Systems and methods for automatic detection and testing of images for clinical relevance

Patent metadata

Field	Value
Publication number	US-9788786-B2
Application number	US-201114112369-A
Country	US
Kind code	B2
Filing date	May 4, 2011
Priority date	May 4, 2011
Publication date	Oct 17, 2017
Grant date	Oct 17, 2017

Abstract

Disclosed herein are systems and methods for automatic detection of clinical relevance of images of an anatomical situation. The method includes comparing a first image and a second image and determining whether a difference between the first and second images is at least one of a local type difference and a global type difference. The local type difference is a local difference of the first image and the second image and the global type difference is a global difference between the first image and the second image. The second image is determined as having a clinical relevance if it is determined that the difference between the first image and the second image comprises a local type difference.

First claim

The invention claimed is:

1. A method for automatic detection on clinical relevance of images of an anatomical situation provided by an imaging device, the method comprising:
taking a first image of the anatomical situation at a first time;
taking a second image of the anatomical situation at a second later time;
comparing the first image and the second image and determining a difference between the first image and the second image by separating each of the first image and the second image into a plurality of corresponding image blocks each comprising a plurality of pixels, and comparing at least one of the corresponding image blocks of the first and second image;
determining whether the difference between the first image and the second image comprises at least one of a local type difference and a global type difference, wherein the local type difference is a local difference of the first image and the second image and wherein the global type difference is a global difference between the first image and the second image, and
determining whether the second image has clinical relevance by determining whether the difference between the first image and the second image comprises a local type difference,
wherein the local type difference is detected if a first difference of a single image block of the first image and a corresponding single image block of the second image exceeds a first predetermined threshold and wherein the global type difference is detected if a second difference of all of the blocks of the plurality of blocks of the first image and all the blocks of the second image exceeds a second predetermined threshold, wherein the second threshold is larger than the first threshold.

2. The method according to claim 1 , wherein the first and second differences are calculated using the same metric.

3. The method according to claim 2 , wherein at least one of the threshold and the first threshold is a locally varying threshold.

4. The method according to claim 2 , wherein at least one of the threshold, the first threshold and the second threshold is a dynamic threshold.

5. The method according to claim 1 , further comprising determining the image block having the maximum local difference and comparing the image block having the maximum local difference with remaining image blocks and determining a local type difference based on the comparison of the image block having the maximum local difference and remaining image blocks.

6. The method according to claim 1 , further comprising determining the image block having a significantly increased local difference and comparing the image block having the significantly increased local difference with image blocks in the vicinity of the image block having a significantly increased local difference and determining a local type difference based on the comparison of the image block having the significantly increased local difference with image blocks in the vicinity of the image block having a significantly increased local difference.

7. The method according to claim 1 , further comprising recognizing relevant objects in the first image and determining a local type difference based on a difference of an image block of the first image including the recognized relevant object and a corresponding image block of the second image.

8. The method according to claim 1 , wherein comparing and determining a difference of the first image and the second image includes determining an average and a variance of at least a part of corresponding image blocks of the first image and the second image and comparing corresponding image blocks with respect to average and variance.

9. The method according to claim 1 , wherein a unique size for all image blocks is dynamically adapted based on a detected size of identified objects of the images.

10. The method according to claim 1 , wherein comparing and determining a difference of the first image and the second image includes combining a plurality of image blocks to an image block cluster and comparing corresponding image block clusters.

11. The method according to claim 10 , wherein the local type difference is determined if at least one of a difference between the image block of the first image and the corresponding image block of the second image and a difference between an image block cluster of the first image and a corresponding image block cluster of the second image exceeds the first threshold.

12. The method according to claim 1 , further comprising detecting a relevant image range, the image range comprising imaged objects, wherein comparing and determining a difference between the first image and the second image is exclusively carried out based on the detected relevant image range.

13. The method according to claim 12 , wherein the detected relevant image range is used as base for a following image.

14. The method according to claim 1 , further comprising detecting a predetermined characteristic of at least one of the first image and second image, wherein the predetermined characteristic is an indicative for at least one of an imaging device type and imaging device manufacturer.

15. The method according to claim 1 , wherein the first image and the second image are generated from a permanent video signal output of an imaging device.

16. The method as claimed in claim 1 wherein the size of the image blocks may be dynamically varied based on the size of objects in the image.

17. The method according to claim 1 , further comprising recognizing relevant objects inside a relevant image range of the first image and determining a local type difference based on a difference of an image block of the first image including the recognized relevant object and a corresponding image block of the second image.

18. A device for automatic detection on clinical relevance of images of an anatomical situation provided by an imaging device, the device comprising:
an image input interface;
a first storage unit for a first digital x-ray image taken at a first time of the anatomical situation taken from the image input interface;
a second storage unit for a second digital x-ray image taken at a second later time of the anatomical situation taken from the image input interface;
a comparator unit for comparing and determining a difference between the first image and the second image, the comparator unit separating each of the first image and the second image into corresponding image blocks each comprising a plurality of pixels, and comparing at least one of the corresponding image blocks of the first and second image;
a difference type evaluation unit for evaluating the difference type of the first image and the second image, wherein the difference type evaluation unit is adapted for determining at least one of a local type difference and a global type difference, wherein the local type difference is a local difference of the first image and the second image and wherein the global type difference is a global difference between the first image and the second image, and wherein the local type difference is detected if a first difference of an image block of the first image and a corresponding image block of the second image exceeds a first predetermined threshold and wherein the global type difference is detected if a second difference of the first image and the second image exceeds a second predetermined threshold, wherein the second threshold is larger than the first threshold; and
an image selection unit for selecting the second image as having a clinical relevance if it is determined that the difference between

Assignees

• Pfrengle Udo
• Kohnen Michael
• Blau Arno
• Epting Thomas
• Stryker European Holdings I Llc

Inventors

• Pfrengle Udo
• Kohnen Michael
• Blau Arno
• Epting Thomas

Classifications

• G06T7/254

  involving subtraction of images · CPC title

• G06T2207/20021

  Dividing image into blocks, subimages or windows · CPC title

• G06T7/0012 Primary

  Biomedical image inspection · CPC title

• G06T2207/20224

  Image subtraction · CPC title

• G06T2207/20004

  Adaptive image processing · CPC title