Deep learning medical systems and methods for medical procedures

Patent metadata

Field	Value
Publication number	US-10242443-B2
Application number	US-201615360410-A
Country	US
Kind code	B2
Filing date	Nov 23, 2016
Priority date	Nov 23, 2016
Publication date	Mar 26, 2019
Grant date	Mar 26, 2019

Abstract

Methods and apparatus for monitoring and improving imaging system operation are provided. An example apparatus includes a first deployed deep learning network (DLN) which operates with an acquisition engine to generate an imaging device configuration. The example apparatus includes a second deployed DLN which operates with a reconstruction engine based on acquired image data. The example apparatus includes a first assessment engine with a third deployed DLN. The assessment engine receives output from at least one of the acquisition engine or the reconstruction engine to assess operation of the respective at least one of the acquisition engine or the reconstruction engine and to provide feedback to the respective at least one of the acquisition engine or the reconstruction engine. The first deployed DLN and the second deployed DLN are generated and deployed from first and second training DLNS, respectively.

First claim

What is claimed is:

1. An apparatus comprising:
a first deployed deep learning network associated with an acquisition engine, the acquisition engine associated with an imaging device, the first deployed deep learning network configured to operate with the acquisition engine to generate a configuration for the imaging device, the first deployed deep learning network generated and deployed from a first training deep learning network;
a second deployed deep learning network associated with a reconstruction engine, the reconstruction engine to receive acquired image data from the imaging device via the acquisition engine and to reconstruct an image from the acquired image data, the second deployed deep learning network to operate with the reconstruction engine based on the acquired image data, the second deployed deep learning network generated and deployed from a second training deep learning network;
a first assessment engine with a third deployed deep learning network, the assessment engine to receive output from at least one of the acquisition engine or the reconstruction engine to assess operation of the respective at least one of the acquisition engine or the reconstruction engine and to provide feedback to the respective at least one of the acquisition engine or the reconstruction engine, the third deployed deep learning network generated and deployed from a third training deep learning network.

2. The apparatus of claim 1 further including:
a fourth deployed deep learning network associated with a diagnosis engine, the diagnosis engine to facilitate diagnosis using the reconstructed image from the reconstruction engine, the fourth deployed deep learning network to operate with the diagnosis engine, the fourth deployed deep learning network generated and deployed from a fourth training deep learning network; and
a second assessment engine with a fifth deployed deep learning network, the assessment engine to receive output from at least one of the acquisition engine, the reconstruction engine, or the diagnosis engine to assess operation of the respective at least one of the acquisition engine, the reconstruction engine, or the diagnosis engine and to provide feedback to the respective at least one of the acquisition engine, the reconstruction engine, or the diagnosis engine.

3. The apparatus of claim 2 , wherein the acquisition engine, the reconstruction engine, and the diagnosis engine exchange feedback to generate an indication of system health.

4. The apparatus of claim 2 , wherein the reconstruction engine is to generate the reconstructed image for human viewing and to process the acquired image data for computer analysis of the image data by the diagnosis engine.

5. The apparatus of claim 1 , wherein the first deployed deep learning network is to generate a configuration for the imaging device based on the acquisition engine, the imaging device, and a patient variable associated with the patient to be imaged.

6. The apparatus of claim 1 , wherein at least one of the first deployed deep learning network, the second deployed deep learning network, or the third deployed deep learning network includes a convolutional neural network.

7. The apparatus of claim 1 , wherein at least one of the first training deep learning network, second training deep learning network, or third training deep learning network is provided with one or more features of interest in training of the corresponding first, second, or third deployed deep learning network.

8. A method comprising:
generating a configuration for an imaging device for image acquisition via a first deployed deep learning network associated with an acquisition engine associated with the imaging device, the first deployed deep learning network generated and deployed from a first training deep learning network;
monitoring, using a second deployed deep learning network, image reconstruction by a reconstruction engine of image data acquired by the imaging device via the acquisition engine, the second deployed deep learning network associated with the reconstruction engine and to operate with the reconstruction engine based on the acquired image data, the second deployed deep learning network generated and deployed from a second training deep learning network;
assessing, by an assessment engine using a third deployed deep learning network, operation of respective at least one of the acquisition engine or the reconstruction engine based on output received from the respective at least one of the acquisition engine or the reconstruction engine, the third deployed deep learning network generated and deployed from a third training deep learning network; and
providing, by the assessment engine using the third deployed deep learning network, feedback to the respective at least one of the acquisition engine or the reconstruction engine.

9. The method of claim 8 , further including:
facilitating, using a fourth deployed deep learning network, diagnosis using the reconstructed image from the reconstruction engine, the fourth deployed deep learning network to operate with a diagnosis engine, the fourth deployed deep learning network generated and deployed from a fourth training deep learning network; and
assessing operation of respective at least one of the acquisition engine, the reconstruction engine, or the diagnosis engine based on output received from the respective at least one of the acquisition engine, the reconstruction engine, or the diagnosis engine to provide feedback to the respective at least one of the acquisition engine, the reconstruction engine, or the diagnosis engine.

10. The method of claim 9 , further including generating an indication of system health based on an exchange of feedback among the acquisition engine, the reconstruction engine, and the diagnosis engine.

11. The method of claim 9 , wherein the reconstruction engine is configured to generate the reconstructed image for human viewing and to process the acquired image data for computer analysis of the image data by the diagnosis engine.

12. The method of claim 8 , wherein the first deployed deep learning network is to generate a configuration for the imaging device based on the acquisition engine, the imaging device, and a patient variable associated with the patient to be imaged.

13. The method of claim 8 , wherein at least one of the first deployed deep learning network, the second deployed deep learning network, or the third deployed deep learning network includes a convolutional neural network.

14. The method of claim 8 , wherein at least one of the first training deep learning network, second training deep learning network, or third training deep learning network is provided with one or more features of interest in training of the corresponding first, second, or third deployed deep learning network.

15. A non-transitory computer readable medium comprising instructions which, when executed, cause a machine to at least:
generate a configuration for an imaging device for image acquisition via a first deployed deep learning network associated with an acquisition engine associated with the imaging device, the first deployed deep learning network generated and deployed from a first training deep learning network;
monitor, using a second deployed deep learning network, image reconstruction by a reconstruction engine of image data acquired by the imaging device via the acquisition engine, the second deployed deep learning network associated with the reconstruction engine and to operate with the reconstruction engine based on the acquired image data, the second deployed deep learning network generated and dep

Assignees

• Gen Electric

Inventors

• Hsieh Jiang
• Avinash Gopal
• Sirohey Saad

Classifications

• G16H30/40 Primary

  for processing medical images, e.g. editing · CPC title

• G16H40/40

  for the management of medical equipment or devices, e.g. scheduling maintenance or upgrades · CPC title

• G06T2207/10004

  Still image; Photographic image · CPC title

• G06T7/0012 Primary

  Biomedical image inspection · CPC title

• G06T2207/20084

  Artificial neural networks [ANN] · CPC title