ECG-based Frame Selection and Curvature-based ROI Detection for Measuring Carotid Intima-Media Thickness

Jianming Liang (Inventor)

Research output: Patent


Carotid Intima-Media Thickness (CIMT) measurement is a sensitive, and highly reproducible technique for cardiovascular risk stratification. CIMT measurements are based on region of interest (ROI) in end-diastolic ultrasound frames (EUFs). Currently, ROIs and EUFs must be manually selected which is both tedious and time consuming, with variable accuracy and speed depending on the user's experience and level of expertise. There are some CIMT measurement programs available; however, with the increased demand for CIMT screening, a more user-friendly and time efficient solution is needed. Researchers at Arizona State University in collaboration with researchers from the Mayo Clinic have developed a novel software system for quickly measuring CIMT. This system automatically selects EUFs and determines ROIs in ultrasound videos, significantly cutting down on the time required for determining CIMT. This software system was evaluated on 23 subjects, each having 4 videos, with 3 EUFs selected in each video for a total of 272 ROIs. The results were compared with a reference provided by an expert for both frame selection and ROI detection and achieved 92.9% sensitivity and 97.6% specificity for EUF selection and 81.2% accuracy in ROI detection. This software system reduces user-dependency by automating and standardizing the CIMT measurement process. The results compare favorably to those achieved by an expert but with a system that is much more user friendly and significantly faster. Potential Applications Carotid intima-media thickness measurements for cardiovascular risk stratification Benefits and Advantages 92.9% sensitivity and 97.6% specificity for EUF selection 81.2% accuracy in ROI detection The selected frame may be manually modified if automatic detection is not optimal by a simple mouse click The edge line may be manually modified if the automatic detection is not optimal Download Original PDF For more information about the inventor(s) and their research, please see Dr. Liang's departmental webpage
Original languageEnglish (US)
Publication statusPublished - Dec 12 2013


Cite this