Method and apparatus for automated determination of contours in iterative reconstruction of image data

1. A method for reconstructing image data and textural features of an examination region to be mapped, comprising:

providing scan projection data to a processor, acquired from the examination region to be mapped;

in said processor, implementing an iterative reconstruction procedure on the acquired scan projection data in order to produce a reconstructed image of the examination region at an end of said iterative reconstruction procedure;

in said iterative reconstruction procedure in said processor, reconstructing preliminary image data and preliminary textural features, determining anticipated preliminary structures in the examination region based on said preliminary textural features, and thereafter iteratively reconstructing said reconstructed image dependent on the anticipated preliminary structures and the preliminary textural features; and

making the reconstructed image available from the processor as an electronic output comprising a data file.

2. A method as claimed in claim 1 comprising determining said anticipated preliminary features in the examination region based on at least one of reconstructed preliminary image data and a structure prediction model.

3. A method as claimed in claim 1 wherein said iterative reconstruction comprises a plurality of iterative steps, and repeating said iterative steps until a quality criterion is satisfied for said preliminary image data and said preliminary textural features.

4. A method as claimed in claim 1 comprising determining said anticipated preliminary structure based on a structure prediction model, and generating said structure prediction model by:

providing training scan projection data to the processor, in said processor, implementing an iterative reconstruction of preliminary training image data based on said training scan projection data, and determining preliminary training textural features during the iterative reconstruction of said preliminary training image data;

in said processor, annotating structures in said preliminary training image data;

in said processor, implementing a renewed iterative reconstruction of new preliminary training image data and determining preliminary training textural features dependent on the annotated structures;

in said processor, classifying most recently reconstructed preliminary training image data and most recently obtained preliminary training textural features, as definitive training image data and definitive training textural features; and

using the definitive training image data and the definitive training textural features to generate said structure prediction model.

5. A method as claimed in claim 4 comprising repeating annotation of said structures and renewed iterative reconstruction of said preliminary training image data and renewed determining of said preliminary training textural features, until a quality criterion is satisfied for at least one of the determined preliminary training image data and the determined preliminary training textural features.

6. A method as claimed in claim 4 comprising determining said structure prediction model using a procedure selected from the group consisting of a machine learning procedure and a deep-learning procedure, based on the definitive training image data and the definitive training textural features.

7. A method as claimed in claim 1 wherein said textural features comprise edge images.

8. A method as claimed in claim 1 wherein said textural features comprise textural feature vectors.

9. A method as claimed in claim 1 wherein said textural features are contours.

10. An image reconstruction apparatus and textural features of an examination region to be mapped, comprising:

a processor provided with scan projection data acquired from the examination region to be mapped;

said processor being configured to implement an iterative reconstruction procedure on the acquired scan projection data in order to produce a reconstructed image of the examination region at an end of said iterative reconstruction procedure;

in said iterative reconstruction procedure, said processor being configured to reconstruct preliminary image data and preliminary textural features, determine anticipated preliminary structures in the examination region based on said preliminary textural features, and thereafter iteratively reconstruct said reconstructed image dependent on the anticipated preliminary structures and the preliminary textural features; and

being configured to make the reconstructed image available from the processor as an electronic output comprising a data file.

11. A computed tomography apparatus comprising:

a computed tomography scan projection data acquisition scanner that acquires projection scan data from an examination region to be mapped;

a processor provided with said scan projection data;

said processor being configured to implement an iterative reconstruction procedure on the acquired scan projection data in order to produce a reconstructed image of the examination region at an end of said iterative reconstruction procedure;

in said iterative reconstruction procedure, said processor being configured to reconstruct preliminary image data and preliminary textural features, determine anticipated preliminary structures in the examination region based on said preliminary textural features, and thereafter iteratively reconstruct said reconstructed image dependent on the anticipated preliminary structures and the preliminary textural features; and

being configured to make the reconstructed image available from the processor as an electronic output comprising a data file.

12. A non-transitory, computer-readable data storage medium encoded with programming instructions, said storage medium being loaded into a computer and said programming instructions causing said computer to:

receive scan projection data acquired from an examination region to be mapped;

implement an iterative reconstruction procedure on the acquired scan projection data in order to produce a reconstructed image of the examination region at an end of said iterative reconstruction procedure;

in said iterative reconstruction procedure, reconstruct preliminary image data and preliminary textural features, determine anticipated preliminary structures in the examination region based on said preliminary textural features, and thereafter iteratively reconstruct said reconstructed image dependent on the anticipated preliminary structures and the preliminary textural features; and

make the reconstructed image available from the processor as an electronic output comprising a data file.