Asthma Condition Estimation Using Symptom-Based Patient Clusters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for evaluating asthma disease conditions require invasive testing and specialized equipment, making them impractical for routine clinical settings, and do not account for individual patient responsiveness to treatment.
Innovation Solution
An asthmatic disease condition estimation server that classifies patients into clusters based on symptom information using machine learning, allowing for the identification of disease conditions without invasive testing, utilizing a storage to associate clusters with disease conditions and a determination unit to estimate disease states from asthma control questionnaire responses.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If invasive testing (bronchoscopy) and gene expression analysis are used to evaluate asthma disease conditions, then measurement precision is improved, but device complexity and ease of operation deteriorate
Solution Approach 1:
The patent creates a virtual model (clone) of the patient's asthma disease condition by training a machine learning model on gene expression data and clinical information from training patients. This virtual model can be copied and applied to evaluate disease conditions without requiring actual invasive testing equipment, thereby maintaining measurement precision while eliminating device complexity
Solution Approach 2:
The patent replaces the mechanical/invasive testing system (bronchoscopy, physical gene expression analysis equipment) with an information processing system that uses machine learning algorithms to analyze gene expression data and clinical information, substituting physical measurement devices with computational models
2Measurement precision
If invasive testing (bronchoscopy) is performed to extract airway epithelial cells, then measurement precision is improved, but object-affected harmful factors increase
Solution Approach 1:
The patent creates a virtual representation of the patient's disease state through machine learning models trained on existing data, eliminating the need for invasive bronchoscopy procedures that carry risks of complications, while still achieving accurate disease condition evaluation
Solution Approach 2:
The patent introduces gene expression data and clinical information as intermediary elements that can be obtained through less invasive means (such as sputum samples or blood tests) and serves as a mediator between the patient and the disease evaluation process, avoiding direct invasive intervention
3Measurement precision
If gene expression analysis is performed to classify asthma phenotypes, then measurement precision is improved, but loss of substance increases
Solution Approach 1:
The patent creates a computationally-derived copy of the gene expression profile that can be generated from smaller, cheaper samples (such as sputum or blood) rather than requiring large amounts of tissue from invasive bronchoscopy, thereby maintaining classification accuracy while reducing material loss and cost
Solution Approach 2:
The patent uses a subset of gene expression data and clinical information that is sufficient for accurate phenotype classification without requiring complete comprehensive analysis, thereby achieving the necessary measurement precision with reduced resource investment
4Measurement precision
If specialized testing equipment and expertise are required for asthma evaluation, then measurement precision is improved, but ease of operation deteriorates
Solution Approach 1:
The patent creates a portable, software-based evaluation system that copies the functionality of complex testing equipment into a machine learning model that can run on standard computers or mobile devices, eliminating the need for specialized hardware and making the system accessible in routine clinical settings
Solution Approach 2:
The patent develops a universal machine learning model that can process multiple types of input data (gene expression data, clinical information, questionnaire responses) and provide comprehensive disease condition evaluation across different asthma phenotypes, replacing multiple specialized tests with a single multi-functional system
Data Source
Figure 1
Figure 2
Figure 3
AI summary
To easily estimate a disease condition of a patient with asthma. An asthmatic disease condition estimation server (1) according to the present invention includes: a storage (2) that stores a plurality of clusters, into which patients with asthma are classified, and a plurality of disease conditions of asthma associated with each cluster; a reception unit (3) that receives symptom information indicating a symptom of a subject; a determination unit (4) that identifies the cluster to which the subject belongs among the plurality of clusters, based on the symptom information, and identifies the disease condition stored in the storage in association with the identified cluster; and an output unit (5) that outputs the identified disease condition as an estimated disease condition of the subject.