Ultrasound system with automated dynamic setting of imaging parameters based on organ detection

Patent metadata

Field	Value
Publication number	US-11950959-B2
Application number	US-201917262435-A
Country	US
Kind code	B2
Filing date	Jul 19, 2019
Priority date	Jul 26, 2018
Publication date	Apr 9, 2024
Grant date	Apr 9, 2024

Abstract

The present disclosure describes ultrasound imaging systems and methods for ultrasonically inspecting biological tissue. An ultrasound imaging system according to the present disclosure may be configured to automatically apply tissue-specific imaging parameter settings ( 312, 314 ) based upon the automatic identification of the type of tissue being scanned. Tissue type identification ( 315 ) may be performed automatically for the images in the live image stream ( 304 ) and thus adjustments to the imaging settings may be applied automatically by using a neural network ( 320 ) and thus dynamically during the exam obviating the need for the sonographer to manually switch presets or adjust the imaging settings when moving to different portion of the anatomy.

First claim

What is claimed is:

1. An ultrasound imaging system with automated setting of imaging parameters during live imaging, the system comprising:
a probe configured to transmit ultrasound toward a subject for generating ultrasound images of biological tissue of the subject;
a processor configured to generate and to cause the ultrasound imaging system to display, in real-time, a live stream of ultrasound images of the biological tissue in accordance with a plurality of imaging parameters of the ultrasound system, wherein the processor is further configured to:
receive, in real-time, an ultrasound image from the live stream of ultrasound images;
receive an identification of a type of the biological tissue in the ultrasound image;
receive subject identification information, user identification information, or a combination thereof;
identify the subject as a recurring subject based at least in part on the subject identification information;
based on the type of the biological tissue and the subject identification information, generate at least one predicted setting for the recurring subject for at least one of the plurality of imaging parameters; and
apply the at least one predicted setting to the respective imaging parameter for subsequent live imaging of the recurring subject,
wherein the processor is configured to generate the at least one predicted setting using an artificial neural network model, and
wherein the artificial neural network model is a model trained using historical data extracted from system logs from multiple prior scans performed by a same user or of the recurring subject.

2. The system of claim 1 , wherein the artificial neural network model is a machine-learning regression model.

3. The system of claim 2 , wherein regression model is trained using data extracted from system logs from multiple ultrasound imaging systems.

4. The system of claim 1 , wherein the processor is configured to generate the at least one predicted setting further based on a current setting for the at least one of the plurality of imaging parameters.

5. The system of claim 4 , wherein the processor is configured to couple to a deep learning model a set of input parameters including a tissue type identification, a plurality of current settings for imaging parameters of the system including a current depth setting, a current gain setting, and a current focus setting, and at least one user identification input and at least one subject identification input.

6. The system of claim 1 , wherein, to generate the at least one predicted setting, the processor is configured to execute a trained neural network model comprising a plurality of layers including a first input layer configured to receive an input of size n+i and an output layer configured to generate an output of size n, and wherein n is equal to a number of the settings defined by the selected preset and i is equal to a number of additional input variables.

7. The system of claim 6 , wherein the trained neural network model comprises a plurality of intermediate fully connected layers, and wherein the output layer is a fully connected regression layer.

8. The system of claim 1 , wherein the processor is further configured to automatically determine the type of the biological tissue in the ultrasound image upon receiving the ultrasound image from the live stream of ultrasound images.

9. The system of claim 8 , wherein the processor is configured to use a machine-learning classification model to identify the type of the biological tissue.

10. The system of claim 9 , wherein the processor is configured to use a machine-learning classification model having a convolutional neural network architecture for identifying the type of the biological tissue.

11. The system of claim 1 , further comprising memory storing a plurality of presets, each defining one or more settings for at least one of the plurality of imaging parameters of the ultrasound imaging system, and wherein the processor is further configured to:
select one of the plurality of stored presets based on the type of the biological tissue; and
provide the one or more settings of the selected preset as inputs to the artificial neural network.

12. The system of claim 11 , wherein the processor is further configured to automatically adjusting corresponding one or more of the plurality of imaging parameters of the ultrasound system to settings defined by the selected preset prior to or while generating the at least one predicted setting.

13. A method of live ultrasound imaging of biological tissue, the method comprising:
receiving, in real-time, by a processor of an ultrasound system, an ultrasound image from a live stream of ultrasound images of a subject, wherein ultrasound images in the live stream, including the ultrasound image, are generated in accordance with current settings of a plurality of imaging parameters of the ultrasound system;
receiving an identification of a type of the biological tissue in the ultrasound image;
receiving subject identification information, user identification information, or a combination thereof;
identifying the subject as a recurring subject based at least in part on the subject identification information;
based on the type of the biological tissue and the subject identification information, generating a predicted setting for the recurring subject for at least one of the plurality of imaging parameters; and
automatically adjusting the current setting of the at least one of the of the plurality of imaging parameters in accordance with the predicted setting for subsequent live imaging of the recurring subject,
wherein the generating the predicted setting is based on use of an artificial neural network model, and
wherein the artificial neural network model is a model trained using historical data extracted from system logs from multiple prior scans performed by a same user or of the recurring subject.

14. The method of claim 13 , wherein the generating includes coupling the type of the biological tissue and the subject identification information to an artificial neural network comprising an input layer configured to receive a multi-dimensional input vector, a plurality of intermediate layers, and an output regression layer configured to output a multi-dimensional output vector having a smaller dimension than the input vector.

15. The method of claim 14 , wherein the output layer and the intermediate layers are fully connected layers.

16. The method of claim 14 , further comprising selecting one of a plurality of presets, each defining one or more settings for at least one of the plurality of imaging parameters of the ultrasound imaging system, from a memory of the ultrasound system, and providing the one or more settings of the selected preset as input to the artificial neural network.

17. The method of claim 16 , further comprising automatically adjusting one or more of the plurality of imaging parameters of the ultrasound system to settings defined by the selected preset prior to or while generating the at least one predicted setting.

18. The method of claim 13 , further comprising determining, by the processor, the type of the biological tissue in the ultrasound image responsive to receiving the ultrasound image from the live stream of ultrasound images.

19. The method of claim 18 , wherein the determining includes coupling the ultrasound image from the live stream to another artificial neural network trained to perform organ classification on an input ultrasound image.

20. A non-t

Assignees

• Koninklijke Philips Nv

Inventors

• Srinivasa Naidu Raghavendra
• Errico Claudia
• Xie Hua
• Wang Haibo
• Dianis Scott William

Classifications

• G06N3/09

  Supervised learning · CPC title

• G06N3/0464

  Convolutional networks [CNN, ConvNet] · CPC title

• A61B8/5207 Primary

  involving processing of raw data to produce diagnostic data, e.g. for generating an image · CPC title

• A61B8/5292 Primary

  using additional data, e.g. patient information, image labeling, acquisition parameters · CPC title

• A61B8/54

  Control of the diagnostic device · CPC title