Fully Automated Segmentation and Tracking of the Intima Media Thickness in Ultrasound Video Sequences of the Common Carotid Artery

The robust measurement of the intima media thickness (IMT) of longitudinal common carotid artery (CCA) has an important clinical value because clinicians often use it as an important predictor to assess the possibility of potential cardiovascular events. The purpose of this study was to develop a fully automated algorithm to measure the IMT in the longitudinal ultrasound B-mode images. A completely automated algorithm for identification and calculation of IMT is proposed in this paper. Based on signal analysis, the algorithm can be divided into four steps. The first step is to automatically identify the lumen-intima (LI) interface points of posterior wall. Starting from the detected LI interface points, the second step uses the gradient-based method to locate the candidate media-adventitia (MA) interface points. The third step applies the canny edge detector to remove the outliers from the candidate points. The last step is to calculate the IMT from the final available points. On 35 ultrasound video sequences of the common carotid artery (CCA) taken from 13 healthy subjects, the results generated by the proposed method were compared to the manual annotated data. The proposed method yielded an IMT of 0.61 mm ± 0.085 (mean ± standard deviation) whereas the corresponding result yielded by the manual annotated ground truth data is 0.60 mm ± 0.1. The proposed method eliminates the need of manual initialization, and measures the IMT of the longitudinal CCA with high precision similar to the ones observed in the manual segmentations. It has the potential to be a suitable replacement for manual segmentation and measurement of the IMT.