A multi-scale method for automatically extracting the dominant features of cervical vertebrae in CT images

Abstract

Abstract-Localization of the dominant points of cervical spines in medical images is important for improving the medical automation in clinical head and neck applications. In order to automatically identify the dominant points of cervical vertebrae in neck CT images with precision, we propose a method based on multi-scale contour analysis to analyzing the deformable shape of spines. To extract the spine contour, we introduce a method to automatically generate the initial contour of the spine shape, and the distance field for level set active contour iterations can also be deduced. In the shape analysis stage, we at first coarsely segment the extracted contour with zero-crossing points of the curvature based on the analysis with curvature scale space, and the spine shape is modeled with the analysis of curvature scale space. Then, each segmented curve is analyzed geometrically based on the turning angle property at different scales, and the local extreme points are extracted and verified as the dominant feature points. The vertices of the shape contour are approximately derived with the analysis at coarse scale, and then adjusted precisely at fine scale. Consequently, the results of experiment show that we approach a success rate of 93.4% and accuracy of 0.37mm by comparing with the manual results.

Automatic initialization for segmentation of medical images based on active contour

“…It has also been shown that the developed strategy outperforms the existing local binary fitting model. Identification of dominant features in cervical vertebrae has been reported in [10] which uses the geodesic active contours for extracting the ROI. The initialization is again through Otsu's threshold and a series of morphological operations for eliminating the air path which is shown as small blobs in the thresholded image.…”

Automatic segmentation of vertebral contours from CT images using fuzzy corners

“…The level set is formulated in terms of intensities by adding a weighted kernel to each pixel, as opposed to [10], where the edge stopping function is defined in terms of Euler Lagrange coefficients. Wu and Lin were able to identify dominant features in cervical vertebrae with a similar thresholdbased initialization approach [11]. To overcome the inherent drawbacks of thresholding, an adaptive 3D region growing method is devised in [12].…”