A Novel 3D Shape Descriptor for Automatic Retrieval of Anatomical Structures from Medical Images

摘要

Content-based image retrieval (CBIR) aims at retrieving from a database objects that are similar to an object provided by a query, by taking into considerationa set of extracted features.While CBIR has been widely appliedin the two-dimensional image domain, theretrieval of3D objectsfrom medical image datasetsusing CBIR remains to be explored. In this context, the development of descriptors that can capture information specific to organs or structures is desirable, In this work, we focus on the retrieval of two anatomical structures commonly imaged by Magnetic Resonance Imaging (MRI) and Computed Tomography (CT)techniques, the left ventricle of the heart and blood vessels. Towards this aim, we developed the Area-Distance Local Descriptor (ADLD), a novel 3D local shape descriptorthatemploysmesh geometry information,namelyfacet area and distance from centroid to surface, to identify shape changes. Because ADLDonly considers surface meshesextractedfrom volumetric medical images,itsubstantially diminishesthe amount of data to be analyzed. A 90% precision rate was obtainedwhenretrieving both convex (left ventricle) and non-convex structures (blood vessels), allowing for detection of abnormalities associated with changes in shape. Thus, ADLD has the potential to aidin the diagnosis of a wide range of vascular and cardiac diseases.