AUTOMATIC-MEASUREMENT OF VERTEBRAL SHAPE USING ACTIVE SHAPE MODELS

In this paper, we describe how Active Shape Models (ASMs) have been us ed to accurately and robustly locate vertebrae in lateral Dual Energy X-ray Absorptiometry (DXA) images of the spine. DXA images are of low spatial resolution, and contain significant random and structural nois e, providing a difficult challenge for object location methods. All ve rtebrae in the image were searched for simultaneously, improving robus tness in location of individual vertebrae by making use of constraints on shape provided by the position of other vertebrae. We show that th e use of ASMs with minimal user interaction allows accuracy to be obta ined which is as good as that achievable by human operators, as well a s high precision. Having located each vertebra, it is desirable to eva luate whether it has been located sufficiently accurately for shape me asurements to be useful. We determined this on the basis of grey-level model fit, which was shown to usefully detect poorly located vertebra e, which should enable accuracy to be improved by rejecting proposed s earch solutions whose grey-level fit was poorer than a threshold. (C) 1997 Elsevier Science B.V.