Multimodality Segmentation and Registration of 3D Pulmonary Exams

Lara Bolito Abrantes
University of Coimbra
Portugal

José Silvestre Silva
University of Coimbra
Portugal

Abstract:
Several medical imaging techniques are used to perform the Pulmonary Embolism diagnose. Multimodality imaging is usually used to achieve better morphological details. In this work we implemented and compared 3D registration algorithms, after the segmentation of thoracic CT exam, ventilation and perfusion SPECT exams. The registration routine is applied using rigid body transformations, affine transformations and B-spline algorithm. In order to evaluate the registration performance, similarity metrics are used.