Stenosis localization

1 . A device for localizing a stenosis in an angiogram, the device comprising:

an image supply configured to provide a first plurality of angiographic images and a second plurality of treatment images;

wherein each of the first plurality of angiographic images comprises image data representative of a region of interest of a vascular structure, wherein the vascular structure comprises at least one vessel with at least a part of a stenosis, and;

wherein each of the second plurality of treatment images comprises image data representative of at least a part of an interventional device arranged within the vascular structure in a state when the stenosis of the vascular structure is treated; and

a data processor configured to:

select a second image from the second plurality of treatment images based on a detection of a property of the interventional device;

for at least one first image from the first plurality of angiographic images, identify a part-stenosis in the first image based on a direct identification of structures in the first image that are most similar to the interventional device as detected in the second image;

delineate the part-stenosis from the at least one first image; and

provide an identification of a full stenosis by merging together the part-stenosis for an annotation of the stenosis in at least one of the first image and second image.

2 . The device according to claim 1 , wherein, the data processor configured to identify a part-stenosis in the first image based on a direct identification of structures in the first image that are most similar to the interventional device as detected in the second image is further configured to identify the interventional device in the second image and to determine at least some of a plurality of geometric parameters relating to the interventional device comprising at least one of the group of shape, size, orientation, bending radius, diameter and direction.

3 . The device according to claim 2 , wherein, the data processor configured to identify a part-stenosis in the first image based on a direct identification of structures in the first image that are most similar to the interventional device as detected in the second image is further configured to analyze whether any of the plurality of geometric parameters relating to the interventional device can be found in vessel structures from the first image.

4 . The device according to claim 1 , wherein the data processor configured to identify the part-stenosis in the first image is configured to provide the identification of the part-stenosis in the first image based on the device-related content present in the second image in a non-registered image based identification procedure.

5 . The device according to claim 1 , wherein the data processor is configured to take the first image as it is and assess the first image in view of structural parameters relating to the interventional device as detected for the interventional device without geometrical transferring or registering procedure.

6 . The device according to claim 1 , wherein the data processor is further configured to provide a segmentation of the part-stenosis in the first image based on device-related content present in the second image.

7 . The device according to claim 1 , wherein the second image comprises image data of the interventional device arranged at least partly in a treatment configuration of the stenosis.

8 . The device according to claim 1 , wherein the interventional device is a balloon device for treating stenosis; and wherein the second image comprises image data of the balloon device arranged in an at least partly inflated state.

9 . The device according to claim 1 , wherein the data processor is configured to provide a self-learning algorithm to be trained with a plurality of learning pairs constituted by the first and the second images as input to the learning algorithm and by a footprint of the stenosis as characterization of the stenosis as the target output of the learning algorithm.

10 . The device according to claim 1 , wherein the data processor comprises a convolutional network configuration; and

wherein the convolutional network configuration is provided with learning pairs of first and second images constituted by the first and the second images as input to the learning algorithm and by a footprint of the stenosis as characterization of the stenosis as target output of the learning algorithm; and the convolutional network configuration is configured for a self-learning process to learn the relationship between the first and second images.

11 . The device according to claim 1 , wherein a device segmentor is provided that is configured to provide a segmentation of the interventional device; and

wherein the data processor is configured to provide the identification of the stenosis in the first image based on the segmentation of the interventional device.

12 . The device according to claim 1 , wherein:

the image supply is configured to provide: a plurality of first images each comprising a part-stenosis; the data processor is configured to delineate or segment each part-stenosis and to merge the several stenosis parts to form a resulting stenosis; and

the data processor is configured to provide an indication of the resulting stenosis; and/or

the image supply is configured to provide a plurality of second images each comprising a part-treated image; and the data processor is configured to merge the part-treated images for annotating the image data representative of at least a part of an interventional device.

13 . A medical system for annotating medical images of stenosis treatment, the system comprising:

a device for localizing a stenosis in an angiogram according to claim 1 ; and

an image acquisition device comprising an X-ray imaging arrangement with an Xray source and an X-ray detector configured to provide at least one of the first and second image to the image supply unit.

14 . A method for localizing a stenosis in an angiogram, the method comprising:

providing a first plurality of angiographic images and a second plurality of treatment images, wherein each of the first plurality of angiographic images comprises image data representative of a region of interest of a vascular structure, wherein the vascular structure comprises at least one vessel with at least a part of a stenosis, and wherein each of the second plurality of treatment images comprises image data representative of at least a part of an interventional device arranged within the vascular structure in a state when the stenosis of the vascular structure is treated;

selecting a second image from the second plurality of treatment images based on a detection of a property of the interventional device

for at least one first image from the first plurality of angiographic images, identifying a part-stenosis in the first image based on a direct identification of structures in the first image that are most similar to the interventional device as detected in the second image;

delineating the part-stenosis from the at least one first image; and

providing an identification of a full stenosis by merging together the part-stenosis for an annotation of the stenosis in at least one of the first image and second image.

15 . A non-transitory computer-readable storage medium having stored a computer program comprising instructions, which, when executed by a processor, cause the processor to:

receive a first plurality of angiographic images and a second plurality of treatment images, wherein each of the first plurality of angiographic images comprises image data representative of a region of interest of a vascular structure, wherein the vascular structure comprises at least one vessel with at least a part of a stenosis, and wherein each of the second plurality of treatment images comprises image data representative of at least a part of an interventional device arranged within the vascular structure in a state when the stenosis of the vascular structure is treated;

select a second image from the second plurality of treatment images based on a detection of a property of the interventional device

for at least one first image from the first plurality of angiographic images, identify a part-stenosis in the first image based on a direct identification of structures in the first image that are most similar to the interventional device as detected in the second image;

delineate the part-stenosis from the at least one first image; and

provide an identification of a full stenosis by merging together the part-stenosis for an annotation of the stenosis in at least one of the first image and second image.