Arterial wall characterization in optical coherence tomography imaging
A method, including: obtaining, by a processor, imaging data from a vessel; detecting, using the processor, an inner wall of the vessel based on the imaging data; identifying, using the processor, a plurality of visible edge portions of an outer wall of the vessel based on the imaging data; fitting, using the processor, a continuous surface model to the plurality of identified visible edge portions of the outer wall; and detecting, using the processor, the outer wall of the vessel based on fitting the continuous surface model to the plurality of identified visible edge portions of the outer wall such that the imaging data has defined therein a wall area between the inner wall and the outer wall of the vessel.
1. A method, comprising:
obtaining, by a processor, imaging data from a vessel;
detecting, using the processor, an inner wall of the vessel based on the imaging data;
identifying, using the processor, a plurality of visible edge portions of an outer wall of the vessel based on the imaging data;
fitting, using the processor, a continuous three-dimensional surface model to the plurality of identified visible edge portions of the outer wall,
the continuous three-dimensional surface model comprising a linear-elastic mesh model; and
detecting, using the processor, the outer wall of the vessel based on fitting the continuous three-dimensional surface model to the plurality of identified visible edge portions of the outer wall such that the imaging data has defined therein a wall area between the inner wall and the outer wall of the vessel.
2. The method of claim 1 , further comprising:
automatically classifying a tissue type of a portion of the wall area using a trained classification model.
3. The method of claim 2 , wherein the tissue type includes at least one of: calcium, lipid tissue, fibrous tissue, mixed tissue, or no visible tissue.
4. The method of claim 2 , wherein the trained classification model comprises a trained convolutional neural network (CNN).
5. The method of claim 4 , wherein the trained CNN is trained using a database of manually scored images.
6. The method of claim 1 , wherein the linear-elastic mesh model comprises an anisotropic linear-elastic mesh including a plurality of interconnected nodes and springs.
7. The method of claim 6 , wherein each of the plurality of interconnected nodes and springs is associated with a respective location of the vessel, and
wherein each of the nodes has a respective force value associated therewith.
8. The method of claim 1 , further comprising:
determining one or more physical parameters of the vessel based on the continuous three-dimensional surface model.
9. The method of claim 1 , further comprising:
classifying a portion of the wall area as non-pathological tissue based on a distance of the portion to at least one of the inner wall or the outer wall of the vessel.
10. The method of claim 1 , wherein the imaging data comprises interferometric data, and
wherein obtaining the imaging data from the vessel further comprises:
obtaining the interferometric data from an OCT probe.
11. The method of claim 1 , wherein detecting the inner wall of the vessel further comprises:
generating a longitudinal cross-sectional image of the vessel based on the imaging data, and
identifying the inner wall of the vessel by applying bilateral filtering and a K-means algorithm to the longitudinal cross-sectional image.
12. The method of claim 1 , wherein detecting the inner wall of the vessel further comprises:
thresholding the imaging data,
identifying, in the imaging data, a plurality of nonzero points closest to a center of the vessel for each of a respective plurality of radial positions, and
fitting lines or a surface to the plurality of nonzero points for each of the plurality of cross-sectional images to identify the inner wall of the vessel.
13. The method of claim 1 , further comprising:
transmitting information regarding the wall area to a user.
14. A system, comprising:
at least one hardware processor that is programmed to:
obtain imaging data from a vessel;
detect an inner wall of the vessel based on the imaging data;
identify a plurality of visible edge portions of an outer wall of the vessel based on the imaging data;
fit a continuous three-dimensional surface model to the plurality of identified visible edge portions of the outer wall,
the continuous three-dimensional surface model comprising a linear-elastic mesh model; and
detect the outer wall of the vessel based on fitting the continuous three-dimensional surface model to the plurality of identified visible edge portions of the outer wall such that the imaging data has defined therein a wall area between the inner wall and the outer wall of the vessel.
15. The system of claim 14 , wherein the processor is further to:
automatically classify a tissue type of a portion of the wall area using a trained classification model.
16. The system of claim 15 , wherein the tissue type includes at least one of: calcium, lipid tissue, fibrous tissue, mixed tissue, or no visible tissue.
17. The system of claim 15 , wherein the trained classification model comprises a trained convolutional neural network (CNN).
18. The system of claim 17 , wherein the trained CNN is trained using a database of manually scored images.
19. The system of claim 14 , wherein the linear-elastic mesh model comprises an anisotropic linear-elastic mesh including a plurality of interconnected nodes and springs.
20. The system of claim 19 , wherein each of the plurality of interconnected nodes and springs is associated with a respective location of the vessel, and
wherein each of the nodes has a respective force value associated therewith.
21. The system of claim 14 , further comprising:
determining one or more physical parameters of the vessel based on the continuous three-dimensional surface model.
22. The system of claim 14 , wherein the processor is further to:
classify a portion of the wall area as non-pathological tissue based on a distance of the portion to at least one of the inner wall or the outer wall of the vessel.
23. The system of claim 14 , wherein the imaging data comprises interferometric data, and
wherein the processor, when obtaining the imaging data from the vessel, is further to:
obtain the interferometric data from an OCT probe.
24. The system of claim 14 , wherein the processor, when detecting the inner wall of the vessel, is further to:
generate a longitudinal cross-sectional image of the vessel based on the imaging data, and
identify the inner wall of the vessel by applying bilateral filtering and a K-means algorithm to the longitudinal cross-sectional image.
25. The system of claim 14 , wherein the processor, when detecting the inner wall of the vessel, is further to:
threshold the imaging data,
identify, in the imaging data, a plurality of nonzero points closest to a center of the vessel for each of a respective plurality of radial positions, and
fit lines or a surface to the plurality of nonzero points for each of the plurality of cross-sectional images to identify the inner wall of the vessel.
26. The system of claim 14 , wherein the processor is further to:
transmit information regarding the wall area to a user.