Interactive coronary labeling using interventional X-ray images and deep learning
A method for classifying a vasculature comprises training a training device with an initial model of a vasculature using diagnostic image data representing a geometry for a plurality of vessels of a vessel tree and including a respective vessel labeling for each vessel, providing at least one diagnostic image of a patient's vessel tree and identifying a variation between the vessel tree represented by the initial model and the patients vessel tree. This variation is checked and labeled in order to improve the trained model. The process may be repeated iteratively until reaching an accurate patient-specific model of the vasculature.
1. A method for classifying a vasculature, the method comprising:
a) training a training device with an initial model of the vasculature using diagnostic image data representing a first vessel tree, the diagnostic image data comprising a corresponding vessel labeling for at least one vessel of the first vessel tree;
b) inputting at least one diagnostic image representing a second vessel tree;
c) identifying at least one deviation between the first vessel tree and the second vessel tree;
d) in response to the identifying, outputting an indication of the at least one deviation to a user and providing at least one labeling for the at least one deviation; and
e) adjusting, based on the at least one deviation and the at least one labeling, the initial model to classify the vasculature;
wherein the deviation comprises a variation between a geometry of the first vessel tree in the initial model and a geometry of the second vessel tree in the diagnostic image.
2. The method according to claim 1 , wherein the providing the at least one labeling comprises receiving a first user input from the user indicating the at least one labeling.
3. The method according to claim 1 , wherein the adjusting the initial model to classify the vasculature comprises iteratively repeating b) to e) for a plurality of diagnostic images.
4. The method according to claim 1 , wherein the adjusting the initial model comprises retraining the training device using the diagnostic image data representing the first vessel tree along with the at least one diagnostic image representing the second vessel tree and the at least one labeling.
5. The method according to claim 1 , further comprising identifying a geometry of the second vessel tree in the at least one diagnostic image by segmenting the at least one diagnostic image; identifying, based on the segmenting, respective centerline information for a plurality of vessels of the second vessel tree; and extracting the centerline information.
6. The method according to claim 5 , wherein the centerline information is input along with the at least one diagnostic image.
7. The method according to claim 1 , wherein the diagnostic image data comprises a plurality of images.
8. The method according to claim 1 , wherein the training device is further trained to include externally-caused variations into the initial model.
9. A classification system for classifying a vasculature, the classification system comprising:
a training device configured to be trained with an initial model of the vasculature using diagnostic image data representing a first vessel tree, the diagnostic image data comprising a corresponding vessel labeling for at least one vessel of the first vessel tree;
a processor configured to:
to receive at least one diagnostic image representing a second vessel tree;
identify at least one deviation between the first vessel tree and the second vessel tree;
output, in response to the identifying, an indication of the at least one deviation to a user; and
provide at least one labeling for the at least one deviation;
wherein the training device is configured to adjust, based on the at least one deviation and the at least one labeling, the initial model to classify the vasculature; and
wherein the deviation comprises a variation between a geometry of the first vessel tree and a geometry of the second vessel tree.
10. The classification system according to claim 9 , further comprising a display configured to display the indication of the at least one deviation to the user; and a user interface configured to receive a first user input from the user indicating the at least one labeling.
11. The classification system according to claim 9 , wherein the processor is further configured to generate a medical representation, the medical representation comprising at least the vasculature, the vessel labeling and the at least one labeling; and to transmit the medical representation to a database.
12. The classification system according to claim 11 , wherein the database comprises an electronic medical record (EMR); and wherein the medical representation is generated according to a pre-determined format of the electronic medical record.
13. A non-transitory computer-readable medium having stored a computer program comprising instructions to classify a vasculature, the instructions, when executed by at least one processor, cause the at least one processor to:
train a training device with an initial model of the vasculature using diagnostic image data representing a first vessel tree, the diagnostic image data comprising a corresponding vessel labeling for at least one vessel of the first vessel tree;
receive at least one diagnostic image representing a second vessel tree;
identify at least one deviation between the first vessel tree and the second vessel tree;
output, in response to the identifying, an indication of the at least one deviation to a user; and
provide at least one labeling for the at least one deviation;
wherein the training device is configured to adjust, based on the at least one deviation and the at least one labeling, the initial model to classify the vasculature; and
wherein the deviation comprises a variation between a geometry of the first vessel tree and a geometry of the second vessel tree.
14. The method according to claim 7 , wherein the plurality of images is between 100 and 10,000 images.
15. The method according to claim 14 , wherein the plurality of images is between 100 and 1000 images.