Clinical Decision Support System Secondary Parameter Approximation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Clinical decision support systems face challenges in providing timely and accurate secondary parameters, such as pressure and blood flow velocity, due to the complexity and time-consuming nature of computational fluid dynamics simulations.
Innovation Solution
A method utilizing a machine learning mechanism, specifically deep learning, to approximate secondary parameters from shape data sets, reducing the need for computationally expensive calculations and enabling faster provision of relevant information to clinicians.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If computational fluid dynamics simulation is used to calculate pressure and blood flow velocity, then measurement precision is improved, but loss of time increases
Solution Approach 1:
The patent pre-calculates and stores mapping relationships between shape parameters and secondary parameters (pressure, blood flow velocity) in a database during system initialization. When a shape data set is input, the system quickly retrieves pre-computed results based on shape similarity, avoiding time-consuming CFD simulations while maintaining measurement precision.
Solution Approach 2:
The patent creates simplified shape representations (shape data sets with extracted geometric features) that capture essential characteristics without full computational complexity. These shape copies enable rapid comparison and retrieval of pre-computed secondary parameters, providing accurate results without repeated complex simulations.
2Reliability
If computational fluid dynamics simulation is used to calculate secondary parameters, then reliability is improved, but device complexity increases
Solution Approach 1:
The patent divides the problem into two independent modules: (1) shape analysis module that extracts geometric features from shape data sets, and (2) parameter retrieval module that queries pre-computed results from database. This segmentation maintains reliability by preserving accurate shape-characteristic relationships while reducing system complexity through modular design and pre-computation.
Solution Approach 2:
The system performs complex CFD simulations once during database population, storing results for rapid retrieval. This preliminary computation ensures reliability of secondary parameters while avoiding repeated complex calculations, thereby reducing operational system complexity.
3Productivity
If machine learning mechanism is used to approximate secondary parameters, then productivity is improved, but manufacturing precision may deteriorate
Solution Approach 1:
The patent uses shape data sets as simplified copies that capture essential geometric characteristics without full computational complexity. These shape representations enable rapid machine learning processing while maintaining sufficient precision for clinical decision support by focusing on clinically relevant geometric features.
Solution Approach 2:
The patent transforms complex shape data into extracted geometric parameters (shape descriptors) that serve as input features for machine learning. This parameter transformation maintains approximation accuracy by preserving clinically significant geometric relationships while enabling fast machine learning inference for high productivity.
Data Source
AI summary
A method for providing a secondary parameter, a decision support system, a computer-readable medium and a computer program product are disclosed. In an embodiment, the method is for providing a secondary parameter in a decision support system providing a primary parameter, in particular in a clinical decision support system. The method includes: providing an input data set; approximating a secondary parameter based on the input data set by using a sub-system being trained by a machine learning mechanism, in particular by a deep learning mechanism; and providing the approximated secondary parameter. The input data set is a shape data set.