Machine Learning Biomarker Models for Multiple Sclerosis Activity Assessment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods lack effective tools to accurately assess multiple sclerosis activity in individuals, particularly for diagnosing MS, tracking disease activity, identifying relapses, determining treatment response, and monitoring remission, with magnetic resonance imaging (MRI) primarily reflecting historical damage rather than dynamic biological processes.
Innovation Solution
Development and validation of predictive models using machine learning techniques such as random forest, stochastic gradient boosting, and Lasso algorithms applied to quantitative expression values of specific biomarkers in blood samples to diagnose MS, assess disease activity, relapse, flare, and remission, with biomarkers including PON1, Myoglobin, PAI1, TIMP1, SDF1, IL6Rbeta, Cystatin B, IgE, MIP3beta, and others, to generate scores indicative of MS activity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If MRI is used to assess multiple sclerosis, then historical damage can be visualized, but dynamic biological processes cannot be detected
Solution Approach 1:
The patent replaces the mechanical/imaging-based MRI system with a biochemical detection system using blood-based biomarkers and machine learning algorithms. This substitution enables detection of dynamic biological processes through molecular signatures in blood samples, achieving the goal of detecting active disease processes without relying on structural imaging that only shows historical damage.
Solution Approach 2:
The patent introduces blood-based biomarkers as intermediary molecules that reflect dynamic biological processes in the central nervous system. These biomarkers serve as mediators between the complex neurological processes and the diagnostic tool, allowing indirect but dynamic assessment of disease activity through accessible blood samples rather than direct brain imaging.
2Ease of operation
If a non-invasive blood test is developed, then ease of operation improves, but measurement precision must be maintained
Solution Approach 1:
The patent transforms the diagnostic approach by changing from detecting structural parameters (MRI images) to measuring biochemical parameters (biomarker expression levels). This parameter change enables use of simple blood collection while achieving precise measurement of disease activity through quantitative biomarker analysis and machine learning classification.
Solution Approach 2:
The patent develops a universal blood-based assay platform that can assess multiple aspects of MS activity (disease activity, treatment response, relapse risk) through a single sample type. This multi-functional approach maintains high measurement precision across different clinical questions while improving ease of operation by using the same simple blood collection method for all assessments.
Data Source
AI summary
Markers useful for determining multiple sclerosis activity in a human subject are provided, along with kits for measuring quantitative expression values of the markers. Also provided are computer systems and software embodiments of predictive models for scoring and determining multiple sclerosis activity in human subjects based on the quantitative expression values of the markers.