Methods and systems for determining hemodynamic parameters

Patent metadata

Field	Value
Publication number	US-12591970-B2
Application number	US-202318401378-A
Country	US
Kind code	B2
Filing date	Dec 30, 2023
Priority date	Dec 30, 2022
Publication date	Mar 31, 2026
Grant date	Mar 31, 2026

Abstract

Some embodiments of the present disclosure provide methods and systems for determining a hemodynamic parameter. The method may include: obtaining image data of a subject being acquired in a rest state; obtaining a trained machine learning model; and determining, based on the trained machine learning model, at least one target hemodynamic parameter of the subject. The trained machine learning model may be obtained based on multiple sets of sample image data. Each set of the multiple sets of sample image data may include a first image data and at least one of a second image data or a third image data. The first image data may be acquired in a rest state of a first sample subject, the second image data may be acquired in a hyperemic state of the first sample subject, and the third image data may be acquired in a hyperemic state of a second sample subject including the first sample subject.

First claim

What is claimed is:

1 . A system for determining hemodynamic parameters, comprising:
at least one storage medium including a set of instructions; at least one processor in communication with the at least one storage medium, wherein when executing the set of instructions, the at least one processor is directed to cause the system to perform operations including:
obtaining image data of a subject being acquired in a rest state;
obtaining a trained machine learning model; and
determining, based on the trained machine learning model, at least one target hemodynamic parameter of the subject, wherein the trained machine learning model is obtained based on multiple sets of sample image data, each set of the multiple sets of sample image data includes a first image data and at least one of a second image data or a third image data, the first image data is acquired in a rest state of a first sample subject, the second image data is acquired in a hyperemic state of the first sample subject, and the third image data is acquired in a hyperemic state of a second sample subject associated with the first sample subject.

2 . The system of claim 1 , where the determining, based on the trained machine learning model, at least one hemodynamic parameter of the subject includes:
determining, based on the image data of the subject, a structure model of the subject; the image data of the subject includes the first image data and at least one of the second image data or the third image data, the first image data is acquired in a rest state of the first sample subject, the second image data is acquired in a hyperemic state of the first sample subject, and the third image data is acquired in a hyperemic state of a second sample subject associated with the first sample subject; determining one or more boundary conditions associated with the subject; and determining the at least one target hemodynamic parameter of the subject by inputting the structure model, and the one or more boundary conditions into the trained machine learning model.

3 . The system of claim 1 , where the determining, based on the trained machine learning model, at least one hemodynamic parameter of the subject includes:
determining the at least one target hemodynamic parameter of the subject by inputting the image data of the subject into the trained machine learning model; the image data of the subject includes the first image data and at least one of the second image data or the third image data, the first image data is acquired in a rest state of the first sample subject, the second image data is acquired in a hyperemic state of the first sample subject, and the third image data is acquired in a hyperemic state of a second sample subject associated with the first sample subject.

4 . The system of claim 1 , where the trained machine learning model is obtained based on multiple groups of training samples, training samples in each group are determined based on one set of the multiple sets of sample image data, a training sample in each group includes a computational fluid dynamics (CFD) result determined based on the first image data, or the second image data, or the third image data and the first image data, or the third image data and the second image data, and a reference hemodynamic parameter corresponding to the CFD result.

5 . The system of claim 4 , wherein the training samples in each group include a first sample and at least one of a second sample, a third sample, or a fourth sample,
the first sample includes a first CFD result determined based on the first image data; the second sample includes a second CFD result determined based on the second image data; the third sample includes a third CFD result determined based on the first image data and the third image data; and the fourth sample includes a fourth CFD result determined based on the second image data and the third image data.

6 . The system of claim 5 , where the first sample is obtained according to operations including:
determining, based on the first image data, a first structure model representing the first sample subject; determining, based on the first structure model, one or more first boundary conditions of the first sample subject; and determining, based on the one or more first boundary conditions and the first structure model, the first CFD result.

7 . The system of claim 5 , where the second sample is obtained according to operations including:
determining, based on the second image data, a second structure model representing the first sample subject; determining, based on the second structure model or the second image data, one or more second boundary conditions of the first sample subject; and determining, based on the one or more second boundary conditions and the second structure model, the second CFD result.

8 . The system of claim 7 , wherein the determining, based on the second image data, a second structure model representing the first sample subject includes:
obtaining a second trained machine learning model; and determining the second structure model representing the first sample subject in the hyperemic state using the second trained machine learning model.

9 . The system of claim 5 , wherein the third sample is obtained according to operations including:
determining, based on the first image data, a first structure model representing the first sample subject; determining, based on the third image data, one or more third boundary conditions of the first sample subject; and determining, based on the one or more third boundary conditions and the first structure model, the third CFD result.

10 . The system of claim 9 , wherein the determining, based on the third image data, one or more third boundary conditions of the first sample subject includes:
determining a first region from the first image data; determining a parametric image representing functional indexes of different portions of the first sample subject based on the third image data; determining a second region in the first region from the parametric image by registering the parametric image with the first image data; and determining, based on the second region, the one or more third boundary conditions.

11 . The system of claim 5 , wherein the fourth sample is obtained according to operations including:
determining, based on the second image data, a second structure model representing the first sample subject; determining, based on the third image data, one or more third boundary conditions of the first sample subject; and determining, based on the one or more third boundary conditions and the second structure model, the fourth CFD result.

12 . The system of claim 5 , wherein the second sample or the fourth sample is obtained according to operations including:
for one of the second sample and the fourth sample; registering the second image data with the first image data to obtain a deformation field; and obtaining the one of the second sample and the fourth image based on the deformation field.

13 . A system for determining hemodynamic parameters, comprising:
at least one storage medium including a set of instructions; at least one processor in communication with the at least one storage medium, wherein when executing the set of instructions, the at least one processor is directed to cause the system to perform operations including:
obtaining first image data of a subject being acquired in a rest state;
obtaining second image data of the subject acquired in a hyperemic state;
determining, based on the first image data, a structure model representing the subject;
determining, based on

Assignees

• Shanghai United Imaging Healthcare Co Ltd

Inventors

• Wang Xiaodong
• Guo Jian
• Qin Peiming
• Kong Deyuan

Classifications

• G06T2207/30104

  Vascular flow; Blood flow; Perfusion · CPC title

• A61B5/02028

  Determining haemodynamic parameters not otherwise provided for, e.g. cardiac contractility or left ventricular ejection fraction · CPC title

• Y02T90/00

  Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation · CPC title

• G06T7/0012 Primary

  Biomedical image inspection · CPC title