Abstract: Detection of intravascular plaque in OCT images is carried out by obtaining images of vascular tissue from a vascular component by OCT either in a static mode of a single image or in a dynamic mode where the images are obtained by scanning. The method acts by dividing the OCT image into different regular regions, calculating different texture features for each of the above regions with a reduced set of less than a full set of the 26 Haralick textural features, using a clustering algorithm to segment the image defined by its texture features calculated above into different regions and transforming the segmented image back from its representation using texture features to its space-domain representation. The method uses three or four texture features where the reduced sets can be f1, f 2, and f14 (ASM at 0°, Inertia at 0° and ASM at 90°).

Abstract: Detection of intravascular plaque in OCT images is carried out by obtaining images of vascular tissue from a vascular component by OCT either in a static mode of a single image or in a dynamic mode where the images are obtained by scanning. The method acts by dividing the OCT image into different regular regions, calculating different texture features for each of the above regions with a reduced set of less than a full set of the 26 Haralick textural features, using a clustering algorithm to segment the image defined by its texture features calculated above into different regions and transforming the segmented image back from its representation using texture features to its space-domain representation. The method uses three or four texture features where the reduced sets can be f1, f 2, and f14 (ASM at 0°, Inertia at 0° and ASM at 90°).