Registration of renal SPECT and 2.5D US images
Résumé
The main goal of this study is the registration of renal SPECT (Single Photon Emission Computerized Tomography) and 2.5D US (Ultrasound) images. In the proposed approach, the matching is performed after kidney segmentation in both images. The SPECT segmentation is achieved using a deformable model based on a simplex mesh. And the 2.5D US image segmentation is carried out in every 2D slice by mean of a deformable contour. Next, the registration is carried out using a nonlinear optimization algorithm, and this registration was also used to correct the movements in the US image caused by the patient respiration during the acquisition. The registration was evaluated quantitatively comparing the distance between a manual segmentation in the US image and the model extracted from de SPECT image. Qualitative expertise is currently been realized.