Method for providing medical imaging decision support data and method for providing ground truth in 2D image space
One or more example embodiments of the present invention relates in one aspect to a computer-implemented method includes receiving 2D topogram data of a patient; generating 2D topogram annotation data by applying a machine learning algorithm for topogram analysis onto the 2D topogram data; generating the medical imaging decision support data based on the 2D topogram annotation data; and providing the medical imaging decision support data.
1. A computer-implemented method for providing a training pair for a training of a machine learning algorithm, the method comprising:
receiving 3D annotation data in relation to 3D image data of an examination region, the examination region comprising an anatomical structure, the 3D annotation data being indicative of a characteristic of the anatomical structure;
receiving 2D projection image data, the 2D projection image data being related to the 3D image data through a projection geometry;
generating 2D annotation data in relation to the 2D projection image data based on the projection geometry and the 3D annotation data, the 2D annotation data being indicative of the characteristic of the anatomical structure; and
providing the training pair, the training pair comprising the 2D projection image data and the 2D annotation data,
wherein the 2D projection image data is a CT topogram of the examination region.
2. The method of claim 1 , further comprising:
receiving the 3D image data; and
calculating the 3D annotation data in relation to the 3D image data by applying a 3D annotation algorithm onto the 3D image data.
3. The method of claim 2 , wherein
the 3D annotation data includes a 3D representation of the anatomical structure, and
the 2D annotation data includes a 2D representation of the anatomical structure.
4. The method of claim 1 , further comprising:
receiving the 3D image data; and
calculating the 2D projection image data based on the projection geometry and the 3D image data.
5. The method of claim 4 , wherein
the 3D annotation data includes a 3D representation of the anatomical structure, and
the 2D annotation data includes a 2D representation of the anatomical structure.
6. The method of claim 1 , wherein
the 3D annotation data includes a 3D representation of the anatomical structure, and
the 2D annotation data includes a 2D representation of the anatomical structure.
7. The method of claim 1 , wherein
the anatomical structure includes a set of vertebrae,
the 3D annotation data including, for each vertebrae of the set of vertebrae, a 3D representation of that vertebrae, and
the 2D annotation data comprising, for each vertebrae of the set of vertebrae, a 2D representation of that vertebrae.
8. The method of claim 7 , wherein
the 3D annotation data includes, for each vertebrae of the set of vertebrae, a quantitative bone density information, the quantitative bone density information being indicative of a value of a bone density of that vertebrae, and
the 2D annotation data includes, for each vertebrae of the set of vertebrae, the quantitative bone density information.
9. A computer-implemented method for training a machine learning algorithm for topogram analysis, the method comprising:
receiving a plurality of training pairs, each training pair of the plurality of training pairs being provided using the method of claim 1 ; and
training the machine learning algorithm based on the plurality of training pairs.
10. The method of claim 1 , wherein the 2D annotation data is automatically generated from the 3D image data.
11. The method of claim 10 , wherein the 2D annotation data is automatically generated based on one or more annotation algorithms available for 3D images.