Automatic classification of eardrum shape

Patent metadata

Field	Value
Publication number	US-10275644-B2
Application number	US-201715453394-A
Country	US
Kind code	B2
Filing date	Mar 8, 2017
Priority date	Mar 8, 2017
Publication date	Apr 30, 2019
Grant date	Apr 30, 2019

Abstract

Eardrums are automatically classified based on a feature set from a three-dimensional image of the eardrum, such as might be derived from a plenoptic image captured by a light field otoscope. In one aspect, a grid is overlaid onto the three-dimensional image of the eardrum. The grid partitions the three-dimensional image into cells. One or more descriptors are calculated for each of the cells, and the feature set includes the calculated descriptors. Examples of descriptors include various quantities relating to the depth and/or curvature of the eardrum. In another aspect, isocontour lines (of constant depth) are calculated for the three-dimensional image of the eardrum. One or more descriptors are calculated for the isocontour lines, and the feature set includes the calculated descriptors. Examples of descriptors include various quantities characterizing the isocontour lines.

First claim

What is claimed is:

1. A method for automatic classification of eardrum shape from a three-dimensional image of an eardrum, the method performed by a computer system having a processor and memory, the method comprising:
calculating a feature set for the eardrum based on the three-dimensional image of the eardrum, wherein calculating the feature set comprises:
overlaying a two-dimensional grid onto the three-dimensional image of the eardrum by extending the two-dimensional grid along a depth dimension, thereby partitioning the three-dimensional image into cells; and
calculating one or more descriptors for each of the cells, the feature set comprising the calculated descriptors; and

classifying a three-dimensional shape of the eardrum based on the feature set.

2. The method of claim 1 wherein the two-dimensional grid is one of an elliptical grid, a circular grid with an unregistered center and a circular grid with a registered center.

3. The method of claim 1 wherein at least one of the descriptors is based on the eardrum depth values within each cell.

4. The method of claim 1 wherein at least one of the descriptors is based on the eardrum curvatures within each cell.

5. A method for automatic classification of eardrum shape from a three-dimensional image of an eardrum, the method performed by a computer system having a processor and memory, the method comprising:
calculating a feature set for the eardrum based on the three-dimensional image of the eardrum, wherein calculating the feature set comprises:
calculating isocontour lines for the three-dimensional image of the eardrum; and
calculating one or more descriptors for the isocontour lines, the feature set comprising the calculated descriptors; and

classifying a three-dimensional shape of the eardrum based on the feature set.

6. The method of claim 5 wherein the descriptors are not calculated for isocontour lines that have a length below a threshold.

7. The method of claim 5 further comprising: for isocontour lines that are open and end at a boundary of the three-dimensional image, closing the isocontour line.

8. The method of claim 5 wherein the descriptors are calculated only for closed isocontour lines that enclose a center of the eardrum.

9. The method of claim 5 wherein at least one of the descriptors is based on the number of isocontour lines.

10. The method of claim 5 wherein at least one of the descriptors is based on a depth of the isocontour lines.

11. The method of claim 5 wherein at least one of the descriptors is based on a length of the isocontour lines, on an area enclosed by the isocontour lines, or on a shape of the isocontour lines.

12. The method of claim 1 wherein at least some of the descriptors are calculated based on values for the three-dimensional image of the eardrum, and at least one of the descriptors is calculated based on a statistical average of values.

13. The method of claim 1 wherein at least some of the descriptors are calculated based on values for the three-dimensional image of the eardrum, and at least one of the descriptors comprises a histogram of values.

14. The method of claim 1 wherein classifying the three-dimensional shape of the eardrum distinguishes between a bulging eardrum and a not bulging eardrum.

15. The method of claim 1 wherein classifying the three-dimensional shape of the eardrum distinguishes between a retracted eardrum and a not retracted eardrum.

16. The method of claim 1 wherein classifying the three-dimensional shape of the eardrum distinguishes between acute otitis media (AOM), otitis media with effusion (OME) and no effusion (NOE).

17. The method of claim 1 wherein the three-dimensional image of the eardrum is a registered three-dimensional image of the eardrum, wherein registration compensates for out-of-plane rotation, in-plane rotation and center localization.

18. A non-transitory computer-readable storage medium storing executable computer program instructions for automatic classification of eardrum shape from a three-dimensional image of an eardrum, the instructions executable by a processor and causing the processor to perform a method comprising:
calculating a feature set for the eardrum based on the three-dimensional image of the eardrum, wherein calculating the feature set comprises:
overlaying a two-dimensional grid onto the three-dimensional image of the eardrum by extending the two-dimensional grid along a depth dimension, thereby partitioning the three-dimensional image into cells; and
calculating one or more descriptors for each of the cells, the feature set comprising the calculated descriptors; and

classifying a three-dimensional shape of the eardrum based on the feature set.

19. The method of claim 5 wherein classifying the three-dimensional shape of the eardrum distinguishes between a bulging eardrum and a not bulging eardrum.

20. The method of claim 5 wherein classifying the three-dimensional shape of the eardrum distinguishes between a retracted eardrum and a not retracted eardrum.

Assignees

• Spinoulas Leonidas
• Karygianni Sofia
• Martinello Manuel
• Frossard Pascal
• Tosic Ivana
• Ricoh Co Ltd

Inventors

• Spinoulas Leonidas
• Karygianni Sofia
• Martinello Manuel
• Frossard Pascal
• Tosic Ivana

Classifications

• G06V40/172

  Classification, e.g. identification · CPC title

• G06V40/168 Primary

  Feature extraction; Face representation · CPC title

• G06V10/50

  by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis · CPC title

• G06K9/00

  Physics · mapped topic

• G06K9/00288 Primary

  Physics · mapped topic