Method and apparatus for generating a universal atlas database

1. A method of generating an atlas for a global standardised atlas database, the method comprising:

a) providing a medical scan image;

b) performing a universal auto-contouring operation using the global standardised atlas database on the medical scan image, to generate a set of universal contours for the medical scan image;

c) performing an institution specific auto-contouring customisation operation on the medical scan image, to generate a set of local contours for the medical scan image, approval or editing of the set of local contours, and adding a local atlas to the local atlas database;

d) standardising the set of local contours, using a trained model to compensate for biases in the set of local contours, thereby creating a set of standardised global contours for the medical scan image;

and adding an atlas to the global standardised atlas database, the atlas comprising the medical scan image and the set of standardised global contours for the medical scan image.

2. The method of claim 1 , further comprising:

The institution specific auto-contouring customisation operation comprising adaptation of the universal contours to local guidelines of a local institution; and

the local institution providing the medical scan image and the local atlas database, and performing the universal auto-contouring operation and the institution specific auto-contouring customisation operation.

3. The method of claim 1 ,

wherein the local atlas added to the local atlas database comprising the medical scan image and the set of local contours, after the approval or editing.

4. The method of claim 1 , further comprising:

repeating the method for a plurality of medical scan images, thereby creating a set of standardised global contours for each medical scan image; and

adding, to the global standardised atlas database, an atlas for each of the plurality of medical scan images, each atlas comprising the medical scan image and the set of standardised global contours for the medical scan image.

5. The method of claim 1 , further comprising:

adding the first atlas to a second local database, the second local database being available only to the local institution.

6. The method of claim 1 , further comprising:

using the set of universal contours in a training step, the training step comprising comparing the set of universal contours to the set of local contours, in order to further optimise the trained model that performs the standardising of the set of local contours.

7. The method of claim 1 , further comprising:

performing the universal auto-contouring operation on the medical scan image with input from the global standardised atlas database.

8. The method of claim 7 , wherein:

the input from the global standardised atlas database to the universal auto-contouring operation comprises a stored atlas from the global standardised atlas database.

9. The method of claim 8 , wherein:

the input from the global standardised atlas database to the universal auto-contouring operation comprises a set of variations between a plurality of atlases stored in the global standardised atlas database.

10. The method of claim 1 , further comprising:

performing the institution specific auto-contouring customisation operation on the medical scan image with input from a local atlas database.

11. The method of claim 10 , wherein:

the input from the local atlas database to the institution specific auto-contouring customisation operation comprises a local atlas from the local atlas database.

12. The method of claim 1 , wherein the universal auto-contouring operation comprises at least one from, or a combination of:

an atlas-based operation; a shape/appearance model; a machine learning approach; or a deep learning algorithm.

13. An apparatus for generating an atlas for a global standardised atlas database, the apparatus comprising:

a memory, the memory configured to store a medical scan image;

a processor, the processor configured to:

a) perform a universal auto-contouring operation using the global standardised atlas database on the medical scan image, thereby generating a set of universal contours for the medical scan image;

b) perform an institution specific auto-contouring customisation operation on the medical scan image, thereby generating a set of local contours for the medical scan image; approval or editing of the set of local contours, adding a local atlas to the local atlas database:

c) standardise the set of local contours, using a trained model to compensate for biases in the set of local contours, thereby creating a set of standardised global contours for the medical scan image, and adding an atlas to the global standardised atlas database, the atlas comprising the medical scan image and the set of standardised global contours for the medical scan image.