The hardware and software evolution in the last 20 years becamepossible the dissemination of computer systems in several areas of thesociety. At the same time of this movement, the graphics processinghas expanded its influence field and the the medical area has been oneof the major benefit. With the growth of three-dimensional modelscurrently available in this area, efficient retrieval mechanisms arenecessary in order to assist health professionals in their diagnosis.From historical images similar to that related to the patient´s exams,the specialist can get a second opinion on the existence or absence ofan anomaly. This project aims to develop techniques for content-basedthree-dimensional medical image retrieval, generating a prototypesystem for queries. To reach the proposed objectives the followingsteps are planned: literature review, database definition, extractorsand similarity functions development, construction of a prototype forretrieving images, tests with medical imaging and result analysis. Itis expected that the results obtained from the retrieval cal berelevant for health professionals, helping them to compose diagnosisas well as these results can offer a contribution to Computer Sciencerelated to developing techniques in order to increase the content-based retrieval considering three-dimensional models context.
News published in Agência FAPESP Newsletter about the scholarship: