Granted Patent B2

US 12,400,752 · App. 17/423,179 · Granted Aug 26, 2025

Inventors: Tobias Wissel (Lübeck, DE); Irina Waechter-Stehle (Hamburg, DE); Frank Michael Weber (Hamburg, DE); Arne Ewald (Hamburg, DE)

Assignee: KONINKLIJKE PHILIPS N.V.

G16H30/40G06N20/00G06T7/0014G06T7/11G06T7/136G06T7/143G06T7/174G06T2207/10132

Abstract

Claims (51)

1. An image analysis method for determining failures in an image segmentation procedure that identifies and segments one or more anatomical features in an image, the method comprising:

obtaining different image segmentations of two or more images of an anatomical region of a patient, representative of the anatomical region at different respective time points, wherein each of the two or more images is segmented using the image segmentation procedure;

determining one or more measures of consistency between the different image segmentations based on a comparison of the different image segmentations; and

determining a measure of accuracy of the image segmentation procedure based on the one or more measures of consistency and on reference to a pre-determined normal degree of inconsistency between the different image segmentations of the two or more images,

wherein the one or more measures of consistency are distinct from the measure of accuracy,

wherein the pre-determined normal degree of inconsistency is representative of expected variations or movement of the anatomical region at the different respective time points, and

wherein the different respective time points correspond to defined time points in a periodic or cyclic movement pattern of the anatomical region.

2. The method as claimed in claim 1 , wherein determining the one or more measures of consistency comprises comparing one or more similar segmented features, segmented in each of the different image segmentations, and determining the one or more measures of consistency between the one or more similar segmented features in the different image segmentations.

3. The method as claimed in claim 2 , wherein determining the one or more measures of consistency comprises comparing at least one of a position, a shape, or one or more dimensions of at least a portion of the different image segmentations.

4. The method as claimed in claim 1 , wherein the measure of accuracy is further based on applying one or more pre-determined minimum thresholds for the one or more measures of consistency.

5. The method as claimed in claim 1 , wherein at least one of determining the one or more measures of consistency or determining the measure of accuracy comprises use of a machine learning algorithm.

6. The method as claimed in claim 5 , wherein determining the measure of accuracy comprises use of a machine learning algorithm, and

wherein the machine learning algorithm is trained for distinguishing between normal and abnormal levels of inconsistency between image segmentations of the anatomical region at different time points.

7. The method as claimed in claim 5 , wherein both determining the one or more measures of consistency and determining the measure of accuracy comprise use of a machine learning algorithm.

8. The method as claimed in claim 1 , wherein the measure of accuracy comprises a probabilistic measure of accuracy.

9. The method as claimed in claim 1 , wherein the method is performed in real time with obtaining of the different image segmentations.

10. The method as claimed in claim 1 , wherein the method is an ultrasound image analysis method such that the two or more images of the anatomical region are ultrasound images.

11. The method as claimed in claim 1 , wherein the obtaining of the different image segmentations comprises:

receiving the two or more images of the anatomical region; and

processing the two or more images with the image segmentation procedure.

12. The method as claimed in claim 1 , wherein obtaining the one or more image segmentations comprises:

receiving ultrasound data;

processing the ultrasound data to derive the two or more images of the anatomical region, representative of the anatomical region at different respective time points; and

processing the derived two or more images with the image segmentation procedure.

13. The method as claimed in claim 1 , wherein the measure the accuracy is configured to indicate whether a variation in the one or more measures of consistency is representative of:

the failures in the image segmentation procedure; or

natural movement of the anatomical region.

14. The method as claimed in claim 1 , further comprising outputting a first representation of the measure of accuracy to a display.

15. The method as claimed in claim 14 , further comprising outputting a second representation of a recommendation for changing acquisition of images of the anatomical region.

16. The method as claimed in claim 1 ,

wherein, when the measure of accuracy for the different image segmentations is unacceptable, at least one of the two or more images of the anatomical region are not output to a display, and

wherein, when the measure of accuracy for the different image segmentations is acceptable, the two or more images of the anatomical region are output to a display.

17. An image analysis device for determining failures in an image segmentation procedure that identifies and segments one or more anatomical features in an image, the image analysis device comprising:

an acquisition module configured to obtain different image segmentations of two or more images of an anatomical region of a patient, representative of the anatomical region at different respective time points, wherein each of the two or more images is segmented using the image segmentation procedure, wherein the different respective time points correspond to defined time points in a periodic or cyclic movement pattern of the anatomical region;

a comparison module configured to determine a measure of consistency between the different image segmentations based on a comparison of the different image segmentations; and

an analysis module configured to determine a measure of accuracy of the image segmentation procedure based on the measure of consistency and on reference to a pre-determined normal degree of inconsistency between the different image segmentations of the two or more images,

wherein the measure of consistency is distinct from the measure of accuracy, and

wherein the pre-determined normal degree of inconsistency is representative of expected variations or movement of the anatomical region at the different respective time points.

18. An ultrasound system, comprising:

an ultrasound transducer unit; and

the image analysis device as claimed in claim 17 , operatively coupled with the ultrasound transducer unit for use in obtaining different image segmentations.

19. An image analysis device for determining failures in an image segmentation procedure that identifies and segments one or more anatomical features in an image, the image analysis device comprising:

one or more processors; and

one or more non-transitory storage media storing instructions that, when executed by the one or more processors, cause the one or more processors to:

receive two or more images of an anatomical region of a patient representative of the anatomical region at different respective time points;

perform image segmentation of the two or more images using the image segmentation procedure to provide different image segmentations of the two or more images, wherein the different respective time points correspond to defined time points in a periodic or cyclic movement pattern of the anatomical region;

determine a measure of consistency between the different image segmentations based on a comparison of the different image segmentations; and

determine a measure of accuracy of the image segmentation procedure based on the measure of consistency and on reference to a pre-determined normal degree of inconsistency between the different image segmentations of the two or more images,

wherein the measure of consistency is distinct from the measure of accuracy, and

wherein the pre-determined normal degree of inconsistency is representative of expected variations or movement of the anatomical region at the different respective time points.

20. The image analysis device as claimed in claim 19 , wherein determining the measure of consistency comprises comparing one or more similar segmented features, segmented in each of the different image segmentations, and determining the measure of consistency between the one or more similar segmented features in the different image segmentations.

Assignments (1)

Priority Claims (1)

Continuity (1)

References Cited (19)