Digital pathology system with low-latency analytics

• US 8,774,518 B2
• Filed: 08/01/2012
• Issued: 07/08/2014
• Est. Priority Date: 08/02/2011
• Status: Active Grant

• Current Assignee: NEC Corporation
• Original Assignee: NEC Laboratories America Inc (NEC Corporation)
• Inventors: Cosatto, Eric, Malon, Christopher, Graf, Hans P.
• Primary Examiner(s): Desire, Gregory M
• Application Number:
  US13/564,437
• Publication Number:
  US 20130034304A1
• Time in Patent Office: 706 Days
• Field of Search: 382/128, 382/133, 382/173, 382/190
• US Class Current: 382/190
• CPC Class Codes: G06V 10/95 structured as a network, e.g. client-server architectures

Abstract

Methods and systems for digital pathology with low-latency analytics include determining potential regions of interest within an image in accordance with one or more high-priority analyses, dividing the potential regions of interest into a plurality of sub-sections optimized for parallel computation, analyzing the sub-sections using one or more execution nodes, each including one or more processors, using a copy of the image stored in a shared memory according to the one or more high-priority analyses, and storing an intermediate analysis result based on analysis results from the one or more execution nodes in a shared memory.

Claims

1. A method for preprocessing images for analysis, the method comprising:

determining potential regions of interest within an image in accordance with one or more high-priority analyses;

dividing the potential regions of interest into a plurality of sub-sections optimized for parallel computation;

analyzing the sub-sections using one or more execution nodes, each including one or more processors, using a copy of the image stored in a shared memory according to the one or more high-priority analyses;

storing an intermediate analysis result based on analysis results from the one or more execution nodes in a shared memory;

receiving a selection of a region of the image and a request for analysis of the selection from an interface layer;

determining what further analysis beyond the intermediate results is needed to meet the request;

analyzing further sub-sections of the image in accordance with the determination of further needed analysis at the one or more execution nodes;

combining the intermediate results with the results of the analysis of said further sub-sections to produce final results; and

transferring said final results to the interface layer.

2. The method of claim 1, wherein results are progressively transferred to the interface layer to provide a user with real-time responses.

3. The method of claim 1, wherein potential regions of interest are determined using a trained classifier that is trained with training images and selections from said training images associated with particular forms of analysis.

4. The method of claim 1, further comprising transferring information regarding the intermediate results to an interface layer.

5. The method of claim 4, wherein said information regarding the intermediate results includes the determined potential regions of interest which underwent analysis for display at the interface layer in conjunction with the image.

6. The method of claim 1, wherein dividing the potential regions of interest comprises:

determining sub-section size constraints;

forming an initial rectangular grid; and

iteratively conforming grid elements to each potential region of interest.

7. The method of claim 6, wherein determining sub-section size constraints comprises determining a minimum width and height based on the request for analysis.

8. The method of claim 6, wherein determining sub-section size constraints comprises determining a maximum width and height based on available system resources.

9. The method of claim 6, wherein iteratively conforming grid elements to the selection includes resizing grid elements in accordance with the determined size constraints.

10. A non-transitory computer readable storage medium comprising a computer readable program, wherein the computer readable program when executed on a computer causes the computer to perform:

determining potential regions of interest within an image in accordance with one or more high-priority analyses;

dividing the potential regions of interest into a plurality of sub-sections optimized for parallel computation;

analyzing the sub-sections using one or more execution nodes, each including one or more processors, using a copy of the image stored in a shared memory according to the one or more high-priority analyses;

storing an intermediate analysis result based on analysis results from the one or more execution nodes in a shared memory;

receiving a selection of a region of the image and a request for analysis of the selection from an interface layer;

determining what further analysis beyond the intermediate results is needed to meet the request;

analyzing further sub-sections of the image in accordance with the determination of further needed analysis at the one or more execution nodes;

combining the intermediate results with the results of the analysis of said further sub-sections to produce final results; and

transferring said final results to the interface layer.

11. A system for preprocessing images for analysis, the system comprising:

an interpretation layer configured to pre-process images to provide intermediate results that can be quickly provided upon a user'"'"'s request, said interpretation layer comprising;

a shared memory configured to store full images and analysis results;

an execution controller configured to determine potential regions of interest within an image in accordance with one or more high-priority analyses, to divide the potential regions of interest into a plurality of sub-sections optimized for parallel computation; and

one or more execution nodes, each including one or more processors, configured to analyze the sub-sections using a copy of the image stored in the shared memory according to the one or more high-priority analyses,

wherein the results from the one or more execution nodes are combined to form intermediate execution nodes stored in the shared memory, and

wherein the interpretation layer is further configured to receive a selection of a region of the image and a request for analysis of the selection from an interface layer, wherein the execution controller is further configured to determine what further analysis beyond the intermediate results is needed to meet the request, and wherein the one or more execution nodes are further configured to analyze further sub-sections of the image in accordance with the determination of further needed analysis at the one or more execution nodes.

12. The system of claim 11, wherein the execution controller is further configured to combine the intermediate results with the results of the analysis of said further sub-sections to produce final results and wherein the interpretation layer is further configured to transferring said final results to the interface layer.

13. The system of claim 11, wherein the interpretation layer is further configured to transfer information regarding the intermediate results to an interface layer.

14. The system of claim 13, wherein said information regarding the intermediate results includes the determined potential regions of interest which underwent analysis for display at the interface layer in conjunction with the image.

15. The system of claim 11, wherein the execution controller is further configured to determine sub-section size constraints, form an initial rectangular grid, and iteratively conform grid elements to the selection to divide the selected region of the image.

16. The system of claim 15, wherein the execution controller is further configured to determine a minimum width and height based on the request for analysis to determine sub-section size constraints.

17. The system of claim 15, wherein the execution controller is further configured to determine a maximum width and height based on available system resources to determine sub-section size constraints.

18. The system of claim 15, wherein the wherein the execution controller is further configured to resize grid elements in accordance with the determined size constraints to conform grid elements to the selection.