Traitement en cours

Veuillez attendre...

Paramétrages

Paramétrages

Aller à Demande

1. WO2020109006 - APPAREIL D'IDENTIFICATION DE RÉGIONS DANS UNE IMAGE CÉRÉBRALE

Note: Texte fondé sur des processus automatiques de reconnaissance optique de caractères. Seule la version PDF a une valeur juridique

[ EN ]

CLAIMS:

1. A regions identifying apparatus for identifying regions in an image showing a brain, wherein the regions identifying apparatus (1; 101) comprises:

an image providing unit (5; 105) configured to provide a three-dimensional image showing a brain of a living being,

a model providing unit (6) configured to provide a three-dimensional model (30) of a head of a living being, wherein the three-dimensional model (30) includes Alberta Stroke Program Early CT Score (ASPECTS) regions (34...37) of a brain,

a regions identifying unit (7) configured to identify ASPECTS regions in the three-dimensional image by applying the three-dimensional model (30) to the three-dimensional image.

2. The regions identifying apparatus as defined in claim 1, wherein the model providing unit (6) is configured to provide the three-dimensional model (30) such that it also includes at least one of the skull (31), the left hemisphere (33), the right hemisphere (32) and ventricles (38).

3. The regions identifying apparatus as defined in claim 2, wherein the model providing unit (6) is configured to provide the three-dimensional model (30) such that it comprises the left hemisphere (33) and the right hemisphere (32) of the brain, wherein the left hemisphere and the right hemisphere of the model (30) are symmetric to each other.

4. The regions identifying apparatus as defined in any of claims 2 and 3, wherein the model providing unit (6) is configured to provide the three-dimensional model (30) such that it comprises the left hemisphere (33) and the right hemisphere (32) of the brain, wherein a same structure in the left hemisphere (33) and the right hemisphere (32) of the brain has a same topology.

5. The regions identifying apparatus as defined in any of claims 1 to 4, wherein the model providing unit (6) is configured to provide the three-dimensional model (30) such that it also defines slices of the brain, wherein the regions identifying unit (7) is further configured to determine two-dimensional slices (20, 21) of the brain in the three-dimensional image based on the three-dimensional model (30) applied to the three-dimensional image and to identify the ASPECTS regions in the determined two-dimensional slices (20, 21).

6. The regions identifying apparatus as defined in claim 5, wherein the regions identifying unit (7) is configured to determine only two two-dimensional slices (20, 21) of the brain in the three-dimensional image based on the three-dimensional model (30) applied to the three-dimensional image and to identify the ASPECTS regions in the determined two two-dimensional slices.

7. The regions identifying apparatus as defined in any of claims 1 to 6, wherein the regions identifying apparatus (1) further comprises a score determination unit (8) configured to determine a score, which is indicative of ischemic changes in the brain, based on image values within the ASPECTS regions.

8. The regions identifying apparatus as defined in claim 7, wherein the score determination unit (8) is configured to compare, for each ASPECTS region, the image values with image values of a control person or a control population for determining the score.

9. The regions identifying apparatus as defined in claim 8, wherein the control person or the control population corresponds to a healthy brain, wherein the score determination unit (8) is configured to

compare the image values of a respective ASPECTS region of the living being with the image values of the respective ASPECTS region of the control person or the control population by applying a similarity measure to the image values,

determine candidate ASPECTS regions having image values which are not similar to the image values of corresponding ASPECTS regions of the control person or the control population given the similarity measure,

determine, among the candidate ASPECTS regions, candidate ASPECTS regions showing new ischemic stroke lesions,

provide an initial score value,

determine the score by subtracting, for each candidate ASPECTS region, which has been determined as showing a new ischemic stroke lesion, a predefined value from the provided initial score.

10. The regions identifying apparatus as defined in any of claim 7 to 9 and any of claims 5 and 6, wherein the score determination unit (8) is configured to determine a score, which is indicative of ischemic changes in the brain, based on image values within the ASPECTS regions in the determined two-dimensional slices.

11. The regions identifying apparatus as defined in any of claims 1 to 10, wherein the model providing unit (6) is configured to provide the three-dimensional model (30) of the brain such that it includes 10 ASPECTS regions (34...37), wherein the regions identifying unit (7) is configured to identify the 10 ASPECTS regions (34...37) in the three-dimensional image by applying the three-dimensional model (30) to the three-dimensional image.

12. The regions identifying apparatus as defined in any of claims 1 to 11, wherein the image providing unit (5; 105) is configured to provide a computed tomography image as the three-dimensional image.

13. A regions identifying method for identifying regions in an image showing a brain, wherein the regions identifying method comprises:

providing a three-dimensional image showing a brain of a living being by an image providing unit (5; 105),

providing a three-dimensional model (30) of a head of a living being by a model providing unit (6), wherein the three-dimensional model (30) includes

ASPECTS regions (34...37) of a brain,

identifying ASPECTS regions in the three-dimensional image by applying the three-dimensional model (30) to the three-dimensional image by a regions identifying unit (7).

14. A computer program for identifying regions in an image showing a brain, the computer program comprising program code means for causing a regions identifying apparatus (1) as defined in claim 1 to carry out the steps of the regions identifying method as defined in claim 13, when the computer program is run on a computer controlling the regions identifying apparatus.