Patient-Specific Blood Flow Modeling With Reduced-Order ML

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for calculating blood flow characteristics in vascular systems, such as the coronary arteries, face challenges in achieving accuracy while maintaining computational efficiency, particularly in localized regions where simplified one-dimensional models are inadequate.

Innovation Solution

A method involving reduced order models and machine learning algorithms is employed to generate patient-specific models, incorporating impedance values and geometric features, which are used to create feature vectors for predicting blood flow characteristics, and then refined using machine learning to improve accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of time

If simplified one-dimensional models are used to calculate blood flow characteristics, then computational time is reduced, but measurement precision and manufacturing precision deteriorate in localized regions

Engineering Contradiction:
Improvecomputational timeVSAvoidblood flow characteristics accuracy
Core Design Contradiction:
Loss of timeVSMeasurement precision

Solution Approach 1:

The vascular system is segmented into different regions with different levels of model complexity. One-dimensional models are used for bulk regions where computational efficiency is prioritized, while three-dimensional models are applied to localized regions (such as stenosis areas, bifurcations, or regions with complex geometry) where high measurement precision is required. This segmentation allows the system to achieve both computational efficiency and accuracy where needed.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Different parts of the vascular model are assigned different qualities of detail. The overall model structure uses simplified one-dimensional representations for computational efficiency, but specific localized regions are enhanced with three-dimensional geometry and detailed Navier-Stokes equations to maintain high measurement precision in critical areas. This local quality approach resolves the contradiction by applying complexity only where measurement precision is most important.

Inventive Principle:
Principle #3Local quality

2Measurement precision

If three-dimensional Navier-Stokes equations are solved for accurate blood flow calculation, then measurement precision improves, but computational time increases significantly

Engineering Contradiction:
Improveblood flow characteristics accuracyVSAvoidcomputational time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The computational domain is segmented into regions that can be solved using different methods. Three-dimensional Navier-Stokes equations are solved only in localized regions requiring high precision (such as stenosis regions or bifurcations), while one-dimensional models are used in the remaining bulk regions. This segmentation dramatically reduces the total computational time while maintaining accuracy where it matters most.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Instead of applying the computationally intensive three-dimensional Navier-Stokes equations throughout the entire vascular system (excessive action), the method applies them only partially to the specific regions where high measurement precision is required. This partial action approach achieves the necessary accuracy in critical regions while avoiding the excessive computational cost of applying the full three-dimensional model everywhere.

Inventive Principle:
Principle #16Partial or excessive action

3Device complexity

If reduced order models are used to simplify geometry, then device complexity is reduced, but manufacturing precision and measurement precision deteriorate

Engineering Contradiction:
Improvegeometric complexityVSAvoidgeometric accuracy
Core Design Contradiction:
Device complexityVSManufacturing precision

Solution Approach 1:

The vascular geometry is segmented into simplified representations for most regions and detailed three-dimensional representations for critical regions. The reduced order models provide sufficient geometric accuracy for bulk regions while the full three-dimensional geometry is retained in localized regions where precise measurement of blood flow characteristics is required, thus balancing device complexity reduction with manufacturing precision requirements.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Different levels of geometric detail are assigned to different regions. The overall vascular model uses simplified geometry to reduce device complexity, but specific localized regions maintain high manufacturing precision through detailed three-dimensional geometry. This local quality approach ensures that geometric accuracy is preserved where it directly impacts blood flow measurement precision while allowing complexity reduction elsewhere.

Inventive Principle:
Principle #3Local quality

Data Source

PatentUS20250359757A1Systems and methods for estimation of blood flow characteristics using reduced order model and/or machine learning
Publication Date: 2025.11.27 HEARTFLOW INC
  • US20250359757A1 patent drawing
  • US20250359757A1 patent drawing
  • US20250359757A1 patent drawing

AI summary

Systems and methods are disclosed for determining blood flow characteristics of a patient. One method includes: receiving, in an electronic storage medium, patient-specific image data of at least a portion of vasculature of the patient having geometric features at one or more points; generating a patient-specific reduced order model from the received image data, the patient-specific reduced order model comprising estimates of impedance values and a simplification of the geometric features at the one or more points of the vasculature of the patient; creating a feature vector comprising the estimates of impedance values and geometric features for each of the one or more points of the patient-specific reduced order model; and determining blood flow characteristics at the one or more points of the patient-specific reduced order model using a machine learning algorithm trained to predict blood flow characteristics based on the created feature vectors at the one or more points.