Content-based retrieval for dental image database

In this work, we report a prototype of a clinical radiograph image database to be indexed by image content. User-identified regions, such as a lesion in a dental radiograph, are matched automatically to similar image regions in a large clinical database. The underlying content-based search engine is based on modal shape descriptions, a physically-motivated method which describes objects in terms of generalized symmetries, as defined by each object's eigenmodes. Modes provide a global-to-local ordering of shape deformation and is used for object recognition and categorization, where similarities are expressed as the amounts of modal deformation energy needed to align two shapes. We investigated the statistical characteristics of modal shape description, and obtained positive evidence for applying modal shape description in classifying non-rigid shape deformations due to different pathologies reflected in the dental image. The system is implemented using WWW technologies, including a client-server system implemented using communication between Java applets and Java server application process. Such system can facilitate diagnosis process in both medical practice and education, and particularly fit for telemedicine scenario.