System, method and computer-accessible medium for quantification of blur in digital images

Patent metadata

Field	Value
Publication number	US-11776235-B2
Application number	US-202017134886-A
Country	US
Kind code	B2
Filing date	Dec 28, 2020
Priority date	Aug 30, 2016
Publication date	Oct 3, 2023
Grant date	Oct 3, 2023

Abstract

The present disclosure discusses systems and methods to detect blur in digital images. The solution can be incorporated into the quality control systems of pathology and other slide scanners or can be a stand-alone solution. The solution can identify scanned images that include blur and cause the scanner to automatically rescan the blurry image. The solution can also identify regions of the scanned image that include blur. The solution can generate blur maps for each of the scanned images that identify regions of the scanned image that include blur.

First claim

What is claimed is:

1. A method, comprising:
identifying, by one or more processors, a first plurality of patches from a first digital image of a pathology slide of a tissue sample obtained from a subject, the first digital image acquired via a camera at a magnification factor of a plurality of magnification factors, each of the first plurality of patches comprising a corresponding portion of the first digital image;
determining, by the one or more processors, using a classifier, a first dwell time identifying a predicted amount of time that each patch of the first plurality of patches is to be viewed, the classifier trained using a second plurality of patches of a second digital image and a second dwell time identifying an amount of time that each patch of the second plurality of patches was viewed;
generating, by the one or more processors, a map for the first digital image based on the first dwell time for each patch of the first plurality of patches; and
storing, by the one or more processors, an association between the map and the first plurality of patches of the first digital image.

2. The method of claim 1 , further comprising providing, by the one or more processors, for display, the map with the first plurality of patches of the first digital image, the map indicating the first dwell time for each of the first plurality of patches.

3. The method of claim 1 , wherein determining the first dwell time further comprises determining, for at least one patch of the first plurality of patches, a plurality of first dwell times for a corresponding plurality of sub-patches from the at least one patch by using the classifier, each of the plurality of first dwell times identifying an estimated amount of time that a sub-patch of the corresponding plurality of sub-patches is to be viewed.

4. The method of claim 1 , further comprising categorizing, by the one or more processors, at least one patch of the first plurality of patches as a first classification instead of a second classification based on a comparison between the first dwell time for the at least one patch and a threshold.

5. The method of claim 1 , further comprising generating, by the one or more processors, an annotation for each patch of the first plurality of patches based on the corresponding first dwell time.

6. A method, comprising:
identifying, by one or more processors, a first plurality of patches from a first digital image of a pathology slide of a tissue sample obtained from a subject, the first digital image acquired via a camera at a magnification factor of a plurality of magnification factors, each of the first plurality of patches comprising a plurality of pixels from a corresponding portion of the first digital image;
determining, by the one or more processors, a first dwell time for each patch of the first plurality of patches, the first dwell time indicating a first amount of time that a corresponding patch of the first plurality of patches was viewed;
training, by the one or more processors, a classifier to determine a second dwell time identifying a predicted amount of time that each patch of the first plurality of patches is to be viewed, using (i) the first dwell time for each patch of the first plurality of patches and (ii) the plurality of pixels from the corresponding portion of the first digital image; and
using, by the one or more processors, the classifier to generate a map for a second digital image based on a second plurality of patches of the second digital image.

7. The method of claim 6 , wherein determining the first dwell time further comprises identifying, for the corresponding patch of the first plurality of patches, at least one of a gaze position or a motion of an eye of an observer viewing the corresponding patch.

8. The method of claim 6 , wherein using the classifier further comprises using the classifier to categorize each patch of the second plurality of patches as a first classification instead of a second classification based on a comparison between the second dwell time for each patch and a threshold.

9. The method of claim 6 , wherein using the classifier further comprises using the classifier to generate the map, the map identifying, for at least one patch of the first plurality of patches, a plurality of second dwell times for a corresponding plurality of sub-patches from the at least one patch by applying the classifier to each of the plurality of sub-patches.

10. A system, comprising:
one or more processors coupled with memory, configured to:
identify a first plurality of patches from a first digital image of a pathology slide of a tissue sample obtained from a subject, the first digital image acquired via a camera at a magnification factor of a plurality of magnification factors, each of the first plurality of patches comprising a corresponding portion of the first digital image;
determine, using a classifier, a first dwell time identifying a predicted amount of time that each patch of the first plurality of patches is to be viewed, the classifier trained using a second plurality of patches of a second digital image and a second dwell time identifying an amount of time that each patch of the second plurality of patches was viewed;
generate a map for the first digital image based on the first dwell time for each patch of the first plurality of patches; and
store an association between the map and the first plurality of patches of the first digital image.

11. The system of claim 10 , wherein the one or more processors are configured to provide, for display, the map with the first plurality of patches of the first digital image, the map indicating the first dwell time for each of the first plurality of patches.

12. The system of claim 10 , wherein the one or more processors are configured to determine, for at least one patch of the first plurality of patches, a plurality of first dwell times for a corresponding plurality of sub-patches from the at least one patch by using the classifier, each of the plurality of first dwell times identifying an estimated amount of time that a sub-patch of the corresponding plurality of sub-patches is to be viewed.

13. The system of claim 10 , wherein the one or more processors are configured to categorize at least one patch of the first plurality of patches as a first classification instead of a second classification based on a comparison between the first dwell time for the at least one patch and a threshold.

14. The system of claim 10 , wherein the one or more processors are configured to generate an annotation for each patch of the first plurality of patches based on the corresponding first dwell time.

Assignees

• Memorial Sloan Kettering Cancer Center

Inventors

• Campanella Gabriele
• Schüffler Peter J
• Fuchs Thomas

Classifications

• G06V10/30 Primary

  Noise filtering · CPC title

• G06F18/00

  Pattern recognition · CPC title

• G06F18/2415

  based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate · CPC title

• G06V20/695

  Preprocessing, e.g. image segmentation · CPC title