System and method for targeted object enhancement to generate synthetic breast tissue images

Patent metadata

Field	Value
Publication number	US-11445993-B2
Application number	US-201816497767-A
Country	US
Kind code	B2
Filing date	Mar 28, 2018
Priority date	Mar 30, 2017
Publication date	Sep 20, 2022
Grant date	Sep 20, 2022

Abstract

A method for processing breast tissue image data includes obtaining image data of a patient's breast tissue, processing the image data to generate a set of image slices, the image slices collectively depicting the patient's breast tissue; feeding image slices of the set through each of a plurality of object-recognizing modules, each of the object-recognizing modules being configured to recognize a respective type of object that may be present in the image slices; combining objects recognized by the respective object-recognizing modules to generate a synthesized image of the patient's breast tissue; and displaying the synthesized image.

First claim

What is claimed is:

1. A method for processing breast tissue image data, comprising:
processing image data of a patient's breast tissue to generate a set of image slices that collectively depict the patient's breast tissue, wherein at least two image slices in the set of image slices comprise at least a first object of a first object type and a second object of a second object type;
performing a plurality of object-recognition processes on the set of image slices, wherein each of the plurality of object-recognition processes is configured to recognize a respective type of object that may be present in the set of image slices;
based at least in part on the performance of the plurality of object-recognition processes, recognizing the first object of the first object type and the second object of the second object type in the at least two image slices in the set of image slices;
based on a determination that the first object of the first object type and the second object of the second object type are likely to overlap in a synthesized image, generating the synthesized image based at least on the at least two image slices in the set of image slices, wherein generating the synthesized image includes fusing the first object and the second object such that at least a portion of each of the first and the second objects are included in the synthesized image; and
causing the synthesized image to be displayed.

2. The method of claim 1 , wherein the plurality of object-recognition processes are performed on the image slices in a sequence.

3. The method of claim 2 , further comprising assigning a respective weight to each of the plurality of object-recognition processes, wherein:
the assigned weight corresponds to a significance of the type of object recognized by a particular one of the plurality of object-recognition processes; and
the respective weights assigned to the object-recognition processes determine an order of image slices in the set of image slices processed by the plurality of object-recognition processes.

4. The method of claim 3 , wherein:
the plurality of object-recognition processes comprise a first object-recognition process having a first weight and a second object-recognition process having a second weight that is higher than the first weight;
the second object of the second object type is recognized by the second object-recognition process; and
the method further comprises:
performing the first object-recognition process on the set of image slices prior to generating the synthesized image; and
recognizing a third object of a third object type based on performing the first object-recognition process on the set of image slices.

5. The method of claim 4 , further comprising determining whether the third object of the third object type is likely to overlap the second object of the second object type in the synthesized image.

6. The method of claim 5 , further comprising based on a determination that the third object of the third object type and the second object of the second object type are likely to overlap, including only the at least the portion of the second object of the second object type fused with the at least the portion of the first object of the first object type in the synthesized image.

7. The method of claim 5 , further comprising based on a determination that the third object of the third object type and the second object of the second object type are likely to overlap, emphasizing the at least the portion of the second object of the second object type relative to the third object of the third object type in the synthesized image.

8. The method of claim 1 , wherein the plurality of object-recognition processes are performed in parallel on the image slices.

9. The method of claim 8 , wherein fusing the first object with the second object such that at least the portion of each of the first and the second objects are included in the synthesized image comprises fusing the first object with the second object such that at least the portion of each of the first and the second objects are enhanced and included in the synthesized image.

10. The method of claim 9 , wherein the first object of the first object type is fused with the second object of the second object type using a linear combination technique.

11. The method of claim 9 , wherein the first object of the first object type is fused with the second object of the second object type using a non-linear combination technique.

12. The method of claim 1 , wherein a first subset of object-recognition processes in the plurality of object-recognition processes are performed sequentially on the set of image slices to recognize a first subset of object types, and a second subset of object-recognition processes in the plurality of object recognition processes are performed in parallel on the set of image slices to recognize a second subset of object types.

13. The method of claim 12 , wherein the first subset of object types includes abnormal breast tissue malignancies, and the second subset of object types include normal breast tissue structures or predetermined image patterns.

14. The method of claim 1 , further comprising displaying target object types associated with the plurality of object-recognition processes in a graphical user interface.

15. The method of claim 14 , wherein the graphical user interface provides options for an end user to select one or more target object types to be recognized and included in the synthesized image.

16. The method of claim 15 , wherein the graphical user interface provides options for allowing an end user to input an order of importance for displaying selected target object types in the synthesized image.

17. The method of claim 15 , wherein:
the graphical user interface provides options for allowing an end user to input a weight factor for each of one or more target object types; and
user input weight factors are considered when generating and displaying user-selected target object types in the synthesized image.

18. The method of claim 17 , wherein the options for allowing an end user to input a weight factor for each of one or more target object types are based at least in part on one or more of age, gender, ethnicity, race or genetic characteristics of the patient.

19. An image processing system configured to perform the method of claim 1 , wherein the system is configured to allow for adding further object-recognition processes to the plurality of the object-recognition processes in order to recognize and display further types of objects.

20. The method of claim 1 , further comprising recognizing, based at least in part on the performance of the plurality of object-recognition processes, a third object of a third object type in at least a third image slice in the set of image slices, wherein generating the synthesized image based at least on the at least two image slices in the set of image slices comprises generating the synthesized image based at least on the at least two image slices and the at least third image slice in the set of image slices.

Assignees

• Hologic Inc

Inventors

• Chui Haili
• Wei Liyang
• Ge Jun
• Gkanatsios Nikolaos

Classifications

• A61B6/5217 Primary

  extracting a diagnostic or physiological parameter from medical diagnostic data · CPC title

• A61B6/025 Primary

  Tomosynthesis · CPC title

• A61B6/5205

  involving processing of raw data to produce diagnostic data · CPC title

• G06T2207/30068

  Mammography; Breast · CPC title

• G06T2200/24

  involving graphical user interfaces [GUIs] · CPC title