Method and apparatus for aiding in the diagnosis of otitis media by classifying tympanic membrane images

Patent metadata

Field	Value
Publication number	US-9636007-B2
Application number	US-201314418509-A
Country	US
Kind code	B2
Filing date	Jun 11, 2013
Priority date	Aug 3, 2012
Publication date	May 2, 2017
Grant date	May 2, 2017

Abstract

A method of aiding the diagnosis of otitis media in a patient includes obtaining image data in a processor apparatus of a computing device, the image data being associated with at least one electronic image of a tympanic membrane of the patient, calculating a plurality of image features, each image feature being calculated based on at least a portion of the image data, classifying the at least one electronic image as a particular type of otitis media using the plurality of image features, and outputting an indication of the particular type of otitis media. Also, a system for implementing such a method that includes an output device and a computing device.

First claim

What is claimed is:

1. A method of aiding a diagnosis of otitis media in a patient, comprising:
obtaining image data in a processor apparatus of a computing device, the image data being associated with at least one electronic image of a tympanic membrane of the patient;
calculating a plurality of image features, each image feature being calculated based on at least a portion of the image data, wherein the plurality of image features includes: (i) a concavity feature which indicates a degree of concavity of a region located centrally in the tympanic membrane, (ii) a translucency feature which indicates a degree of translucency of the tympanic membrane, (iii) an amber level feature which indicates a degree of amber color present in the tympanic membrane, (iv) a grayscale variance feature which indicates a degree of variance of intensities across a grayscale version of the at least one electronic image, (v) a bubble presence feature which indicates a degree to which bubbles are present in the tympanic membrane, and (vi) a light feature which indicates a degree of non-uniformity of illumination in the at least one electronic image;
classifying the at least one electronic image into one of a plurality of predetermined clinical diagnostic categories for otitis media using the plurality of image features and a predetermined decision process; and
outputting an indication of the one of the plurality of predetermined clinical diagnostic categories.

2. The method according to claim 1 , wherein calculating the concavity feature comprises identifying a central region in a grayscale version of the at least one electronic image, defining a bright region in the central region and a dark region in the central region and comparing the bright region to the dark region.

3. The method according to claim 1 , wherein calculating the translucency feature comprises measuring a grayness of the tympanic membrane using a color-assignment technique.

4. The method according to claim 1 , wherein calculating the amber level feature comprises employing a color-assignment technique.

5. The method according to claim 1 , wherein calculating the grayscale variance feature comprises measuring a variance of pixel intensities in the grayscale version of the at least one electronic image.

6. The method according to claim 1 , wherein calculating the bubble presence feature comprises obtaining red and green channels of the at least one electronic image, performing edge detection thereon to create a binary image X b in between edges, and defining as the bubble presence feature the mean of X b .

7. The method according to claim 1 , wherein calculating the light feature comprises calculating a ratio of brightly-lit to the darkly-lit regions of the at least one electronic image.

8. The method according to claim 1 , wherein the obtaining image data comprises receiving unprocessed image data and preprocessing the unprocessed image data to produce the image data.

9. The method according to claim 8 , wherein the preprocessing employs an automated segmentation process to locate the tympanic membrane in the at least one electronic image.

10. The method according to claim 1 , wherein the classifying includes a first level based on the concavity feature, translucency feature and the light feature to classify the at least one electronic image into one of two superclasses: AOM/OME and NOE/OME, and a second level to classify the at least one electronic image into one of AOM, OME and NOE based on a weighted combination of the amber level feature, bubble presence feature, the translucency feature and grayscale variance feature.

11. The method according to claim 1 , further comprising capturing the at least one electronic image.

12. A system for aiding a diagnosis of otitis media in a patient, comprising:
an output device; and
a computing device having a processor apparatus structured and configured to
obtain image data, the image data being associated with at least one electronic image of a tympanic membrane of the patient;
calculate a plurality of image features, each image feature being calculated based on at least a portion of the image data, wherein the plurality of image features includes: (i) a concavity feature which indicates a degree of concavity of a region located centrally in the tympanic membrane, (ii) a translucency feature which indicates a degree of translucency of the tympanic membrane, (iii) an amber level feature which indicates a degree of amber color present in the tympanic membrane, (iv) a grayscale variance feature which indicates a degree of variance of intensities across a grayscale version of the at least one electronic image, (v) a bubble presence feature which indicates a degree to which bubbles are present in the tympanic membrane, and (vi) a light feature which indicates a degree of non-uniformity of illumination in the at least one electronic image;
classify the at least one electronic image into one of a plurality of predetermined clinical diagnostic categories for otitis media using the plurality of image features and a predetermined decision process; and
cause the output device to output an indication of the one of the plurality of predetermined clinical diagnostic categories.

13. The system according to claim 12 , wherein the output device is a display.

14. The system according to claim 12 , further comprising an image capture device structured to capture the at least one electronic image.

15. The system according to claim 12 , wherein the computing device is structured and configured to calculate the concavity feature by identifying a central region in a grayscale version of the at least one electronic image, defining a bright region in the central region and a dark region in the central region and comparing the bright region to the dark region.

16. The system according to claim 12 , wherein the computing device is structured and configured to calculate the translucency feature by measuring a grayness of the tympanic membrane using a color-assignment technique.

17. The system according to claim 12 , wherein the computing device is structured and configured to calculate the amber level feature by employing a color-assignment technique.

18. The system according to claim 12 , wherein the computing device is structured and configured to calculate the grayscale variance feature by measuring a variance of pixel intensities in the grayscale version of the at least one electronic image.

19. The system according to claim 12 , wherein the computing device is structured and configured to calculate the bubble presence feature by obtaining red and green channels of the at least one electronic image, performing edge detection thereon to create a binary image X b in between edges, and defining as the bubble presence feature the mean of X b .

20. The system according to claim 12 , wherein the computing device is structured and configured to calculate the light feature by calculating a ratio of brightly-lit to the darkly-lit regions of the at least one electronic image.

21. The system according to claim 12 , wherein the computing device is structured and configured to classify the at least one electronic image using a first level based on the concavity feature, translucency feature and the light feature to classify the at least one electronic image into one of two superclasses: AOM/OME and NOE/OME, and a second level based on a weighted combination of the amber level feature, bubble presence feature, the translucency feature and grayscale

Assignees

• Univ Of Pittsburgh—Of The Commonwealth System Of Higher Education
• Univ Carnegie Mellon

Inventors

• Hoberman Alejandro
• Kovacevic Jelena
• Shaikh Nader
• Kuruvilla Anupama

Classifications

• A61B2576/02

  specially adapted for a particular organ or body part · CPC title

• A61B5/7425

  Displaying combinations of multiple images regardless of image source, e.g. displaying a reference anatomical image with a live image · CPC title

• G06T7/0012

  Biomedical image inspection · CPC title

• G16H30/20

  for handling medical images, e.g. DICOM, HL7 or PACS · CPC title

• G06T7/40

  Analysis of texture (depth or shape recovery from texture G06T7/529) · CPC title