IMAGE DATA FRAUD DETECTION SYSTEMS

• US 20110176712A1
• Filed: 07/24/2009
• Published: 07/21/2011
• Est. Priority Date: 07/25/2008
• Status: Active Grant

• Current Assignee: IXICO Technologies Limited (IXICO Plc)
• Original Assignee: IXICO Technologies Limited (IXICO Plc)
• Inventors: Hill, Derek Lionel Glendon, Hajnal, Joseph Vilmos, McLeish, Kate
• Granted Patent: US 9,298,877 B2
• Time in Patent Office: Days
• US Class Current: 382/128
• CPC Class Codes: G16H 30/40 for processing medical images, e.g. editing

Abstract

A system is described for detecting fraud from medical imaging data comprising image data and associated meta-data. The system comprises input means arranged to receive image data from at least one source, and processing means arranged to analyse the imaging data to determine whether it meets a fraud criterion, and if it does to generate a fraud indicator output.

Claims

1. A system for detecting fraud from medical imaging data including image data and associated metadata, the system comprising input means arranged to receive image data from at least one source, and processing means arranged to analyze the imaging data to determine whether it meets a fraud criterion, and if it does to generate a fraud indicator output.

2. A system according to claim 1 wherein the fraud criterion relates at least partially to the metadata and the processing means is arranged to analyze the metadata to determine whether the fraud criterion is met.

3. A system according to claim 1 wherein the fraud criterion relates at least partially to the image data and the processing means is arranged to analyze the image data to determine whether the fraud criterion is met.

4. A system according to claim 1 wherein the image data comprises values for each of a plurality of pixels or voxels that make up an image.

5. A system according to claim 1 wherein the processing means is arranged to compare at least one parameter of the image data with a corresponding reference parameter, and the fraud criterion relates to the similarity of the parameters.

6. A system according to claim 5 wherein the reference parameter is derived from reference image data.

7. A system according to claim 5 wherein the reference parameter is defined in the metadata.

8. A system according to claim 7 wherein the reference parameter is derived from a protocol including at least one of:

subject identity, acceptable ranges of scan time points, scanner type and scan type.

9. A system according to claim 6 wherein the reference image data is from a same trial as an image being analyzed.

10. A system according to claim 6 wherein the reference image data is from a different trial from an image being analyzed.

11. A system according to claim 5 wherein the processing means is arranged to compare one image with another using at least one of:

image registration;

image subtraction;

taking an intensity ratio and using an information theoretic based comparison.

12. A system according to claim 11 wherein the fraud criterion relates to residuals of the image subtraction.

13. A system according to claim 1 wherein the processing means is arranged to perform analysis of noise properties of an image being analyzed.

14. A system according to claim 13 wherein the fraud criterion is arranged to be indicative of image warping.

15. A system according to claim 11 where the image registration method is based on the use of b-splines to generate a model from cohorts.

16. A system according to claim 6 wherein the reference image data is from a same subject as an image being analyzed.

17. A system according to claim 16 wherein the processing means is arranged to compare all images indicated as being from single subject and to identify any inconsistencies indicating that they do not appear to be from the same subject.

18. A system according to claim 16 wherein the processing means is arranged to identify changes between images from the subject and to compare the changes with changes in the images from a plurality of reference subjects.

19. A system according to claim 1 arranged to present a user with a fraud probability rating for each subject, or each investigator in a trial.

20. A system according to claim 1 wherein the processing means is arranged to identify, in response to a fraud indicator output, a subset of the imaging data which met the fraud criterion and identify for further processing a group of imaging data that excludes the identified subset.

21. A system according to claim 20 wherein the identified subset includes all data relating to at least one of:

a specific subject;

a specific trial site;

a specific range of scan time points;

a specific scanner type; and

a specific scan type.

22. A system according to claim 20 wherein the processing means is arranged to produce respective clinical trial results from the imaging data and the identified subset of imaging data.

23. A system according to claim 22 wherein the processing means is arranged to compare the respective trial results to determine the significance of the excluded identified subset.

24. A system for detecting fraud from medical imaging data comprising data acquisition means arranged to acquire image data, to receive metadata relating to the image data, and to modify the image data in a manner determined by the metadata so as to code the metadata in the image data for subsequent extraction.

25. A system according to claim 24 further comprising processing means arranged to receive the metadata and the image data, to extract the coded metadata from the image data and compare it with the received metadata.

26. A system according to claim 24 wherein the image data comprises intensity data for each of a plurality of image elements, and the step of modifying the image data comprises modifying the intensity data.

27. A computer implemented method of detecting fraud from medical imaging data including image data and associated metadata, the method comprising receiving image data from at least one source, and analyzing the imaging data to determine whether it meets a fraud criterion, and if it does to generate a fraud indicator output.

28. A computer readable medium comprising computer interpretable instructions which, when executed on a computer apparatus, cause the computer apparatus to implement the method of claim 27.