Omics-Inferred Body Index for Metabolic Health Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Body Mass Index (BMI) is inadequate for capturing the heterogeneous metabolic health states of individuals, leading to misclassification and failing to reflect actual metabolic health, particularly in identifying metabolically unhealthy normal-weight and metabolically healthy obese groups, and does not account for variations in body composition and ethnic differences.
Innovation Solution
A computer-implemented method using machine learning models to generate an omics-inferred body index from blood analyte data, combining omics data with anthropomorphic data to classify subjects into more reflective metabolic health categories, providing a system for determining an omics-inferred body index that better captures metabolic health states.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If BMI is used to classify obesity, then the classification is simple and easily understood, but it misclassifies individuals with high muscle-to-fat ratio and fails to capture heterogeneous metabolic health states
Solution Approach 1:
The patent segments the homogeneous BMI classification into heterogeneous metabolic health states by using multiple omics layers (proteomics, metabolomics, lipidomics) to distinguish different metabolic phenotypes within BMI categories. This allows differentiation of metabolically healthy obese, metabolically unhealthy obese, and other metabolic states that BMI cannot distinguish.
Solution Approach 2:
The patent transitions from a single-dimensional BMI measure to a multi-dimensional omics-based classification system. By incorporating multiple omics data types and machine learning algorithms, the system adds biological and metabolic dimensions to the classification, enabling precise identification of metabolic health states while maintaining BMI as a reference.
2Stability of the object's composition
If BMI thresholds are used for obesity diagnosis, then the diagnosis is consistent across populations, but it does not account for ethnic differences and body composition variations
Solution Approach 1:
The patent applies local quality by customizing metabolic health assessment for different ethnic populations and body composition types. The omics-based classification system identifies distinct metabolic phenotypes specific to different populations, allowing tailored health recommendations that account for ethnic differences and body composition variations while maintaining overall framework consistency.
3Device complexity
If single targeted metrics or specific biomarkers are used, then the analysis is focused and simple, but it cannot comprehensively bridge the gaps between BMI and heterogeneous physiological states
Solution Approach 1:
The patent merges multiple omics data types (proteomics, metabolomics, lipidomics) with BMI and other clinical measurements in an integrated classification system. This combination allows comprehensive capture of metabolic health states by synthesizing information from different biological layers, overcoming the limitations of single biomarker or single-metric approaches.
Solution Approach 2:
The patent creates a universal classification system that can identify multiple metabolic health states (metabolically healthy obese, metabolically unhealthy obese, normal weight with metabolic abnormalities) using a single integrated omics-based framework. The system serves multiple functions: classification, risk stratification, and intervention guidance, while maintaining comprehensive coverage of metabolic heterogeneity.
Data Source
AI summary
Provided are computer-implemented methods, systems and products of determining omic body index and class of a subject.