Automatic Diagnosis of Atherosclerosis for Cerebrovascular Accident Prevention

Overview:

This project developed a method to aid in the diagnosis of atherosclerosis through ultrasound imaging analysis.

Main Steps:

• CCA detection – the CCA is located based on the whereabouts of the lumen centerline.
  • Allows to select the region of interest for further analysis.

• Intima-media region segmentation – the interfaces between the CCA wall layers (lumen-intima and intima-media) are segmented based on edge detection and dynamic programming optimization.
  • Allows to measure intima-media thickness (IMT), and degree of stenosis.

• Plaque segmentation – the plaque limits are located on the basis of medical consensus criteria, using the previously segmented intima-media regions
  • Allows to measure plaque burden
  • Allows further processing to evaluate plaque vulnerability and stroke risk

• Region normalization – the intensity of the plaque regions are normalized by locating reference media and lumen points.
  • Improve the repeatability of intensity-based markers characterizing plaque contents.