for generating a personalized biological age prediction model are proposed. The method and the system generate a model capable of predicting a biological age for each individual by obtaining an excess age for a chronological age for each age on the basis of medical checkup data. More particularly, the method and the system build a biological age prediction model by gender and chronological age group in consideration of aging mechanisms different from each other according to the gender and the chronological age group, and enable predicting the biological age according to the biological age prediction model for each age group.