Forecast of MRI images by means of a forecast model trained by supervised learning

Patent metadata

Field	Value
Publication number	US-12148163-B2
Application number	US-202318340051-A
Country	US
Kind code	B2
Filing date	Jun 23, 2023
Priority date	Sep 18, 2019
Publication date	Nov 19, 2024
Grant date	Nov 19, 2024

Abstract

The present disclosure deals with the quickening of MRI examinations. Subjects of the present disclosure are a method, a system, a computer program product, a use, a contrast agent for use and a kit.

First claim

We claim:

1. A computer-implemented method comprising:
receiving at least two magnetic resonance imaging (MRI) images, wherein at least one received MRI image of the at least two MRI images shows an examination region before administering a contrast agent, and at least one received MRI image of the at least two MRI images shows the examination region after administering the contrast agent, wherein the contrast agent is a hepatobiliary contrast agent comprising a disodium salt of gadolinium ethoxybenzyl diethylenetriaminepentaacetic acid;
feeding the at least two MRI images to a prediction model, the prediction model having been trained by means of supervised learning to predict based on a multiplicity of reference MRI images, some of which show the examination region during a first time span, some of which show the examination region during a second time span, the first time span starting before administering the contrast agent and ending at a time span after administering the contrast agent, and the second time span following the first time span chronologically;
generating one or more predicted MRI images showing the examination region during the second time span by means of the prediction model, the second time span following the first time span chronologically; and
displaying the one or more predicted MRI images, outputting the one or more predicted MRI images, or storing the one or more predicted MRI images in a data storage medium.

2. The method of claim 1 , wherein the examination region is a liver or a portion of a liver of a mammal.

3. The method of claim 2 , wherein the first time span is chosen such that the at least two MRI images show the examination region in different phases, wherein the phases comprise a native phase, an arterial phase, a portal-vein phase, and/or a late phase, wherein at least one MRI image is received which shows the examination region in the native phase, and at least one MRI image is received which shows the examination region during the arterial phase, and/or at least one MRI image is received which shows the examination region in the portal-vein phase, and/or at least one MRI image is received which shows the examination region in the late phase.

4. The method of claim 1 , wherein the at least two MRI images show a liver or a portion of a liver of a mammal prior to a time point TP0 and during a time span from TP0 to TP1 and/or during a time span from TP1 to TP2 and/or during a time span from TP2 to TP3 and/or during a time span from TP3 to TP4, at time point TP0 the contrast agent is administered intravenously as a bolus and then reaches liver cells via liver arteries and liver veins, at time point TP1 the contrast agent in the liver arteries reaches a maximum concentration, at time point TP2 a signal intensity generated in the liver veins by the contrast agent assumes a value which is the same size as a value of a signal intensity generated in the liver arteries by the contrast agent, at time point TP3 the contrast agent in the liver veins reaches a maximum concentration, at time point TP4 a signal intensity generated in the liver cells by the contrast agent assumes a value which is the same size as a value of a signal intensity generated in the liver veins by the contrast agent.

5. The method of claim 1 , wherein the first time span starts within a time span of from one minute to one second before administration of the contrast agent or with the administration of the contrast agent, and lasts for a time span of from 2 minutes to 15 minutes from the administration of the contrast agent.

6. The method of claim 1 , wherein the second time span is within a hepatobiliary phase.

7. The method of claim 1 , wherein the second time span starts at least 10 minutes after administration of the contrast agent.

8. The method of claim 1 , wherein the prediction model is an artificial neural network.

9. A system comprising:
a receiving unit;
a control and calculation unit; and
an output unit; wherein
the control and calculation unit being configured to prompt the receiving unit to receive at least two magnetic resonance imaging (MRI) images, wherein at least one received MRI image of the at least two MRI images shows an examination region before administering a contrast agent, and at least one received MRI image of the at least two MRI images shows the examination region after administering the contrast agent, wherein the contrast agent is a disodium salt of gadolinium ethoxybenzyl diethylenetriaminepentaacetic acid,
the control and calculation unit being configured to predict one or more MRI images by means of a prediction model based on the at least two MRI images,

wherein the prediction model was trained by means of supervised learning to predict based on a multiplicity of reference MRI images, some of which show the examination region during a first time span, some of which show the examination region during a second time span, the first time span starting before administering the contrast agent and ending at a time span after administering the contrast agent, and the second time span following the first time span chronologically,
wherein the one or more predicted MRI images show the examination region during a second time span, the second time span following the first time span chronologically, and
the control and calculation unit being configured to prompt the output unit to display the one or more predicted MRI images, to output the one or more predicted MRI images, or to store the one or more predicted MRI images in a data storage medium.

10. A computer program product comprising at least one non-transitory computer-readable medium comprising one or more instructions that, when executed by a computer system, the computer system to execute the following:
receiving at least two magnetic resonance imaging (MRI) images, wherein at least one received MRI image of the at least two MRI images shows an examination region before administering a contrast agent, and at least one received MRI image of the at least two MRI images shows the examination region after administering the contrast agent, wherein the contrast agent is a disodium salt of gadolinium ethoxybenzyl diethylenetriaminepentaacetic acid;
feeding the at least two MRI images to a prediction model, the prediction model having been trained by means of supervised learning to predict based on a multiplicity of reference MRI images, some of which show the examination region during a first time span, some of which show the examination region during a second time span, the first time span starting before administering the contrast agent and ending at a time span after administering the contrast agent, and the second time span following the first time span chronologically;
generating one or more predicted MRI images showing the examination region during the second time span by means of the prediction model, the second time span following the first time span chronologically; and
displaying the one or more predicted MRI images, outputting the one or more predicted MRI images, or storing the one or more predicted MRI images in a data storage medium.

11. The computer program product of claim 10 , wherein the first time span is chosen such that the at least two MRI images show the examination region in different phases, wherein the phases comprise a native phase, an arterial phase, a portal-vein phase, and/or a late phase, wherein at least one MRI image is received which shows the examination region in the native phase, and at least one MRI image is received which shows the examination region during the arterial phase, and/or at least one MRI image is received which shows the examination region in the portal-vein phase, and

Assignees

• Bayer Ag

Inventors

• Rohrer Martin
• Uber Iii Arthur

Classifications

• A61B5/055

  involving electronic [EMR] or nuclear [NMR] magnetic resonance, e.g. magnetic resonance imaging · CPC title

• G06N3/0464

  Convolutional networks [CNN, ConvNet] · CPC title

• G06N3/09

  Supervised learning · CPC title

• G06T2207/30101

  Blood vessel; Artery; Vein; Vascular · CPC title

• G06T2207/30056

  Liver; Hepatic · CPC title