| 摘要 |
An automatic or semiautomatic method of processing intravascular Optical Coherence Tomography (OCT) images, comprising: pre-processing the primary image by segmenting the guidewire in the lumen to exclude it from the image, determining whether the lumen is open or closed; if the lumen is determined to be closed, performing a first automatic lumen segmentation procedure on the pre-processed image to define a perimeter line of the lumen, corresponding to the inner edge of the vessel wall, and determine a centroid CL of the lumen area; if the lumen is determined not to be closed, sequentially performing an automatic lumen closing procedure on the pre-processed image, and a second automatic lumen segmentation procedure, to define a perimeter line of the lumen, corresponding to the inner edge of the vessel wall, and determine a centroid CL of the lumen area; automatically finding an outer edge of the vessel wall and defining a first region of interest ROI1 between the inner edge and the outer edge of the lumen; automatically segmenting a fibrous plaque in the first region of interest ROI1 thereby defining a second region of interest ROI2 that contains the fibrous plaque, and segmenting a hyporeflective plaque, contained in a radial region that falls within ROI1 and is external to ROI2. |
