Method and apparatus for expanding the use of existing computer-aided detection code

• US 7,477,766 B1
• Filed: 03/13/2006
• Issued: 01/13/2009
• Est. Priority Date: 11/21/2001
• Status: Active Grant

• Current Assignee: Hologic Incorporated
• Original Assignee: Hologic Incorporated
• Inventors: Hasegawa, Akira, Zhang, Xiaolan, Chui, Haili, Roehrig, Jimmy, Pourjavid, Sussan
• Primary Examiner(s): AZARIAN, SEYED H
• Application Number:
  US11/375,379
• Time in Patent Office: 1,037 Days
• Field of Search: 382128-133, 382/162, 382/168, 382/169, 382/194, 382/196, 382/203, 382/209, 382/232, 382/234, 382/237, 382/240, 382/255, 382/274, 382/277, 382/291, 382/305, 382/100, 382/134, 382/155, 382/224, 382/276, 382/294, 382/312, 600/408, 600/409, 378/28, 378/37, 378/4, 378/21, 707/102, 356/443, 345/424, 345/427
• US Class Current: 382/128
• CPC Class Codes: G06T 2207/20081 Training; Learning
• CPC Class Codes: G06T 2207/30068 Mammography; Breast
• CPC Class Codes: G06T 7/0012 Biomedical image inspection

Abstract

A method and apparatus for analyzing a medical image obtained from one of a plurality of digital modalities, the method comprising transforming or mapping the initial medical image to create a uniform contrast response and appearance regardless of the original modality of the image.

Claims

1. A computer program product embodied in a computer-readable medium for analyzing a first medical image to detect anatomical abnormalities therein, the first medical image acquired from a first type of imaging system having a first response characteristic, comprising:

computer-aided detection (CAD) code comprising at least one neural network-based automated anatomical abnormality detection algorithm trained using training images acquired from a second type of imaging system having a second response characteristic different than the first response characteristic; and

transforming code based on said first and second response characteristics for transforming pixel values of the first medical image to approximate values that would have been obtained if the first medical image had been acquired using the second type of imaging system such that said CAD code is applicable for the first medical image.

2. The computer program product of claim 1, wherein said transforming code transforms said pixel values using a first lookup table computed using said first and second response characteristics.

3. The computer program product of claim 2, further comprising a second lookup table used by said transforming code for transforming pixel values of a second medical image acquired from a third type of imaging system having a third response characteristic different than said first and second response characteristics to approximate values that would have been obtained if the second medical image had been acquired using the second type of imaging system, whereby said CAD code is also applicable for said second medical image despite having been acquired using a different type of imaging system than said training or first medical images.

4. The computer program product of claim 3, wherein said transforming code further comprises:

identifying code for identifying which type of imaging system was used to acquire each of said medical images;

a first remapping code for utilizing said first lookup table to remap medical images acquired using said first type of imaging system; and

a second remapping code for utilizing said second lookup table to remap medical images acquired using said third type of imaging system.

5. The computer program product of claim 4, wherein said identifying code further comprises code for reading DICOM headers associated with said medical images.

6. The computer program product of claim 1, wherein said first type of imaging system comprises a film-based detector, and wherein said second type of imaging system comprises a digital detector.

7. The computer program product of claim 1, wherein said first type of imaging system comprises a first film-based detector, and wherein said second type of imaging system comprises a second film-based detector different than said first film-based detector.

8. The computer program product of claim 1, wherein said first type of imaging system comprises a first digital detector, and wherein said second type of imaging system comprises a second digital detector different than said first digital detector.

9. The computer program product of claim 1, wherein said first type of imaging system comprises a digital detector, and wherein said second type of imaging system comprises a film-based detector.

10. A method for analyzing a first medical image to detect anatomical abnormalities therein using a computer-aided detection (CAD) algorithm including at least one neural network-based automated anatomical abnormality detection algorithm, the first medical image acquired from a first type of imaging system having a first response characteristic, the CAD algorithm trained using training images acquired from a second type of imaging system having a second response characteristic different than the first response characteristic, comprising:

transforming pixel values of the first medical image to approximate pixel values that would have been obtained if the first medical image had been acquired using the second type of imaging system such that the CAD code is applicable for said first medical image.

11. The method of claim 10, wherein said transforming pixel values comprises using a first lookup table designed using said first and second response characteristics.

12. The method of claim 11, further comprising using a second lookup table for transforming pixel values of a second medical image acquired from a third type of imaging system having a third response characteristic different than said first and second response characteristics to approximate values that would have been obtained if the second medical image had been acquired using the second type of imaging system such that said CAD code is also applicable for said second medical image.

13. The method of claim 11, further comprising:

identifying which type of imaging system was used to acquire each of said medical images;

applying said first lookup table for medical images acquired using said first type of imaging system; and

applying said second lookup table for medical images acquired using said third type of imaging system.

14. The method of claim 13, wherein said identifying further comprises reading DICOM headers associated with said medical images.

15. The method of claim 10, wherein said first type of imaging system comprises a film-based detector, and wherein said second type of imaging system comprises a digital detector.

16. The method of claim 10, wherein said first type of imaging system comprises a first film-based detector, and wherein said second type of imaging system comprises a second film-based detector different than said first film-based detector.

17. The method of claim 10, wherein said first type of imaging system comprises a first digital detector, and wherein said second type of imaging system comprises a second digital detector different than said first digital detector.

18. The method of claim 10, wherein said first type of imaging system comprises a digital detector, and wherein said second type of imaging system comprises a film-based detector.

19. A method for analyzing medical images to detect anatomical abnormalities therein, comprising:

receiving a medical image;

applying a computer-aided detection (CAD) algorithm to said medical image to detect anatomical abnormalities therein, wherein the CAD algorithm is a neural network-based automated anatomical abnormality detection algorithm; and

prior to said applying said CAD algorithm, performing the steps of;

identifying a first response characteristic associated with an acquisition of the medical image;

determining whether the first response characteristic is substantially similar to a second response characteristic associated with an acquisition of a plurality of training images used to train the CAD algorithm; and

if said first response characteristic is not substantially similar to said second response characteristic, transforming pixel values of said medical image to approximate pixel values that would have been obtained if the medical image had been acquired in a manner associated with said second response characteristic, such that said CAD code is applicable for the medical image.

20. The method of claim 19, wherein said first response characteristic is similar to a film-based image acquisition response characteristic, and wherein said second response characteristic is similar to a digital image acquisition response characteristic.

21. The method of claim 19, wherein said first response characteristic is similar to a first film-based image acquisition response characteristic, and wherein said second response characteristic is similar to a second film-based image acquisition response characteristic.

22. The method of claim 19, wherein said first response characteristic is similar to a first digital image acquisition response characteristic, and wherein said second response characteristic is similar to a second digital image acquisition response characteristic.

23. The method of claim 19, wherein said first response characteristic is similar to a digital image acquisition response characteristic, and wherein said second response characteristic is similar to a film-based acquisition response characteristic.