3D web-based annotation

1. A method comprising:

reading a three dimensional medical imaging study from a data store, the three dimensional medical imaging study comprising a first plurality of two dimensional medical images;

selecting a second plurality of two dimensional medical images from the first plurality of two dimensional medical images, the second plurality of two dimensional medical images being a strict subset of the first plurality of two dimensional medical images;

providing the second plurality of two dimensional medical images to a remote user;

receiving from the remote user regional annotations for each of the second plurality of medical images;

generating a volumetric description of the three dimensional imaging study by interpolation of the regional annotations;

providing the volumetric description of the three dimensional imaging for rendering and display to the remote user, wherein:

said selecting of the second plurality of two dimensional images minimizes an estimated error rate of the interpolation by selecting the second plurality of two dimensional images to represent clusters of the first plurality of two dimensional images, each cluster having a similar interslice propagation error, and

the estimated error rate is based on differences in intensity between a plurality of pairs of two dimensional images of the first plurality of two dimensional images.

2. The method of claim 1 , wherein the data store comprises a PACS.

3. The method of claim 1 , wherein selecting the second plurality of two dimensional medical images comprises applying k-means clustering to the first plurality of two dimensional medical images.

4. The method of claim 3 , wherein applying k-means clustering comprises clustering the first plurality of two dimensional medical images based on a minimization of the error rate.

5. The method of claim 3 , wherein applying k-means clustering comprises iteratively initializing centers of a plurality of clusters of the first plurality of two dimensional medical images and selecting a clustering having a minimum error rate.

6. The method of claim 1 , wherein the second plurality of two dimensional medical images corresponds to centers of clusters obtained by applying k-means clustering to the first plurality of two dimensional medical images.

7. The method of claim 1 , wherein the estimated error rate corresponds to a matching between manually segmented images and warped manually segmented images.

8. The method of claim 1 , wherein the error rate is based on a predetermined region of the first plurality of two dimensional images.

9. The method of claim 1 , further comprising:

providing a web application to the remote user, the web application being adapted to receive input of the regional annotations from the remote user.

10. The method of claim 1 , wherein the second plurality of two dimensional medical images is provided to the remote user via the internet.

11. The method of claim 1 , further comprising:

providing a web application to the remote user, the web application being adapted to render the volumetric description and display the rendering to the remote user.

12. A system comprising:

a data store;

a computing node comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor of the computing node to cause the processor to perform a method comprising:

reading a three dimensional medical imaging study from the data store, the three dimensional medical imaging study comprising a first plurality of two dimensional medical images;

selecting a second plurality of two dimensional medical images from the first plurality of two dimensional medical images, the second plurality of two dimensional medical images being a strict subset of the first plurality of two dimensional medical images;

providing the second plurality of two dimensional medical images to a remote user;

receiving from the remote user regional annotations for each of the second plurality of medical images;

generating a volumetric description of the three dimensional imaging study by interpolation of the regional annotations;

providing the volumetric description of the three dimensional imaging for rendering and display to the remote user, wherein

said selecting of the second plurality of images minimizes an estimated error rate of the interpolation by selecting the second plurality of two dimensional images to represent clusters of the first plurality of two dimensional images, each cluster having a similar interslice propagation error, and

the estimated error rate is based on differences in intensity between a plurality of pairs of two dimensional images of the first plurality of two dimensional images.

13. The system of claim 12 , wherein selecting the second plurality of two dimensional medical images comprises applying k-means clustering to the first plurality of two dimensional medical images.

14. The system of claim 13 , wherein applying k-means clustering comprises clustering the first plurality of two dimensional medical images based on a minimization of the error rate.

15. The system of claim 13 , wherein applying k-means clustering comprises iteratively initializing centers of a plurality of clusters of the first plurality of two dimensional medical images and selecting a clustering having a minimum error rate.

16. The system of claim 12 , wherein the second plurality of two dimensional medical images corresponds to centers of clusters obtained by applying k-means clustering to the first plurality of two dimensional medical images.

17. The system of claim 12 , wherein the estimated error rate corresponds to a matching between manually segmented images and warped manually segmented images.

18. The system of claim 12 , wherein the error rate is based on a predetermined region of the first plurality of two dimensional images.

19. A computer program product for medical imagery annotation, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor to cause the processor to perform a method comprising:

reading a three dimensional medical imaging study from a data store, the three dimensional medical imaging study comprising a first plurality of two dimensional medical images;

selecting a second plurality of two dimensional medical images from the first plurality of two dimensional medical images, the second plurality of two dimensional medical images being a strict subset of the first plurality of two dimensional medical images;

providing the second plurality of two dimensional medical images to a remote user;

receiving from the remote user regional annotations for each of the second plurality of medical images;

generating a volumetric description of the three dimensional imaging study by interpolation of the regional annotations;

providing the volumetric description of the three dimensional imaging for rendering and display to the remote user, wherein

said selecting of the second plurality of images minimizes an estimated error rate of the interpolation by selecting the second plurality of two dimensional images to represent clusters of the first plurality of two dimensional images, each cluster having a similar interslice propagation error, and

the estimated error rate is based on differences in intensity between a plurality of pairs of two dimensional images of the first plurality of two dimensional images.

20. The computer program product of claim 19 , wherein selecting the second plurality of two dimensional medical images comprises applying k-means clustering to the first plurality of two dimensional medical images.