NLP Matrix for Tumor Microenvironment Characterization

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current methods for characterizing a patient's tumor microenvironment (TME) are time-consuming, prone to human bias, and lack comprehensive information, making it difficult to determine the most effective cancer treatment.

Innovation Solution

A computer-implemented machine-learning method that uses a trained natural language processing model to extract relationship information between cell types and gene names from biomedical text, and generates a matrix for accurate TME characterization, reducing human bias and increasing efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If bulk RNA sequencing is used to characterize TME, then gene expression levels can be determined, but cell type-specific information is lost and manual literature research is required

Engineering Contradiction:

Improvecompleteness of TME characterizationVSAvoidtime for manual literature research

Core Design Contradiction:

Quantity of substanceVSLoss of time

Solution Approach 1:

The system uses machine learning models to automatically extract and process TME characterization information from literature, eliminating the need for manual literature research while maintaining comprehensive cell type-specific information

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent replaces manual mechanical literature research with automated machine learning-based information extraction systems that process scientific texts to obtain TME data

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Reliability

If manual literature research is used to find marker genes, then established marker genes can be identified, but the process is time-consuming and prone to human bias

Engineering Contradiction:

Improvereliability of marker gene identificationVSAvoidspeed of marker gene discovery

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The system replaces manual literature research with machine learning models that automatically extract and validate marker genes, improving both speed and reducing human bias while maintaining reliability through systematic processing

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The machine learning system processes literature data systematically and can validate findings against established knowledge, providing feedback loops that ensure reliability while maintaining high productivity

Inventive Principle:

Principle #23Feedback

3Measurement precision

If human researchers manually process TME data, then comprehensive analysis can be performed, but the process is slow and biased

Engineering Contradiction:

Improveprecision of TME characterizationVSAvoidtime for data processing

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent replaces manual data processing with automated machine learning systems that maintain measurement precision through systematic extraction and processing of TME information while dramatically reducing processing time

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

4Manufacturing precision

If comprehensive TME characterization is performed, then accurate patient stratification is achieved, but the process is complex and time-consuming

Engineering Contradiction:

Improveprecision of patient stratificationVSAvoidcomplexity of characterization process

Core Design Contradiction:

Manufacturing precisionVSDevice complexity

Solution Approach 1:

The system replaces complex manual characterization processes with machine learning models that automatically process data, maintaining high precision in patient stratification while simplifying the overall process complexity through automation

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS20250054567A1Machine-learning extraction of biomedical information and optimized characterization of a tumor micro environment of a patient

Publication Date: 2025.02.13 NEC LAB EURO GMBH

AI summary

A computer-implemented machine-learning method characterizes a tumor micro environment. The method includes: using a trained natural language processing machine learning model (NLP-model), extracting facts from biomedical text indicating relationship information between cell types and found gene names; using a reference database having gene names and aliases, grouping the extracted facts according to associated genes to generate extracted and grouped information; and generating a matrix from the extracted and grouped information with a first axis representing cell types and second axis representing genes. Each value of the matrix is calculated based on an importance of an associated gene taken and an associated weight. The associated weight is based associated publication meta information and/or an associated detection method's robustness and reliability. The method has applications including, but not limited to, use cases in drug development, medical artificial intelligence (AI)/healthcare for optimization of predictions or to support decision making.